WO2025231291A1

WO2025231291A1 - Photo-based monitoring of dental occlusion

Info

Publication number: WO2025231291A1
Application number: PCT/US2025/027373
Authority: WO
Inventors: Chao Shi; Christopher E. Cramer; Yun Gao; Chad Clayton Brown; Guotu Li; Elena AGILINA; Mitra Derakhshan; Jeeyoung Choi; John Y. Morton; Irina Ivanova
Original assignee: Align Technology Inc
Current assignee: Align Technology Inc
Priority date: 2024-05-01
Filing date: 2025-05-01
Publication date: 2025-11-06
Anticipated expiration: 2026-11-01

Abstract

A system includes a computing device comprising a memory and one or more processors. The computing device is configured to access a first image of a patient's dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; process the first image to generate a second image of the patient's dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image; perform one or more oral diagnostics measurements of the patient's dentition using the new contours; and output a result of the one or more oral diagnostics measurements.

Description

PHOTO-BASED MONITORING OF DENTAL OCCLUSION

TECHNICAL FIELD

[0001] The instant specification generally relates to systems and methods for characterizing a dental occlusion and/or bite classification based on image data of a patient’s dentition.

BACKGROUND

[0002] People may develop different types of dental occlusions, including class I malocclusion, class II malocclusion, class III malocclusion, crossbite, deep bite, open bite, and so on. Traditionally, such malocclusions are diagnosed by a doctor based on patient examination and direct measurement of the patient’s dentition.

[0003] Virtual Care aims to provide photo-based remote oral diagnostics based on patient- provided 2D images. Unfortunately, it is challenging to infer various types of oral conditions from 2D images, especially when some teeth are only partially visible in provided images. Furthermore, it can be difficult to identify a full distribution of malocclusions from images of patients undergoing treatment.

SUMMARY

[0004] The below summary is a simplified summary of the disclosure in order to provide a basic understanding of some aspects of the disclosure. This summary is not an extensive overview of the disclosure. It is intended neither to identify key or critical elements of the disclosure, nor delineate any scope of the particular embodiments of the disclosure or any scope of the claims. Its sole purpose is to present some concepts of the disclosure in a simplified form as a prelude to the more detailed description that is presented later.

[0005] In a first example implementation, a method for characterizing a dental occlusion of a patient comprises: receiving a first image of a patient’s dentition depicting upper and lower jaws of the patient in a bite-open arrangement and a second image of the patient’s dentition depicting the upper and lower jaws of the patient in a bite-closed arrangement; determining one or more first digital measurements of the patent’s dentition from the first image and one or more second digital measurements of the patient’s dentition from the second image; and characterizing a level of the dental occlusion between opposing teeth of the upper and lower jaws of the patient based at least in part on the one or more first digital measurements and the one or more second digital measurements.

[0006] In a second example implementation, a method for measuring a bite classification comprises: receiving a bite-closed image of a patient’s dentition; registering one or more 3D models of the patient’s dentition to the image; identifying a first reference point on the bite-closed image associated with one or more maxillary teeth of the patient; identifying a second reference point on the bite-closed image associated with one or more mandibular teeth of the patient; identifying, within an image space of the bite-closed image, a virtual marker corresponding to a distance between the first reference point and the second reference point; projecting the virtual marker from the image space to the one or more 3D models; determining a physical measurement corresponding to the projected virtual marker; and determining the bite classification for the patient based on the physical measurement.

[0007] In a third example implementation, a method for measuring a bite classification comprises: receiving, from an image sensor, image data comprising a bite-closed image of a patient’s upper dental arch and lower dental arch; registering a first 3D model of the patient’s upper dental arch to the bite- closed image; registering a second 3D model of the patient’s lower dental arch to the bite-closed image; identifying a first reference point on the first 3D model; identifying a second reference point on the second 3D model; determining a physical measurement corresponding to a distance between the first reference point and the second reference point; and determining the bite classification for the patient based on the physical measurement.

[0008] In a fourth example implementation, a method for measuring a bite classification comprises: receiving, from an image sensor, image data comprising a bite-closed image of a patient’s upper dental arch and lower dental arch; processing the image data using a trained machine learning model, wherein the trained machine learning model outputs an estimate of a physical measurement corresponding to a distance between a first reference point on the patient’s upper dental arch and a second reference point on the patient’s lower dental arch; and determining the bite classification for the patient based on the physical measurement.

[0009] In a fifth example implementation, a method for measuring an amount of posterior crossbite comprises: receiving a bite-closed image of a patient’s upper dental arch and lower dental arch; segmenting the bite-closed image into a plurality of teeth; measuring a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient in the bite- closed image; determining a first ratio between the first tooth height and the second tooth height; determining a second ratio between a third tooth height of the maxillary tooth and a fourth tooth height of the mandibular tooth in one or more 3D models of the patient’s upper dental arch and lower dental arch; determining a difference between the first ratio and the second ratio; and determining the amount of posterior crossbite based on the difference.

[0010] In a sixth example implementation, a method for measuring an amount of posterior crossbite comprises: receiving a bite-closed image and an bite-open image of a patient’s upper dental arch and lower dental arch; segmenting the bite-closed image and the bite-open image into a plurality of teeth; measuring a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient in the bite-closed image; determining a first ratio between the first tooth height and the second tooth height; measuring a third tooth height of the maxillary tooth of the patient and a fourth tooth height of the opposing mandibular tooth of the patient in the bite-open image; determining a second ratio between the third tooth height and the fourth tooth height; determining a difference between the first ratio and the second ratio; and determining the amount of posterior crossbite based on the difference.

[0011] In a seventh example implementation, a method comprises: receiving a first image of a patient’s dentition, the first image comprising a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth; processing the first image of the patient’s dentition to generate a second representation of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the first image; performing one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth that is occluded in the first image; and outputting a result of the one or more oral diagnostics measurements.

[0012] In an eighth example implementation, s method comprises: receiving a first image of a patient’s dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; determining one or more three-dimensional (3D) models of the patient’s dentition; projecting the one or more 3D models onto an image plane defined by the first image to generate a second image comprising a second representation of at least the one or more second teeth, wherein the contours of the more second teeth that are occluded by the one or more first teeth in the first image are shown in the second image; and training an artificial intelligence (Al) model using the first image as an input and the second image as a target output, wherein the Al model is trained to process input images comprising representations of dentition including occluded teeth and to generate output images showing contours of the occluded teeth that are occluded in the input images.

[0013] In a nineth example implementation, a method comprises: receiving a first image of a patient’s dentition, the first image comprising a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth; processing the first image of the patient’s dentition to generate a second representation of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the first image; performing one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth that is occluded in the first image; determining whether to recommend palatal expansion treatment based on the one or more oral diagnostics measurements of the patient’s dentition; and outputting a recommendation with respect to palatal expansion treatment.

[0014] A tenth example implementation may further extend any of the first through nineth example implementations. In the tenth example implementation, a non-transitory computer readable medium comprises instructions that, when executed by a processing device, cause the processing device to perform the method of any of the first through sixth implementations.

[0015] An eleventh example implementation may further extend any of the first through nineth example implementations. In the eleventh example implementation, a computing device comprises: a memory configured to store instructions; and a processing device configured to execute the instructions from the memory to perform the method of the first through sixth implementations.

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Aspects and embodiments of the present disclosure will be understood more fully from the detailed description given below and from the accompanying drawings of various aspects and embodiments of the disclosure, which, however, should not be taken to limit the disclosure to the specific aspects or embodiments, but are for explanation and understanding only.

[0017] FIG. 1A illustrates an example process for characterizing a dental occlusion, according to some embodiments of the present disclosure.

[0018] FIG. 1 B is a flow chart of an example method for characterizing a dental occlusion of a patient, according to some embodiments of the present disclosure.

[0019] FIG. 1C is a flow chart of a first example method for measuring a level of malocclusion between opposing teeth of the upper and lower jaws of a patient, according to some embodiments of the present disclosure.

[0020] FIG. 1 D is a flow chart of a second example method for measuring a level of malocclusion between opposing teeth of the upper and lower jaws of a patient, according to some embodiments of the present disclosure.

[0021] FIG. 2A illustrates an example bite-open anterior image used by the processes of FIGS.

1A-D, according to some embodiments of the present disclosure.

[0022] FIG. 2B illustrates an example bite-closed anterior image used by the processes of FIGS. 1A-D, according to some embodiments of the present disclosure.

[0023] FIG. 3 illustrates an example correction subprocess, according to some embodiments of the present disclosure.

[0024] FIG. 4 illustrates an example orientation of a camera to a dental arch of a patient, according to some embodiments of the present disclosure. [0025] FIG. 5 illustrates an example process for characterizing a dental occlusion, according to some embodiments of the present disclosure.

[0026] FIG. 6 illustrates a flow diagram of an example method for determining a bite classification for a patient from image data of the patient, in accordance with some embodiments of the present disclosure.

[0027] FIG. 7A illustrates an example bite-closed side view image of a patient, according to some embodiments of the present disclosure.

[0028] FIG. 7B illustrates an example bite-closed side view image of a patient used by the method of FIG. 6, according to some embodiments of the present disclosure.

[0029] FIG. 8 illustrates a flow diagram of an example method for determining a bite classification for a patient from image data of the patient, in accordance with some embodiments of the present disclosure.

[0030] FIG. 9 illustrates a flow diagram of an example method for determining a bite classification for a patient from image data of the patient using one or more trained machine learning model, in accordance with some embodiments of the present disclosure.

[0031] FIG. 10 illustrates a flow diagram of an example method for determining an amount of posterior crossbite for a patient from image data of the patient, in accordance with some embodiments of the present disclosure.

[0032] FIG. 11 illustrates a flow diagram of an example method for determining an amount of crossbite for a patient from a lateral image of a patient, in accordance with some embodiments of the present disclosure.

[0033] FIG. 12 illustrates a flow diagram of an example method for determining an amount of crossbite for a patient from an anterior image of a patient, in accordance with some embodiments of the present disclosure.

[0034] FIG. 13 illustrates an example process for characterizing a dental occlusion using generative techniques, according to some embodiments of the present disclosure.

[0035] FIG. 14 illustrates a flow diagram of an example method for performing oral diagnostics measurements of a patient’s dentition, in accordance with some embodiments of the present disclosure.

[0036] FIG. 15A illustrates a flow diagram of an example oral diagnostics measurement, in accordance with some embodiments of the present disclosure.

[0037] FIG. 15B illustrates a flow diagram of an example bite class assessment, in accordance with some embodiments of the present disclosure. [0038] FIG. 16A illustrates a flow diagram of an example method for assessing tooth crowding and recommending dental treatment, in accordance with some embodiments of the present disclosure. [0039] FIG. 16B illustrates a flow diagram of an example method for determining whether to recommend palatal expansion treatment, in accordance with some embodiments of the present disclosure.

[0040] FIG. 17 illustrates a flow diagram of an example method for determining whether to recommend palatal expansion treatment, in accordance with some embodiments of the present disclosure.

[0041] FIG. 18A illustrates an example overbite measurement process using a single input image, according to some embodiments of the present disclosure.

[0042] FIG. 18B illustrates an example overbite measurement process using pair of input images, according to some embodiments of the present disclosure.

[0043] FIG. 18C illustrates an example crowding measurement process using a single input image, according to some embodiments of the present disclosure.

[0044] FIG. 18D illustrates an example measurement of a synthetic image, according to some embodiments of the present disclosure.

[0045] FIG. 19 illustrates a flow diagram of an example method for training an Al model to generate modified images of dentition usable for oral diagnostics measurements, in accordance with some embodiments of the present disclosure.

[0046] FIG. 20A illustrates a work flow for generating training data usable to train an Al model to generate modified images of dentition, in accordance with some embodiments of the present disclosure.

[0047] FIG. 20B illustrates another work flow for generating training data usable to train an Al model to generate modified images of dentition, in accordance with some embodiments of the present disclosure.

[0048] FIG. 21 illustrates a flow diagram of an example method for modifying an orthodontic treatment plan based on photo-based monitoring of a patient’s dentition, in accordance with some embodiments of the present disclosure.

[0049] FIG. 22 illustrates an example system architecture capable of supporting occlusion monitoring of a treatment plan, in accordance with one embodiment of the present disclosure.

[0050] FIG. 23 shows a block diagram of an example system for virtual dental care associated with an orthodontic treatment, in accordance with some embodiments.

[0051] FIG. 24 depicts a block diagram of an example processing device operating in accordance with one or more embodiments of the present disclosure. DETAILED DESCRIPTION

[0052] Embodiments described herein cover automated assessment of multiple different types of malocclusion. Examples of malocclusions that may be assessed include bite classifications such as class I malocclusion (e.g., including crowding, spacing, tooth rotations, etc.), class II malocclusion (e.g., excessive overbite, overjet), class III malocclusion (e.g., underbite), crossbite, and so on. Traditionally, malocclusions are diagnosed and monitored by doctors based on clinical examination of a patient’s teeth, jaws and/or facial structures, dental x-rays and/or cone-beam computed tomography (CBCT). However, in order for a doctor to identify a malocclusion and assess a severity of the malocclusion for a patient, the patient generally needs to visit the doctor’s office. This can impose a barrier to tracking a change in malocclusion over time.

[0053] In an example, if a patient undergoes orthodontic treatment to correct one or more malocclusions, the patient periodically visits their doctor to enable the doctor to assess a progress of the treatment against a treatment plan. However, the patient may need to take time off work for such visits, and such visits may occupy too much of the doctor’s time. As a result, patient visits may be few and far between because of the inconvenience associated with such patient visits.

[0054] Embodiments address the challenges associated with treatment monitoring by performing automated assessment of occlusion classification (s) of a patent based on images of the patient’s dentition. The patient may generate the images themselves using their own equipment (e.g., digital camera or mobile computing device such as a mobile phone, tablet computer, and so on). Alternatively, the images may be generated by the doctor or by a dental technician. The image(s) may be processed using processing logic configured for performing dentition assessment (e.g., occlusion assessment, such as crowding assessment, crossbite assessment, overbite assessment, underbite assessment, and so on) of images of dentition. In some embodiments, provided image(s) are processed using an artificial intelligence (Al) model (e.g., a generative Al model) that generates one or more new representations of the patient’s dentition based on processing of the provided image(s). The new representations may include new visual representations in one or more new images and/or may include non-visual representations (e.g., such as dimensions and/or coordinate locations of one or more features of the patient’s dentition). In some embodiments, oral diagnostics measurements may be performed using the one or more new representations of the patient’s dentition (e.g., new images). The processing logic may output identifications of one or more malocclusion classifications (e.g., crossbite, overbite, underbite, crowding, etc.), severity of the malocclusion classification(s), trends or changes in the severity of malocclusion classification(s), and so on. The output of the processing logic may be usable to determine a patient’s treatment progress for orthodontic treatment and/or palatal expansion treatment, to identify whether unanticipated malocclusions have developed, to recommend dental treatment (e.g., orthodontic treatment and/or palatal expansion treatment), and so on.

[0055] In the example of a class II malocclusion with a deep bite, the characterization of the deep bite may be done by automatically assessing the vertical overlap of the upper front teeth over the lower front teeth in embodiments. Deep bite may be measured based on assessment of 3D models and/or of bite-open and/or bite-closed images (e.g., where the jaws are in occlusion). The amount that the upper incisors overlap the lower incisors as projected into the vertical dimension can be measured and reported in embodiments. This may be measured in millimeters (or some other absolute distance measurement characterizing the amount of overlap) or as a percentage of the lower teeth covered by the upper teeth. The overlap can be characterized based on the lowest visible point on the upper incisors as compared to the highest visible point on the lower incisors (e.g., as viewed in an image taken from an angle generally normal to the patient’s dentition), or could be measured from a standard point such as the FACC-tip point (i.e., the point on the incisal edge of each tooth that is along the facialaxis).

[0056] Monitoring the progress of deep bite correction during orthodontic treatment presents some challenges. One of the challenges in this process is the need for precise intrusion of both the upper and lower incisors. If teeth do not move as planned, it may lead to complications. Additionally, there is a risk of the upper and/or lower incisors tilting forward (proclination). This can not only affect the bite but also the fit of orthodontic aligners. Poorly fitting aligners are less effective and can prolong treatment time. Regular check-ups with an orthodontist may typically be performed to assess the movement of the teeth and make adjustments to the treatment plan as necessary. Embodiments enable an amount of overbite to be automatically assessed based on one or more images of a patient’s dentition, which may eliminate a need for the patient to make regular doctor visits during orthodontic treatment treating an overbite. This may include automatically computing an amount of the lower teeth that are obscured by the upper teeth and/or automatically computing a percentage of the lower teeth that are obscured by the upper teeth based on assessment of one or more images.

[0057] In the example of a class III malocclusion (e.g., underbite or lower prognathism), the lower jaw extends farther forward than the upper jaw when the teeth are in occlusion. This results in the lower front teeth overlapping the upper front teeth. An underbite can affect the overall facial profile, giving the appearance of a protruding lower jaw and a flattened or recessed upper jaw. This may result in a prominent chin and/or a concave facial appearance, cause difficulty chewing and/or speaking, lead to premature wear of teeth, and/or cause temporomandibular joint (TMJ) disorders. Similar to deep bite, in embodiments, underbite may be automatically measured in 3D models and/or based on analysis of bite-open and/or bite-closed images of a patient. The amount that the lower incisors overlap the upper incisors when the jaw is closed as projected into the vertical dimension can be measured and reported from images of a patient’s dentition in embodiments. This may be measured in millimeters (or some other absolute distance measurement characterizing the amount of overlap) or as a percentage of the upper teeth covered by the lower teeth. The overlap can be defined based on the lowest visible point on the upper incisors as compared to the highest visible point on the lower incisors. Embodiments enable an amount of underbite to be automatically assessed based on one or more images of a patient’s dentition, which may eliminate a need for the patient to make regular doctor visits during orthodontic treatment treating an underbite. This may include automatically computing an amount of the upper teeth that are obscured by the lower teeth and/or automatically computing a percentage of the upper teeth that are obscured by the lower teeth based on assessment of one or more images.

[0058] In the example of a crossbite (e.g., a posterior crossbite, anterior crossbite and/or singletooth crossbite), one or more teeth of the upper dental arch are positioned behind the corresponding teeth on the lower dental arch when the jaws are closed (e.g., when the teeth are in occlusion). In other words, the upper teeth sit behind the lower teeth instead of outside or in front of the lower teeth (which is the normal bite relationship). The two main types of crossbite are anterior crossbite and posterior crossbite. Single-tooth crossbite is also possible. In an anterior crossbite, one or more of the upper front teeth are positioned behind the corresponding lower front teeth. In a posterior crossbite, one or more of the upper posterior (back) teeth are positioned inside or behind the corresponding lower posterior teeth. This can affect one side of the mouth (unilateral crossbite) or both sides of the mouth (bilateral crossbite), and may involve the premolars and/or molars. A posterior crossbite occurs, for example, when the occlusion of the buccal cusps of the upper molars is on the central fossae of their opposing lowers (as opposed to the buccal cusps of the lower molars occluding between the buccal and lingual cusps of the upper molars). Such posterior crossbite may be corrected via use of a palatal expander in embodiments (e.g., as set forth in U.S. Application No. 63/611 ,770, filed December 18, 2023, which is incorporated by reference herein in its entirety). Where the mandibular molars are more buccal than the maxillary molars, the patient is exhibiting posterior crossbite. When the maxillary molars are more buccal than the mandibular molars, the patient is not exhibiting posterior crossbite (e.g., the crossbite has been corrected). In a single-tooth crossbite, a single tooth in the upper dental arch sits behind the corresponding tooth in the lower dental arch.

[0059] Crossbites can lead to various oral health problems, such as uneven wear of teeth, increased risk of tooth decay gum disease, TMD dysfunction, jaw misalignment, facial asymmetry, difficulty chewing and/or speaking, and so on. Embodiments enable an amount of crossbite (e.g., posterior crossbite, anterior crossbite, single-tooth crossbite, etc.) to be automatically assessed based on one or more images of a patient’s dentition, which may eliminate a need for the patient to make regular doctor visits during orthodontic treatment of a crossbite. This may include automatically computing an amount of one or more upper teeth that are obscured by one or more opposing lower teeth and/or automatically computing a percentage of the upper teeth that are obscured by the lower teeth based on assessment of one or more images.

[0060] Tooth crowding occurs when there is not enough space in a patient’s jaw for all of their teeth to fit properly. As a result of the lack of space, teeth overlap, twist, and/or get pushed forward or backward out of alignment. Crowding can occur in the upper jaw and/or lower jaw. Crowding may be mild (with minor overlapping and/or tooth rotations), moderate (with a noticeable overlap and misalignment), or severe (where teeth are significantly twisted, overlapping, or even blocked from emerging). Crowding makes it difficult to maintain proper oral hygiene because it causes teeth to be more difficult to brush and floss, increasing the risk of plaque buildup, tooth decay, and gum disease. Misaligned teeth can also affect occlusion (how the teeth bite together), leading to uneven tooth wear, temporal mandibular joint (TMJ) issues, and/or difficulty chewing. Over time, crowding can worsen as teeth continue to shift. Crowding can be measured for individual pairs or sets of teeth based on tooth overlap in embodiments. Additionally, crowding can be measured based on an arch length analysis, in which the amount of space available in a dental arch and the total amount of space required by the teeth are compared. For example, space available can be a measure of a length of the arch perimeter, and the space required may be a sum of tooth sizes (e.g., mesiodistal tooth widths) plus any required interproximal spacing between teeth. The space available may be subtracted from the space required to determine whether there is crowding. A positive value may indicate crowding, while a negative number may indicate no crowding. Mild crowding may be corrected using orthodontic treatment (e.g., braces or clear aligners) to align teeth without tooth extraction. In some instances, interproximal reduction (I PR) may be performed for mild crowding. Moderate crowding may be treated with orthodontic treatment, optionally after performing I PR and/or palatal expansion treatment. Severe crowding may be treated using braces or aligners paired with extraction of one or more teeth, I PR, palatal expansion, and/or surgical jaw expansion (e.g., if skeletal issues are involved).

[0061] In some embodiments, tooth crowding may be measured for individual teeth and/or for a full upper and/or lower jaw based on assessment of 2D images in which a full view of all teeth may not be available. Such images may be processed to assess an amount of horizontal overlap between teeth, which may be used to determine crowding levels for individual pairs or sets of teeth. These measurements may be aggregated across all of a patient’s teeth to determine an overall crowding level or severity of the patient.

[0062] Aspects and implementations of the present disclosure address the above and other challenges by providing a method and system for consistent automatic image-based measuring and characterization of a patient’s occlusion (e.g., assessing a level of malocclusion). In embodiments, systems and methods for virtual monitoring of orthodontic treatment based on images of a patient’s dentition are provided. The images may be analyzed to determine a patient’s underbite, overbite, deep bite, crossbite, crowding, etc., including a determination of whether such malocclusion is decreasing as called for in a treatment plan, for example. Based on the image data, processing logic may automatically determine whether orthodontic treatment and/or palatal expansion treatment is called for, and/or whether orthodontic treatment and/or palatal expansion treatment is progressing as planned, and/or whether adjustments should be made to a treatment plan. Such adjustments may be automatically determined in embodiments, and may be presented to a doctor for approval prior to updating the treatment plan. Accordingly, in embodiments orthodontic treatment and/or palatal expansion treatment may be recommended and/or may be virtually monitored over extended periods without requiring a patient to visit their doctor (e.g., dentist or orthodontist) in person. Additionally, in embodiments processing logic may automatically determine a type, class, and/or severity of one or more types of malocclusion based on image assessment, and may provide such information to a doctor for their review. Accordingly, embodiments may increase an accuracy and consistency of malocclusion classification for patients across dental practices and/or within dental practices.

[0063] In some embodiments, one or more teeth of a patient are at least partially occluded (e.g., by other teeth of the patient). For example, a received image may include a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth. Such images may be processed in embodiments (e.g., using a trained Al model) to generate a second representation of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the first image. For example, the received image(s) may be processed to generate new images that comprise new contours of the at least one region of the one or more second teeth that is occluded in the first image. One or more oral diagnostics measurements of the patient’s dentition may be performed using the information of the at least one region of the one or more second teeth that is occluded in the first image (e.g., the new contours of the at least one region). The oral diagnostics measurement results may be used to determine a patient’s underbite, overbite, deep bite, crossbite, crowding, etc.

[0064] FIGS. 1A-1 D, 4-6, 8-15B illustrate flow diagrams of methods and processes for automatically characterizing an occlusion class of a patient. FIGS. 16A-17 illustrates a flow diagram for a method of recommending palatal expansion treatment. FIG. 19 illustrates a flow diagram of synthetic image generation and Al model training. The methods and processes described with association to these figures may be performed by a virtual dental care system, in accordance with embodiments of the present disclosure. The methods and processes described in association with these figures may additionally or alternatively be performed by a medical application (e.g., a chairside application) for a doctor. For example, the methods and processes may be performed by a machine running at a doctor’s office, or a server machine that interfaces with machine at a doctor’s office. These methods and processes may be performed, for example, by processing logic that comprises hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (such as instructions run on a processing device), or a combination thereof. In one embodiment, processing logic corresponds to a computing device of an occlusion monitor 2250, treatment plan coordination platform 2220 and/or storage platform 2240 of virtual care system 2270 of FIG. 22.

[0065] FIG. 1A illustrates an example process 100 for characterizing a dental occlusion, according to some embodiments of the present disclosure.

[0066] Process 100 of FIG. 1 A is a process that produces data for characterizing an occlusion (e.g., a level of malocclusion between opposing teeth of the upper and lower jaws) and/or bite for a dentition of a patient, based on image data of the patient.

[0067] At operation 1 .1 of process 100, image segmentation, processing logic can intake image data 102 (e.g., one or more 2D images of a patient’s dentition or teeth), and segment the image data 102 to identify oral structures, such as teeth, gingiva, etc. within the image. The image data may include one or more anterior image (e.g., a front facing or front view image of the patient’s dentition) with the patient’s jaws in a bite-open configuration (e.g., teeth not in occlusion) and/or with the patient’s jaws in a bite-closed configuration (e.g., teeth in occlusion). The image data may additionally or alternatively include one or more side-views or lateral images (e.g., left and/or right views) of the patient’s dentition with the patient’s jaws in a bite-open configuration and/or a bite-closed configuration. Different views may be usable to identify different types of malocclusion in embodiments. The image data may have been generated by a device of a patient (e.g., a camera device such as a phone/tablet, a scanning device) and/or of a dental practice (e.g., a camera device such as a phone/tablet, a clinical scanner) in embodiments.

[0068] In some embodiments, prior to image segmentation 1.1 , processing logic may assess the image data 102 to determine whether images of the image data 102 satisfy one or more image criteria. Image criteria may include a blurriness criterion, a criterion that at least a threshold amount of teeth are showing in the image, a sharpness criterion, a criterion that the image data includes one or more particular views of the patient’s dentition (e.g., an anterior bite-open view, an anterior bite-closed view, a lateral bite-open view, a lateral bite-closed view, etc.), and so on. Those images or sets of images that fail to satisfy image criteria may be filtered out. In some embodiments, images may be scored based on sharpness, amount of teeth visible, etc. In some embodiments, one or more highest scoring images (e.g., images that have a score exceeding a threshold value) of one or more views may be selected for analysis. For example, one or more highest scoring anterior bite-open views, one or more highest scoring anterior bite-closed views, one or more highest scoring lateral bite-open views, and/or one or more highest scoring lateral bite-closed views may be selected in embodiments. In some embodiments, if the image data 102 does not include images that meet image criteria for one or more views, processing logic may output a recommendation to obtain additional images of the one or more views. Processing logic may indicate what the deficiencies of the existing images are to enable an individual to generate images that satisfy the image criteria. In one embodiment, processing logic assesses images and filters out images in accordance with the teachings of U.S. Patent Application No. 17/111 ,264, filed December 3, 2020, which is incorporated by reference herein in its entirety.

[0069] In embodiments, image segmentation 1.1 is performed by one or more trained machine learning (ML) models, such as artificial neural networks (e.g., deep neural networks, convolutional neural networks (CNNs), etc.). During image segmentation 1.1 , image data 102 (e.g., 2D images of a patient’s teeth) can be processed by one or more trained ML models to output segmentation information. In one embodiment, the ML model generates one or more pixel-level segmentation masks of the patient’s teeth. The pixel-level segmentation mask(s) may separately identify each tooth, or may provide a single pixel-level identification of teeth, without separately calling out individual teeth. Accordingly, the ML model may perform semantic segmentation of the image data or instance segmentation of the image data in embodiments. In embodiments, the ML model may output a tooth identification (e.g., a tooth number) for each of the identified teeth. Operation 1 .1 can output the segmented image data (segmented data 104) to operation 1.2. In some embodiments, the segmentation is performed as described in U.S. Patent No. 11 ,759,291 or in U.S. Patent No. 10,997,727, which are incorporated by reference herein in its entirety.

[0070] At operation 1.2, digital measuring, processing logic can produce digital measurements 106 (e.g., digital oral diagnostics images) from the segmented data 104. Digital measurements 106 can include pixel distances of features within the segmented data 104. For instance, digital measurements 106 can include the pixel distance measurements of a tooth, or feature, visible within the segmented image data, or segmented data 104. The digital measurements that are generated may depend on the view of the image (e.g., anterior view, left side view, right side view, etc.) and/or the type of malocclusion to be assessed. In embodiments, multiple different measurements may be made of the segmented image data to assess multiple different classes of malocclusion. In an example, the digital measurements may include measurements of tooth heights of one or more exposed teeth and/or portions of teeth. In another example, the digital measurements may include measurements between two or more features or reference points identified on one or more teeth and/or between one or more teeth. For example, a digital measurement may include a shortest distance between a point on a maxillary canine (e.g., tip of a facial axis of a clinical crown (FACC)) and a boundary line between a mandibular canine and a mandibular first premolar. Multiple other digital measurements may also be generated. In an embodiment, one or more first digital measurements comprise a first digital measurement of a dimension of one or more first teeth of a first dental arch in a bite-open image and a second digital measurement of a dimension of one or more second teeth of a second dental arch in the bite-open image, and one or more second digital measurements comprise a third digital measurement of a dimension of the one or more first teeth in a bite-closed image and a fourth digital measurement of a dimension of the one or more second teeth in the bite-closed image, wherein a portion of the one or more second teeth is occluded in the bite-closed image.

[0071] In an example, to assess an overbite, processing logic may measure a height of a maxillary incisor in a bite-open image (e.g., an anterior bite-open image) and a bite-closed image (e.g., an anterior bite-closed image), and may measure a height of an opposing mandibular incisor in the biteopen image and the bite-closed image. Processing logic may then determine a first sum of the heights of the mandibular incisor and the maxillary incisor in the bite-open image and a second sum of the heights of the mandibular incisor and the maxillary incisor in the bite-closed image. Processing logic may determine a difference between the first sum and the second sum to determine an amount of overbite or deep bite.

[0072] In another example, to assess an underbite, processing logic may measure a height of a maxillary incisor in a bite-open image (e.g., an anterior bite-open image) and a bite-closed image (e.g., an anterior bite-closed image), and may measure a height of an opposing mandibular incisor in the biteopen image and the bite-closed image. Processing logic may then determine a first sum of the heights of the mandibular incisor and the maxillary incisor in the bite-open image and a second sum of the heights of the mandibular incisor and the maxillary incisor in the bite-closed image. Processing logic may determine a difference between the first sum and the second sum to determine an amount of underbite.

[0073] In another example, to assess a crossbite, processing logic may measure a height of a maxillary incisor in a bite-open image (e.g., a lateral bite-open image) and a bite-closed image (e.g., a lateral bite-closed image), and may measure a height of an opposing mandibular incisor in the biteopen image and the bite-closed image. Processing logic may then determine a first sum of the heights of the mandibular incisor and the maxillary incisor in the bite-open image and a second sum of the heights of the mandibular incisor and the maxillary incisor in the bite-closed image. Processing logic may determine a difference between the first sum and the second sum to determine an amount of crossbite. This assessment may be performed for a single tooth (e.g., to determine an amount or level of single-tooth crossbite), for a set of anterior teeth (e.g., to determine an amount or level of anterior crossbite), and/or for a set of posterior teeth (e.g., to determine an amount of level of posterior crossbite).

[0074] In another example, to assess a malocclusion class, processing logic may measure a distance between a tip of a FACC line on a maxillary canine and a point on a boundary between a mandibular canine and a mandibular first premolar in a bite-closed lateral image of the patient’s dentition. In some embodiments, the tip of the FACC line on the maxillary canine and the boundary between the mandibular canine and the mandibular first premolar may be projected onto a 3D model of the patient’s upper and/or lower dental arches. The boundary may be a plane when projected into the 3D model.

[0075] In another example, to assess a malocclusion class, processing logic registers a bite- closed image to a first 3D model of a patient’s upper dental arch and to a second 3D model of the patient’s lower dental arch. These 3D models may have been previously generated (e.g., based on an intraoral scan of the patient). Based on such registration, a relationship between the upper and lower dental arches may be determined. Processing logic may determine a first reference point on the first 3D model (e.g., a tip of an FACC line of a tooth) and a second reference point on the second 3D model (e.g., a boundary between two teeth), and may measure a distance between the two reference points. The distance may be measured in units of physical measurement (e.g., mm), and a bite classification may be determined based on the distance.

[0076] In another example, to assess a crossbite (e.g., a posterior crossbite, a single-tooth crossbite, and//or an anterior crossbite), processing logic may measure a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient from the segmented image data (e.g., from a single anterior or lateral bite-closed image). The type of crossbite assessed may be based on the tooth or teeth being measured. For example, anterior crossbite may be assessed based on measurements of one or more anterior teeth and posterior crossbite may be assessed based on measurement of one or more posterior teeth. If there is a crossbite, then the ratio of the maxillary tooth height to the corresponding mandibular tooth height may be small (e.g., less than 100%), indicating that some portion of the maxillary tooth is occluded by the mandibular tooth. If there is no crossbite, then in some instances the ratio of the maxillary tooth height to the corresponding mandibular tooth height may be large (e.g., greater than 100%), indicating that some portion of the mandibular tooth is occluded by the maxillary tooth. Processing logic may assess a crossbite based on a comparison of a determined ratio of a visible portion of a maxillary tooth height to a visible portion of a mandibular tooth height in a bite-closed image to a reference (e.g., a ratio of the known sizes of the maxillary and mandibular teeth). For example, processing logic may determine a first ratio between the first tooth height and the second tooth height from a single bite-closed image (e.g., anterior bite-closed image or lateral bite-closed image for an assessment of anterior crossbite or a lateral bite-closed image for an assessment of a lateral crossbite). The first ratio may be determined based on measurements of the maxillary and mandibular teeth in pixels, and may be referred to as an image ratio in embodiments. Processing logic may additionally determine a second ratio between a third tooth height of the maxillary tooth and a fourth tooth height of the mandibular tooth in one or more 3D models of the patient’s upper and lower dental arches. The ratio computed between the heights of the maxillary and mandibular teeth in the 3D models may be between facial axis of clinical crowns (FACC) heights of these teeth in some embodiments, which may be the heights of the teeth along their respective FACC lines. The 3D models may include the known dimensions of the maxillary and mandibular teeth. Thus, the actual ratio of the maxillary and mandibular teeth may be determined from the 3D models. Processing logic may determine a difference between the first ratio (e.g., image ratio as determined from a closed-bite image) and the second ratio (e.g., actual ratio as determined from the 3D models), and may determine an amount of posterior crossbite based on the difference. If the image ratio is equivalent to the actual ratio, then the image is showing the full heights of both teeth - either the bite is open, or both mandibular and maxillary teeth are fully visible in the image. If the image ratio is less than the actual ratio, then a portion of the maxillary teeth has been covered and the patient is exhibiting crossbite. When the image ratio is greater than the actual ratio, the patient does not exhibit crossbite for the assessed teeth (e.g., molars/premolars for an assessment of posterior crossbite or incisors/bicuspids for an assessment of anterior crossbite) because the maxillary teeth are covering a part of the mandibular teeth. Accordingly, if the first ratio is smaller than the second ratio, this indicates some level of crossbite. For example, if the first ratio is smaller than the second ratio for a single pair of a maxillary tooth and an opposing mandibular tooth, then a single-tooth crossbite may be determined. If the first ratio is smaller than the second ratio for a set of posterior teeth (e.g., molars and/or premolars), then a posterior crossbite may be determined. If the first ratio is smaller than the second ratio for a set of anterior teeth (e.g., incisors and/or bicuspids), than an anterior crossbite may be determined. If the first ratio is the same as the second ratio, this may indicate a small level of crossbite (e.g., since for a standard occlusion the maxillary tooth should partially occlude the mandibular tooth). If the first ratio is larger than the second ratio, this may indicate that there is no crossbite for a tooth or set of teeth.

[0077] One variation on the above approach is when both a bite-open and a bite-closed image (e.g., lateral images) are available. In this case, image ratios can be separately computed for the bite-open and bite-closed image. These ratios may be referred to as a bite-open ratio and a bite-closed ratio, respectively. The same comparison as discussed above between the image ratio and actual ratio can be performed using the bite-open and bite-closed ratios. In this case, the bite-open ratio may take the place of the actual ratio and the bite-closed ratio may take the place of the image ratio. [0078] In another example, to assess a crossbite (e.g., a posterior crossbite, an anterior crossbite, a single-tooth crossbite, etc.), processing logic may measure a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient from the segmented image data. For example, processing logic may measure the first and second tooth heights of the visible portions of one or more anterior teeth (e.g., incisors, bicuspids, etc.) and/or of one or more posterior teeth (e.g., molars, premolars, etc.) in a bite-closed image. Processing logic may then determine a first ratio (e.g., a bite-closed ratio) between the first tooth height and the second tooth height measured in the bite-closed image. Such measurements may be performed by measuring pixel heights of the visible portions of the maxillary and mandibular teeth in the bite-closed image in embodiments. Processing logic may additionally measure a third tooth height of the maxillary tooth and a fourth tooth height of the mandibular tooth of the patient in a bite-open image, and may determine a second ratio (e.g., bite-open ratio) between the third tooth height and the fourth tooth height. Such measurements may be performed by measuring pixel heights of the maxillary and mandibular teeth in the bite-open image in embodiments. Processing logic may determine a difference between the first ratio (bite-closed ratio) and the second ratio (bite-open ratio), and may determine an amount of crossbite based on the difference. If the bite-closed ratio is equivalent to the bite-open ratio, then the bite-closed image is showing the full heights of both teeth - either the bite is open, or both mandibular and maxillary molars are fully visible in the bite-closed image. If the bite-closed ratio is less than the bite-open ratio, then a portion of the maxillary teeth has been covered and the patient is exhibiting posterior crossbite. When the bite-closed ratio is greater than the bite-open ratio, the patient does not exhibit crossbite for the given teeth because the maxillary teeth are covering a part of the mandibular teeth. This assessment may be performed based on measurements of opposing anterior teeth to determine anterior crossbite and/or of opposing posterior teeth to determine posterior crossbite in embodiments.

[0079] At operation 1.3 (conversion) processing logic may convert the digital measurements 106 into physical measurements 108 (e.g., physical oral diagnostics measurements) in some embodiments. In embodiments, digital measurements 106 can be measurements of features of the image data that are in units of digital measurement, such as pixels. In embodiments, physical measurement 108 can be the same measurements as digital measurements 106, but converted to units for physical, or real- world, measurements. In embodiments, units of physical measurements may be millimeters, centimeters, inches, and so on. Some digital measurements may be unitless measurements, such as ratios or percentages, which may not undergo a unit conversion. For some measurements that are unitless (e.g., such as ratios), no conversion from digital measurements to physical measurements may be performed in some embodiments. [0080] Using the segmented image data and the treatment plan data, a pixel size can be computed for the image data. In embodiments, separate pixel sizes can be computed for each image within the image data. For instance, a separate pixel size can be determined for each image, each tooth, each jaw, and/or each pixel, etc. Pixel sizes can be determined by comparing the known physical size of a feature of the image (e.g., a tooth), to the measured pixel size from image data.

[0081] In embodiments, processing logic registers one or more images to one or more 3D models of the patient’s dental arches. Registration may be performed using shared features in the image data and the 3D model(s). In one embodiment, performing registration includes identifying various features on a surface in the 3D models (e.g., based on segmented information of the 3D models), identifying the same features in the 2D image data, and registering the 2D image data to the 3D model(s) by computing transformations. The registration may include adjusting a scale of one or more tooth in the 2D image data to match a scale of the one or more tooth in the 3D model(s). The registration may be assisted based on the tooth number labels of the 3D model and of the image data in embodiments. For example, registration may be performed by registering one or more teeth having particular tooth numbers from the 3D model to the teeth having the same tooth numbers from the image data.

[0082] The 3D models may have known accurate dimensions in units of physical measurement. Based on such registration of the image data to the 3D model(s), processing logic can determine a conversion factor FCON, for converting between millimeters and pixels (or vice versa). For instance, in embodiments, the conversion factor FCON, can indicate the real-world, or physical width of a pixel of an image (e.g., in millimeters).

[0083] Since each image of image data 102 can be taken from a different camera angle, or perspective, in some embodiments, processing logic can calculate FCON individually for each image, and/or for different regions of each image. Thus FCON can be generated to determine a size of a pixel in real-world measurements (e.g., in mm).

[0084] Once the conversion factor FCON is computed for an image, physical measurements 106 corresponding to the one or more digital measurements 106 can be determined using the pixel distance (e.g., number of pixels between two points as extracted from the image data), and multiplied by the conversion factor FCON. In embodiments, conversion factor FCON can then be used to generate distances and/or measurements from any feature within the image data, as desired. Alternatively, FCON can be individually computed for each tooth as desired. In embodiments, other features, including any known distances apparent in the image data, can be used to generate FCON. These can include, for example, size of teeth in one or more dimensions, etc. More information and additional methods and systems of determining pixel sizes may be found in U.S. Provisional Patent Application No. 63/511 ,635, which is incorporated by reference herein in its entirety. [0085] In some instances, processing logic may perform different techniques for converting from units of digital measurement to units of physical measurement based on one or more image criteria. In one embodiment, the one or more criteria comprise a resolution or pixel size criterion for one or more images in image data 102. For example, a first conversion technique (e.g., “standard conversion” 1.4) may be performed if the image data satisfies a first criterion, and a second conversion technique (e.g., “mapped conversion” 1 .5) may be performed if the image data fails to satisfy the first criterion (or if the image data satisfies a second criterion). These conversion techniques are explained in further detail below. In an example, the standard conversion 1.4 may be performed if the image data has a resolution that meets or exceeds a predetermined resolution threshold (e.g., 50 microns per pixel) and the mapped conversion 1 .5 may be performed if the image data has a resolution that is below the resolution threshold. In this example, the disclosed system may automatically determine that the standard conversion 1.4 be performed when an image is above a predetermined resolution threshold (e.g., a high-resolution image) and that the mapped conversion 1.5 be performed when an image is below a predetermined resolution threshold (e.g., a low-resolution image).The standard conversion 1.4 may include separately converting each digital measurement of an image to a physical measurement based on the conversion factor(s) computed for that image. An amount of one or more teeth of a dental arch that are covered by teeth of an opposing dental arch may then be determined based on one or more first physical measurements (e.g., from a bite-open image) and one or more second physical measurements (e.g., from a bite-closed image). The mapped conversion 1.5 may include converting a first digital measurement to a first physical measurement, determining other digital measurements as functions of the first digital measurement, and converting those other digital measurements to physical measurements based on the determined function and the determined first physical measurement. For example, the mapped conversion 1 .5 may include determining an amount of one or more teeth of a dental arch that are covered by teeth of an opposing dental arch based on one or more first digital measurements from a bite-open image and one or more second digital measurements from a bite- closed image, and converting the amount to a physical measurement.

[0086] The mapped conversion method may reduce a conversion error (as compared to the standard conversion method) associated with conversions between digital and physical measurements for low resolution images by reducing the number of individual conversions that are made. The error in the standard conversion is due to errors (variances) in measuring the tooth (e.g., incisor) pixel heights plus the errors in determining the sizes of pixels (e.g., of the maxillary tooth on the upper jaw in the biteopen image, of the mandibular tooth on the lower jaw of the bite-open image, and of the mandibular tooth on the lower jaw of the bite-closed image). The error of determining pixel sizes may correspond to the error in converting between pixels and units of physical measurement (e.g., error in a determined conversion factor). When the pixel sizes are small (i.e., on a high resolution image) there are more pixels in the incisors and the variances in the pixel sizes (and thus the error in pixel sizes) are smaller and the results are generally accurate. Accordingly, standard conversion 1.4 may be performed for high resolution (e.g., small pixel) images in embodiments. However, when the pixel sizes are large, the variances can be significant and affect the accuracy of the approach. The error in the determination of the sizes of pixels (e.g., of the conversion factors) are larger when there are fewer, larger pixels (as in the case of a low resolution image). Accordingly, one source of error associated with applying different conversion factors to different digital measurements may be eliminated for images with large pixels (e.g., low resolution images) by not requiring the determination of pixel size in the bite-closed image), reducing total error and increasing an accuracy of the measurements in physical units of measurement. Accordingly, in cases where the images have large pixels (e.g., low resolution images), the mapped conversion 1 .5 may be performed in embodiments to reduce the measurement error induced by the variances in the multiple pixel size measurements and conversions.

[0087] Operations 1 .4 and 1 .5 will be further described with respect to FIGS. 2 and 3.

[0088] In some embodiments, a correction operation 1 .6 can be applied to correct for any error introduced by one or more angles between a camera that generated the image data 102 and the imaged dentition of the patient.

[0089] Operation 1 .6 can apply a correction subprocess to the produced physical measurements 108, to produce corrected measurements 110. For example, to introduce further accuracy to measurements taken from image data 102, the angle at which an image sensor (e.g., a camera) is placed with respect to the teeth can be accounted for. For instance, in addition to, or alternatively to, conversion factor FCON, a correction factor FCOR can be computed to account for the changes in physical size attributable to an angle that the capturing image sensor (e.g., a camera) was held at. The correction factor may be a perspective correction factor determined based on an estimated inclination angle. In embodiments, this correction factor FCOR can be used for improving the conversion factor FCON, the pixel size estimate, and/or any of the digital measurements and/or physical measurements produced from the image data. In embodiments, the correction factor FCOR can be computed to correct for inaccuracies caused by a pitch or camera inclination of the camera (e.g., rotation about the y axis, which may be the left to right axis in the coordinate system of the patient’s dentition) and/or to correct for inaccuracies caused by a yaw of the camera (e.g., rotation about the z axis, which may be the vertical axis in the coordinate system of the patient’s dentition). Correction operation 1.6 is described in greater detail with reference to FIGS. 3-4.

[0090] FIG. 1 B is a flow chart of an example method 120 for characterizing a dental occlusion of a patient, according to some embodiments of the present disclosure. At block 122 of method 120, processing logic may access a treatment plan. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Each stage of treatment may be associated with a particular tooth arrangement in embodiments, and may include 3D models of the upper and/or lower dental arches of the patient having a particular tooth arrangement. In embodiments, method 120 may be performed during orthodontic treatment at a stage of the orthodontic treatment. Accordingly, a current stage of orthodontic treatment may be determined for the patient, and the 3D models associated with the current stage of treatment may be retrieved (e.g., loaded from a data store). Alternatively, method 120 may be performed prior to commencement of a treatment plan. For example, method 120 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 122 are omitted.

[0091] At block 124, processing logic may receive image data. The image data may include one or more images (e.g., two-dimensional (2D) images) of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed. The image data may have been generated by an imaging device of the person (e.g., of a patient), by an imaging device of a third party, and/or by an imaging device of a dental practice. In the example of image data captured by a patient during orthodontic treatment, the image data may have been captured by the patient (or a friend or family member of the patient) outside of a doctor office. For example, the patient may have generated one or more self-images during treatment. The images may have been uploaded to a server (e.g., of a virtual dental care system), which may execute method 120 in some embodiments.

[0092] In some embodiments, a patient may include a virtual dental care application on their mobile phone, mobile tablet, or other personal computing device. The virtual dental care application may include information on a treatment plan for the patient, including what types of malocclusion are being treated. Based on the types of malocclusion being treated, different image views may be optimal. Accordingly, in embodiments, the virtual dental care application may direct the patient to take one or more specific images based on information about the treatment plan for the patient. Alternatively, the virtual dental care application may direct the patient to generate one or more set images regardless of the treatment plan being executed. In an example, if an overbite or underbite is being corrected via the treatment plan, then the virtual dental care application may instruct the patient to generate a bite-open anterior view of their dentition and a bite-closed anterior view of their dentition. In another example, if a posterior crossbite is being corrected via the treatment plan, then the virtual dental care application may instruct the patient to generate a bite-closed lateral view of their dentition and/or a bite-open lateral view of their dentition. [0093] At block 126, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth (e.g., segmenting each of the visible teeth into individual segments), gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth.

[0094] At block 128, processing logic may identify pixel sizes in the image(s) and/or resolution of the images. A first unit conversion technique may be more accurate for images having small pixels (e.g., high resolution images), and a second unit conversion technique may be more accurate for images having large pixels (e.g., low resolution images). For example, as explained previously, the first unit conversion technique may be the standard technique and the second unit conversion technique may be the mapped technique (both techniques are described in further detail below). Accordingly, at block 130, processing logic may determine whether the pixel sizes of the image(s) are smaller than a pixel size threshold and/or whether the resolution of the image(s) is higher than a resolution threshold. If the pixel size is lower than the pixel size threshold and/or the resolution is higher than the resolution threshold, then the image(s) may be classified as high-resolution images, and the method may continue to block 132. If the pixel size is higher than the pixel size threshold and/or the resolution is lower than the resolution threshold, then the image(s) may be classified as low-resolution images, and the method may continue to block 134.

[0095] At block 132, processing logic may perform a first measurement technique for measuring an amount of dental occlusion between opposing teeth in the upper and lower jaws. The amount of dental occlusion may be an amount of overbite, an amount of underbite, an amount of posterior crossbite, and so on. The first measurement technique may include measuring the heights of one or more maxillary teeth and one or more mandibular teeth in a bite-open image and in a bite-closed image in units of digital measurement, converting each of the measurements into units of physical measurement, and then calculating an amount of dental occlusion based on the measurements in the units of physical measurement. Once an amount of dental occlusion is determined, at block 136 a correction may be applied to correct for camera inclination in some embodiments, as discussed with reference to FIGS. 3-4. The correction may additionally, or alternatively, be applied to account for (e.g., correct for) the angle of teeth surfaces in some embodiments. For example, a camera may be positioned perfectly normal to the jawline, but teeth may protrude outward. In such instances, the correction may be performed to address angle differences of the camera with respect to the teeth surfaces that the camera is photographing.

[0096] At block 134, processing logic may perform a second measurement technique for measuring an amount of dental occlusion between opposing teeth in the upper and lower jaws. The amount of dental occlusion may be an amount of overbite, an amount of underbite, an amount of posterior crossbite, and so on. The second measurement technique may include measuring the heights of one or more maxillary teeth and one or more mandibular teeth in a bite-open image and in a bite- closed image in units of digital measurement, determining functions that represent one or more measurement in terms of a fraction or amount of another measurement, converting a single measurement into units of physical measurement, and then applying the functions to the single physical measurement to determine the other measurements in units of physical measurement. One or more of the measurements in units of physical measurement may then be used to determine an amount of dental occlusion.

[0097] In some embodiments, a third measurement technique may be performed. The third measurement technique may include measuring heights of one or more mandibular teeth and/or maxillary teeth in units of digital measurement, and then determining ratios and/or percentages between the measurements, resulting in a unitless value that indicates a level of malocclusion. This may be performed instead of, or in addition to, the first or second measurement technique.

[0098] At block 138, processing logic characterizes a level of malocclusion between the opposing upper and lower jaws based on the measurement(s) as measured in units of physical measurement. Different dental occlusion values may correlate to different levels or stages of one or more types of malocclusion. For example, an overbite may be considered to be normal when the vertical overlap covers 30% of the lower teeth (or is, e.g., 2-4 mm). When the overbite is 4-6 mm or more, the overbite may be considered a deep overbite or deep bite. Overbites of 9 mm or more may be classified as severe overbite. Accordingly, the level or severity of dental occlusion may be determined based on the measurements in physical units of measurement (e.g., mm) and/or in the unitless values (e.g., percentages or ratios).

[0099] In some embodiments, processing logic may compare the determined level of malocclusion and/or the measurement(s) of dental occlusion to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of dental occlusion may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments. [0100] At block 140, processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on. Alternatively or additionally, the method 120 may include transmitting to an output device of a computing system associated with the patient and/or the patient’s doctor, or otherwise causing a display on an output device, of information related to the characterization of the level of malocclusion. For example, a visualization, a quantitative (e.g., numerical) characterization, and/or a qualitative characterization (e.g., stating whether or not the patient is on track with respect to a treatment plan) may be displayed on a screen of a computing system (e.g., a tablet, a computer, a phone).

[0101] FIG. 1C is a flow chart for a first example method 150 for measuring a level of malocclusion between opposing teeth of the upper and lower jaws of a patient, according to some embodiments of the present disclosure. The first example method 150 may correspond to the standard conversion 1.4 technique previously discussed in embodiments. The first example method 150 for measuring a level of malocclusion includes, at block 152, measuring in a first segmented image (e.g., a bite-open image), a first tooth height of a tooth in the upper dental arch (e.g., a maxillary tooth) and a first tooth height of an opposing tooth in the lower dental arch (e.g., a mandibular tooth). The first tooth height of the first and second tooth may be measured in digital measurement units (e.g., pixels) in embodiments. At block 154, a second tooth height of the tooth in the upper dental arch and a second tooth height of the opposing tooth in the lower dental arch in the segmented bite-closed image 200B are measured. The second tooth height of the first and second tooth may also be measured in digital measurement units (e.g., pixels) in embodiments.

[0102] FIG. 2A illustrates an example bite-open anterior image 200A, according to some embodiments of the present disclosure. In the bite-open anterior image 200A, a first view of a right maxillary canine (e.g., tooth 6) 206A, a right maxillary lateral incisor (e.g., tooth 7) 207A, a right maxillary central incisor (e.g., tooth 8) 208A, a left maxillary central incisor (e.g., tooth 9) 209A, a left maxillary lateral incisor (e.g., tooth 10) 210A, and a left maxillary canine (e.g., tooth 11) 211A are shown. Additionally, a first view of a right mandibular canine (e.g., tooth 27) 227A, a right mandibular lateral incisor (e.g., tooth 26) 226A, a right mandibular central incisor (e.g., tooth 25) 225A, a left mandibular central incisor (e.g., tooth 24) 224A, a left mandibular lateral incisor (e.g., tooth 23) 223A, and a left mandibular canine (e.g., tooth 22) 222A are shown. FIG. 2B illustrates an example bite- closed anterior image, according to some embodiments of the present disclosure. In the bite-closed anterior image 200B, a second view of the right maxillary canine (e.g., tooth 6) 206B, the right maxillary lateral incisor (e.g., tooth 7) 207B, the right maxillary central incisor (e.g., tooth 8) 208B, the left maxillary central incisor (e.g., tooth 9) 209B, the left maxillary lateral incisor (e.g., tooth 10) 210B, and the left maxillary canine (e.g., tooth 11) 211 B are shown. Additionally, a second view of the right mandibular canine (e.g., tooth 27) 227B, the right mandibular lateral incisor (e.g., tooth 26) 226B, the right mandibular central incisor (e.g., tooth 25) 225B, the left mandibular central incisor (e.g., tooth 24) 224B, the left mandibular lateral incisor (e.g., tooth 23) 223B, and the left mandibular canine (e.g., tooth 22) 222B are shown. As shown, the full height of the mandibular teeth are shown in the bite-open anterior image 200A. However, portions of the mandibular teeth are occluded by the opposing maxillary teeth in the bite-closed image 200B. The first measurement 250A of one or more maxillary teeth (e.g., of tooth 208A and tooth 209A) and the first measurement 252A of one or more mandibular teeth (e.g., of tooth 225A and 224A) in the first image 200A are shown. Additionally, the second measurement 250B of the one or more maxillary teeth (e.g., of tooth 208A and tooth 209A) and the second measurement 252B of one or more mandibular teeth (e.g., of tooth 225A and 224A) in the second image 200B are shown.

[0103] Returning to FIG. 1 C, at block 156, the first image and the second image may be registered to a 3D model of the upper dental arch and to a 3D model of the lower dental arch of the patient. The 3D models may correspond to a current treatment stage, and may be part of a treatment plan. Alternatively, the 3D models may have been generated based on a prior intraoral scanning of the patient’s oral cavity. In some embodiments, the 3D models are generated from a collection of 2D images of the patient’s oral cavity. The 3D models may include accurate information on sizes of oral structures (e.g., teeth, arch width, etc.) on the upper and lower dental arches of the patient. Any suitable method may be used to register a 2D image to a 3D model. For example, the disclosed systems may use an expectation maximization technique, a differentiable rendering technique, and/or a joint jaw pair optimization technique. More information about such registration techniques are described in U.S. Patent No. 11 ,020,205; U.S. Provisional Application No. 63/585,581 ; U.S. Patent No. 11 ,723,748; and U.S. Provisional Application No. 63/511 ,635.

[0104] In some instances, operation 156 may be omitted for the first image and/or the second image. Operation 156 may be omitted, for example, if a height H’u, H’L (shown in FIG. 2A) and/or Hu (shown in FIG. 2B) were previously determined based on images captured during a prior stage of treatment. For example, referencing FIG. 2B, once processing logic has determined the height of Hu when performing a first monitoring event (e.g., at a first stage of treatment), processing logic can use this as a baseline to determine pixel heights at a subsequent monitoring event (e.g., at a third stage of treatment). This may enable processing logic to determine physical units of measurement for Hu, H’u, HL and/or H’L without having to register current images against a 3D model. This may allow for a faster processing, save computing resources, etc. [0105] In some instances, accuracy may go down over time when prior registrations to 3D models are leveraged as the teeth move (thus affecting the angle of photo capture). Accordingly, a reregistration step may be performed periodically (e.g., after every 5 stages, 8 stages, 10 stages, etc., depending on the treatment plan and how much the teeth are expected to move). Thus, in some embodiments, a "batched" approach may be used in which registration (e.g., the operations of block 156) are performed periodically (e.g., every 5 stages, 8 stages, 10 stages, etc.).

[0106] Based on the registration of the images to the 3D models, at block 158 one or more digital to physical measurement unit conversion factors may be determined. For example, the 3D model may have information about the real-world heights (and/or other dimensions) of each tooth and/or other feature, and this information may be used to convert pixel dimensions of the 2D image to real-world dimensions once the images have been registered to a 3D model (e.g., as further describe previously in relation to conversion factor FCON). In an embodiment, a different conversion factor is determined for the first image and the second image.

[0107] In some instances, operations 152, 156 and 158 may be omitted for the first image (e.g., the bite-open image). Operations 152, 156 and/or 158 may be omitted, for example, where real-world heights H’u and/or H’Li were previously determined based on images captured during a prior stage of treatment. For example, referencing FIG. 2A, once processing logic has determined the heights H’u and/or H’L when performing a first monitoring event (e.g., at a first stage of treatment), processing logic can use this as the bite-open tooth heights at a subsequent monitoring event (e.g., at a third stage of treatment). This may enable processing logic to determine physical units of measurement for H’u and/or H’L without having generate a new bite-open image, without having to segment a new bite-open image, without having to measure digital heights (e.g., in pixels) in a new bite-open image, without having register such a new bite-open image against a 3D model, and/or without having to convert digital units of measurement to physical units of measurement for a new bite-open image. This may allow for a faster processing, save computing resources, etc.

[0108] In some instances, accuracy may go down over time when old bite-open images (and analysis of such) are leveraged as the teeth move (thus affecting the angle of photo capture). Accordingly, operations 152, 156, 158 and/or 160 may be performed for bite-open images periodically (e.g., after every 5 stages, 8 stages, 10 stages, etc., depending on the treatment plan and how much the teeth are expected to move). Thus, in some embodiments, a "batched" approach may be used in which new bite-open images are processed periodically (e.g., every 5 stages, 8 stages, 10 stages, etc.).

[0109] At block 160, processing logic converts the first and second measurements of the first tooth and the second tooth from units of digital measurement to units of physical measurement using the determined conversion factor(s). In embodiments, the conversion factors indicate a number of mm per pixel, which may range from a fraction of a mm per pixel to one or more mm per pixel. The unit conversion may be performed by multiplying the digital measurements by the appropriate conversion factor(s).

[0110] At block 162, processing logic may determine a difference between the second tooth height (e.g., of the mandibular tooth) between the first image (e.g., bite-open image) and the second image (e.g., bite-closed image). The difference in the tooth height between the two images is the amount of the occluded tooth that is behind the opposing tooth on the opposite dental arch. In one embodiment, a first combined height of the first and second tooth in the bite-open image (e.g., height 250A plus height 252A) and a second combined height of the visible portions of the first and second tooth in the closed bit image (e.g., height 250B plus height 252B) are determined. A difference between the first combined height and the second combined height may be computed to determine an amount of overbite (e.g., when the top teeth are occluding the bottom teeth in an anterior image), to determine an amount of underbite (e.g., when the bottom teeth are occluding the top teeth in an anterior image), and/or to determine an amount of crossbite (e.g., when the bottom teeth are occluding the top teeth in a lateral image).

[0111] In an example, a level of malocclusion (e.g., an amount of overbite or underbite) can then be characterized by the equation Occlusion Value = H'_v + H'_L - (Hu + H_L ), where H'_v is the height of the upper incisors of the bite-open image or height 250A, H'_L is the height of the lower incisors of the bite-open image or height 252A, Hu is the height of the upper incisors as visible in the bite-closed image or height 250B, and H_L is the height of the lower incisors as visible in the bite- closed image or height 252B. Note that (Hu + H_L ) can correspond to the combined height of the visible teeth. Note that the above equation can be applied after all terms have been converted to units of physical measurement (e.g., millimeters) via respective conversion factors. The disclosed systems and methods may classify a patient’s bite based on the Occlusion Value. For example, as described previously, a certain range of Occlusion Values may correspond to an overbite and/or underbite, and a certain range of Occlusion Values may correspond to a proper or “normal” occlusion. Although FIGS. 2A-2B illustrate an overbite, a similar method can be used to classify and/or monitor underbite, where one or more lower teeth obscure a portion of one or more upper teeth. In the case of an overbite, (H'u + H'_L )is greater than + H_L ) because of a decreased HL value. In the case of an underbite, , H'u + H'_L ) is greater than + H_L ) because of a decreased Hu value.

[0112] In some embodiments, the standard conversion 1 .4 method may also be used to characterize an open bite malocclusion, where the patient’s upper and lower teeth do not overlap in a bite-closed arrangement (e.g., the posterior teeth may be touching, but one or more anterior teeth may not be overlapping). In an open bite case, the expression above ((H'u + H'_L - + H_L )) would yield a value of zero. Using the standard conversion 1.4 method as discussed above, an Occlusion Value of zero may be used in classifying a patient as having an open bite, and the patient may be monitored until the patient has an Occlusion Value that is within a normal range. In some embodiments, in an open bite case, the bite-closed image may be registered to a 3D model and the real-world distance between the upper and lower teeth while in a bite-closed arrangement may be tracked during treatment to monitor progress.

[0113] Mapped conversion 1 .5 can similarly characterize an amount of overbite or underbite. However, mapped conversion can be employed when resolution is low or below a threshold. In embodiments, mapped conversion can be used when resolution is below 50 microns per pixel. [0114] FIG. 1 D is a flow chart for a second example method 170 for measuring a level of malocclusion between opposing teeth of the upper and lower jaws of a patient, according to some embodiments of the present disclosure. In embodiments, method 170 corresponds to mapped measurement 1 .5. The second example method 170 for measuring a level of malocclusion includes, at block 172, measuring in a first segmented image (e.g., a bite-open image), a first tooth height of a first tooth in the upper dental arch (e.g., a maxillary tooth) or lower dental arch (e.g., mandibular tooth) and a first tooth height of an opposing second tooth in the lower dental arch or upper dental arch. The first tooth height of the first and second tooth may be measured in digital measurement units (e.g., pixels) in embodiments. At block 174, a second tooth height of the first tooth and a second tooth height of the opposing second tooth in the segmented bite-closed image 200B are measured. If an overbite is being assessed, the first tooth may be a maxillary tooth on the upper dental arch and the second tooth may be a mandibular tooth on the lower dental arch. If an underbite is being assessed, the first tooth may be a mandibular tooth on the lower dental arch and the second tooth may be a maxillary tooth on the upper dental arch. The second tooth height of the first and second tooth may also be measured in digital measurement units (e.g., pixels) in embodiments.

[0115] At block 176, processing logic may determine a ratio between the second tooth height of the first tooth and the second tooth height of the opposing second tooth. The ratio of the pixel heights of the visible lower incisors to the upper incisors of the bite-closed image 200B can be computed, e.g., by dividing the number of pixels of height 252B by the number of pixels of height 250B. At block 178, processing logic may map the ratio to the second tooth in the bite-open image. This allows the height 252B of the second tooth in the bite-closed image to be represented as a function of the height 250A in the bite-open image times the ratio of the height 252B to the height 250B in the closed byte image. Additionally, the amount of the second tooth that is occluded in the bite-closed image may be represented as a difference between the height of the second tooth in the bite-open image and the height of the second tooth in the bite-closed image.

[0116] In an example, let height 250A be represented as H’u, let height 252A be represented as H’L, let height 250B be represented as Hu, and let height 252B be represented as HL. From the bite- closed image 200B, processing logic can compute the ratio of the pixel heights of the visible lower incisors (W_L) to the upper incisors (Hu), which yields the ratio:

^HFU (¹)

[0117] This ratio can be mapped to the bite-open image according to the formula:

H_L2 _ L m

Uy ~ Hu ' which allows processing logic to solve for height 256A, represented as H’L2 according to the equation:

[0118] The amount of disocclusion (e.g., from an overbite) in pixels in the bite-open image is then characterized by an Occlusion Value as follows:

Occlusion Value = H_L'₁ - Hy — (4)

Hu

[0119] Accordingly, the amount of disocclusion (e.g., overbite) can be represented entirely in units of pixels in the bite-open image. This value may represent the height of the lower incisors that would be covered by the upper incisors.

[0120] Note that the above equations 1-4 provide calculations for determining an amount of overbite in embodiments. The same equations as set forth above may be similarly used to compute an amount of underbite by swapping each variable for a tooth on the lower dental arch (e.g., HL, H’L2, H’LI) with a variable for a tooth on the upper dental arch, and swapping each variable for a tooth on the upper dental arch (e.g., Hu, H’u) with a variable for a tooth on the lower dental arch.

[0121] In an example, let H’ui represent a bite-open maxillary tooth height, let H’u2 represent a visible portion of the closed-bite maxillary tooth height mapped to the bite-open image, let H’L represent a bite-open mandibular tooth height, let height Hu represent a visible closed-bite maxillary tooth height, and let HL represent a closed-bite mandibular tooth height. From the bite-closed image, processing logic can compute the ratio of the pixel heights of the visible upper incisors (Hu) to the lower incisors (H_L , which yields the ratio: [0122] This ratio can be mapped to the bite-open image according to the formula: which allows processing logic to solve for H’u2 according to the equation:

[0123] The amount of disocclusion (e.g., from an underbite) in pixels in the bite-open image is then characterized by an Occlusion Value as follows:

Occlusion Value

[0124] Accordingly, the amount of disocclusion (e.g., underbite) can be represented entirely in units of pixels in the bite-open image. This value may represent the height of the upper incisors that would be covered by the lower incisors.

[0125] At block 180, processing logic registers the first image to 3D models of the upper and lower dental arches. The 3D models may be from a treatment plan and may be associated with a current stage of treatment. Alternatively, the 3D models may not be from a treatment plan. For example, the 3D models may be based on intraoral scanning of the patient’s oral cavity, and may represent a current or prior condition of the patient’s dentition.

[0126] In some instances, operation 180 may be omitted. Operation 180 may be omitted, for example, if a height H’u, H’L (shown in FIG. 2A) and/or Hu (shown in FIG. 2B) were previously determined based on images captured during a prior stage of treatment. For example, referencing FIG. 2B, once processing logic has determined the height of Hu when performing a first monitoring event (e.g., at a first stage of treatment), processing logic can use this as a baseline to determine pixel heights at a subsequent monitoring event (e.g., at a third stage of treatment). This may enable processing logic to determine physical units of measurement for Hu, H’u, HL and/or H’L without having to register current images against a 3D model. This may allow for a faster processing, save computing resources, etc.

[0127] In some instances, accuracy may go down over time when prior registrations to 3D models are leveraged as the teeth move (thus affecting the angle of photo capture). Accordingly, a reregistration step may be performed periodically (e.g., after every 5 stages, 8 stages, 10 stages, etc., depending on the treatment plan and how much the teeth are expected to move). Thus, in some embodiments, a "batched" approach may be used in which registration (e.g., the operations of block 156) are performed periodically (e.g., every 5 stages, 8 stages, 10 stages, etc.). [0128] At block 182, processing logic determines a digital to physical measurement unit conversion factor based on the registration (or based on prior registration of a prior image taken at an earlier treatment stage to 3D models of dental arches). At block 184, processing logic converts the first measurements from digital measurements to units of physical measurement using the conversion factor. At block 186, processing logic may then multiply the determined ratio by the first measurements in physical units to determine a second measurement for the second tooth in physical measurement units. At block 188, processing logic may then determine a difference between the first tooth height in the bite-open image and the bite-closed image to determine an amount of dental occlusion (e.g., overbite). Similar computations may be made to solve for an amount of underbite and/or an amount of crossbite in embodiments.

[0129] In an example of determining an amount of overbite or deep bite, H’Li and H’u may be converted to units of physical measurement, and these values in units of physical measurement may be plugged into equation (4) above to solve for the amount of overbite in units of physical measurement. In an example, the amount of overbite in pixels can then be converted to mm according to the following equation:

Occlusion Value = (H_L'₁ — H_L'₂) * lower pixel size (in bite-open image) (9)

[0130] In an example of determining an amount of underbite, H’ui and H’L may be converted to units of physical measurement, and these values in units of physical measurement may be plugged into equation (8) above to solve for the amount of underbite in units of physical measurement. In an example, the amount of underbite in pixels can then be converted to mm according to the following equation:

Occlusion Value = (Hy_X — H_U' ₂) * upper pixel size (in bite-open image) (10)

[0131] Since the mapped conversion method only involves use of a single pixel size (upper jaw on the bite-open image for an overbite determination or lower jaw on the bite-open image for an underbite determination), it can have significantly less error for lower resolution images (e.g., where the pixel sizes are larger) as compared to the standard conversion method.

[0132] In some embodiments, the mapped conversion 1 .5 method may also be used to characterize an open bite malocclusion. For example, the mapped conversion 1.5 method may be used to characterize an open bite malocclusion by determining whether or not an overlap exists between maxillary and mandibular teeth.

[0133] Similar methods may be employed for crossbite (e.g., anterior crossbite, posterior crossbite, single-tooth crossbite), determining overlaps between relevant upper and lower teeth to classify a malocclusion in a patient and/or to monitor the level of malocclusion as patient is treated. Lateral views may be additionally or alternatively used for determining such crossbite or for monitoring such crossbite (e.g., during treatment). For example, the standard method 1 .4 and/or mapped method 1 .5 may be performed using bite-closed and bite-open lateral views of a patient’s dentition to determine an amount of posterior crossbite, anterior crossbite and/or single-tooth crossbite.

[0134] FIG. 3 illustrates an example correction subprocess 300, according to some embodiments of the present disclosure. In embodiments, the correction subprocess 300 may be performed for the mapped conversion 1.5. The correction subprocess may correspond to block 136 of method 120 in embodiments. FIG. 4 illustrates an example orientation of a camera 402 relative to a 3D model 404 of a dental arch of a patient, according to some embodiments of the present disclosure.

[0135] With reference to FIGS. 3-4, at operation 3.1 , registering, of correction subprocess 300, processing logic can intake image data 302. Processing logic can register a known, 3D model 404 of the patient’s teeth (e.g., of an upper and/or lower dental arch of the patient) to captured image data 302 that was used to determine an amount of dental occlusion. The 3D model(s) 404 may be registered, for example, to a bite-open image and/or to a bite-closed image of the patient’s dentition. Such a registration process can be used to determine an inclination angle, a, of the image source 402, or sensor, used to generate the image data.

[0136] Once an inclination angle, a, has been determined, the overjet, d, of the patient’s teeth can be found using the 3D model(s) 404 of the patient’s dental arches in occlusion. At operation 3.2, correction factor determination, processing logic may use the inclination angle a and overjet d to compute a correctional factor (Of) using trigonometry according to the following equation:

C_F = d tan a (11)

[0137] This correction factor is the additional overbite that is apparent in the image due to the camera inclination. By subtracting the correction factor from the overbite found using the mapped conversion (e.g., the measurement techniques of method 170), a final corrected overbite measurement for the patient can be identified. Similar corrections for camera inclination can be determined for underbite and/or crossbite measurements in embodiments.

[0138] In an example, the camera that generated an image may have been above the plane of the dental arch at the time of imaging. The inclination angle, a, can be found by registering the patient’s 3D dentition (e.g., as may have been performed to determine the conversion factor) to the 2D dentition from the image data. Based on the patient’s treatment plan, a planned relationship between the teeth of the upper and lower dental arches (e.g., including the relative positions of teeth on the upper dental arch and teeth on the lower dental arch) for a stage of treatment may be determined. This may include, for example, determining an amount of overjet or underjet of the patient’s teeth at the treatment stage. The amount of overjet or underjet can be determined as a distance, d, in the x-direction (as indicated in FIG. 4) between the upper and lower jaws. With these values, processing logic can compute a correction factor: c. f. = d tan (a). This correction factor is the additional overbite, underbite, crossbite, etc. that is apparent in the image due to the camera inclination. By subtracting the correction factor from the physical measurements for the overbite, underbite, crossbite, etc. (e.g., optionally found using either the standard or the mapped measurement), processing logic can identify the final overbite, underbite, crossbite, etc. measurement for the patient.

[0139] The correction described with reference to FIGS. 3-4 may additionally, or alternatively, be applied to account for (e.g., correct for) the angle of teeth surfaces in some embodiments. For example, a camera may be positioned perfectly normal to the jawline, but teeth may protrude outward. In such instances, the correction may be performed to address angle differences of the camera with respect to the teeth surfaces that the camera is photographing.

[0140] FIG. 5 illustrates an example process 500 for characterizing a dental occlusion, according to some embodiments of the present disclosure. Process 500 of FIG. 5 illustrates a process that produces data for characterizing an occlusion of a patient, based on image data of the patient. Process 500 may be performed, for example, from a lateral bite-closed image of a patient to determine a malocclusion class and/or a level of malocclusion of the patient. For example, process 500 may be performed to identify a class I malocclusion, a class II malocclusion and/or a class III malocclusion. [0141] At operation 5.1 of process 500, processing logic performs image filtering on received image data 502. The image filtering may be performed to determine whether input image data satisfies one or more image criteria, and to filter out or remove those images that fail to satisfy the one or more image criteria. Image criteria may include a pose criterion (e.g., bite-open arrangement vs. bite-closed arrangement), a view criterion (e.g., anterior view, lateral view, etc.), a blurriness or sharpness criterion, an amount of shown teeth criterion, and so on. An output of operation 5.1 may be a reduced set of filtered image data 503.

[0142] At operation 5.2, image segmentation may be performed on the filtered image data. The segmentation may be performed to generate segmented image data in which the location, shape, size, etc. of individual teeth in one or more images are identified. For example, the segmentation that is performed may be instance segmentation, and may output a mask for each tooth in an image. The mask for a tooth may indicate the pixels of the image that represent the tooth. In embodiments, in addition to segmenting the image into individual teeth, processing logic may determine identification of the teeth, and may assign tooth numbers to the teeth based on an accepted tooth numbering scheme (e.g., such as the Universal Tooth Numbering System, the Zsigmondy tooth numbering system, and so on). In addition to the image(s) being segmented, one or more 3D model(s) of an upper and/or lower dental arch may also be segmented into individual teeth, and tooth numbers may be assigned to the identified teeth in the 3D model(s).

[0143] Operation 5.3, reference point localization, of process 500 can include localizing reference points from segmented image data 504. In embodiments, one or more known, 3D models of the patient’s dental arch(es) (e.g., segmented 3D models) can be registered onto segmented image data 502. The registration may be performed as discussed previously, and tooth identifications may be used to assist the registration process in embodiments. Once registration has been performed, one or more reference points may be identified on the 3D model and may be mapped to the 2D image based on the registration between the 3D model and the 2D image. Accordingly, reference points can be localized on to the 3D model, as opposed to the image data. In embodiments, 3D models of the whole upper and lower jaws can be registered to the 2D image, to determine the relative position of the teeth in upper dental arch to the teeth in the lower dental arch in 3D.

[0144] At operation 5.4, a first reference point can be localized within segmented image data 502. In embodiments, the first reference point on a maxillary tooth is determined in the 3D model of the patient’s upper dental arch, and is projected onto the 2D image. The first reference point can be the FACC line of the maxillary canine, for example. In embodiments, the first reference point can be the tip of the FACC line of the maxillary canine.

[0145] At operation 5.5, a second reference point can be localized within the image data. The second reference point can be a point on the boundary line between two teeth on the lower dental arch in an embodiment. In one example, the second reference point is a point on the boundary line between the mandibular canine and premolar. In embodiments, the second reference point may be the point on the boundary line that is closest to the first reference point. Note that though two reference points are discussed, more than two reference points may be determined on the 2D image.

[0146] Once reference points 504 have been identified on the 2D image, one or more digital measurements between the reference points can be taken. The digital measurements may be measurements in 2D, and may not reflect an actual distance between corresponding points in 3D (e.g., on the 3D models). In embodiments, a virtual marker between the two refence points may be determined, and the length of the virtual marker may be determined (where the length of the virtual marker represents the distance between the two reference points in the 2D image). Accordingly, in some embodiments processing logic projects the virtual marker and/or the digital measurement from an image space of the 2D image to a 3D model space of the one or more 3D models. In embodiments, the projection is performed by determining an angle between the patient’s jaw(s) (e.g., the 3D model(s) of the dental arch(es)) and the image plane of the 2D image. The angle may then be used to adjust the distance between the two points using a trigonometric function in embodiments.

[0147] At operation 5.7, unit conversion, processing logic may convert the units of digital measurement into units of physical measurement. In embodiments, such unit conversion is performed based on the registration between the 2D image and the 3D model(s). The 3D model(s) may be an accurate representation of the patient’s teeth with physical units of measurement. Accordingly, once the 3D model(s) are registered to the 2D image, the conversion factor between a pixel and a unit of length (e.g., mm) may be determined. The conversion factor may be applied to the digital measurements to convert the determined distance in units of digital measurement (e.g., pixels) to units of physical measurement (e.g., length in mm). The measurement in units of physical measurement may indicate an occlusion class and/or a severity of the occlusion class in embodiments.

[0148] FIG. 6 illustrates a flow diagram of an example method 600 for determining a bite classification for a patient from image data of the patient, in accordance with some embodiments of the present disclosure. At block 602 of method 600, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Each stage of treatment may be associated with a particular tooth arrangement in embodiments, and may include 3D models of the upper and/or lower dental arches of the patient having a particular tooth arrangement. In embodiments, method 600 may be performed during orthodontic treatment at a stage of the orthodontic treatment. Accordingly, a current stage of orthodontic treatment may be determined for the patient, and the 3D models associated with the current stage of treatment may be retrieved (e.g., loaded from a data store). Alternatively, method 600 may be performed prior to commencement of a treatment plan. For example, method 600 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 602 are omitted.

[0149] At block 604, processing logic may receive image data. The image data may include one or more images (e.g., two-dimensional (2D) images) of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed. The image data may have been generated by an imaging device of the person (e.g., of a patient), by an imaging device of a third party, and/or generated by an imaging device of a dental practice. In the example of image data captured by a patient during orthodontic treatment, the image data may have been captured by the patient (or a friend or family member of the patient) outside of a doctor office. For example, the patient may have generated one or more selfimages during treatment. The images may have been uploaded to a server (e.g., of a virtual dental care system), which may execute method 600 in some embodiments. In embodiments, the image data comprises one or more lateral bite-closed images of the patient. For example, a lateral bite-closed image showing at least the mandibular canine and premolar and the maxillary canine for one side of the patient’s mouth may be included in the image data.

[0150] At block 605, processing logic processes the image data to identify images that fail to satisfy one or more image criteria, and removes those images that fail to satisfy the image criteria. In one embodiment, processing logic assesses images and filters out images in accordance with the teachings of U.S. Patent Application No. 17/111 ,264, filed December 3, 2020, which is incorporated by reference herein in its entirety.

[0151] At block 606, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth.

Processing logic may additionally perform segmentation of one or more 3D models of the patient’s dental arch(es). For example, processing logic may identify 3D models of the patient’s upper and lower dental arches for a current stage of treatment, and perform segmentation on those 3D models. Alternatively, a treatment plan may include pre-segmented 3D model(s), and processing logic may retrieve such pre-segmented 3D model(s) associated with a current stage of treatment or otherwise associated with the treatment plan and/or the patient. In embodiments, the image may include a maxillary canine and/or molar and a mandibular canine, premolar and first molar for a side of the patient’s mouth.

[0152] At block 607, processing logic registers the image data (e.g., one or more 2D lateral bite- closed images) to the one or more 3D models. In one embodiment, registering the image data to the 3D models includes finding optimal camera parameters and tooth pose parameters to line up teeth of the 2D image with the same teeth in the 3D model(s). In one embodiment, the registration is performed as described in

[0153] At block 608, processing logic identifies a first reference point on the bite-closed image, where the first reference point is associated with one or more maxillary teeth of the patient. In one embodiment, processing logic determines a reference point on the 3D model of the patient’s upper dental arch, and projects the reference point onto the 2D image based on the registration information between the 2D image and the 3D model. For example, processing logic may determine the FACC line for the maxillary canine on the left or right side of the patient’s mouth in the 3D model, and may project the FACC line into the 2D image. The tip of the FACC line (e.g., corresponding to the tip of the maxillary canine) may be determined in the projected FACC line in embodiments. [0154] At block 610, processing logic identifies a second reference point in the bite-closed image, where the second reference point is associated with one or more mandibular teeth of the patient. In one embodiment, processing logic determines a boundary line between two adjacent teeth on the lower dental arch. In one embodiment, the boundary line is a boundary line between the mandibular canine and mandibular premolar on the same side of the patient’s mouth as the maxillary canine associated with the first reference point. A line may be drawn between the first reference point and a point on the boundary line that is closest to the first reference point. The point on the boundary line that it is closest to the first reference point may be the identified second reference point. In embodiments, the line between the first reference point and the second reference point is perpendicular to the boundary line. The line between the first reference point and the second reference may be a virtual marker that corresponds to the distance between the first reference point and the second reference point.

[0155] At block 614, the virtual marker may be projected from the 2D image space of the image to the 3D model space of the 3D model(s). The projection may be performed by determining an angle between the 3D model(s) and an image plane of the 2D image using a trigonometric function in embodiments. Alternatively, the two reference points and/or the virtual marker may be projected onto tooth surfaces in the 3D model, and an updated distance between the two reference points in the 3D model space of the 3D model(s) may be determined.

[0156] In embodiments, the determined distance between the reference points (e.g., length of the projected virtual marker) in units of digital measurement are converted to units of physical measurement (e.g., mm) at block 616. At block 618, processing logic may determine a bite classification and/or severity/level of the bite classification based on the physical measurement. Different physical measurement values may correlate to different levels or stages of one or more types of malocclusion.

[0157] In some embodiments, processing logic may compare the determined bite classification and/or the measurement(s) to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of the malocclusion may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments.

[0158] Processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on.

[0159] FIG. 7A illustrates an example bite-closed side view image 700 of a patient, according to some embodiments of the present disclosure. As shown, bite class can be determined either by the molar position(s) 710 or by the canine position(s) 705. However, it can be difficult for patients to take images that fully show their molars. Accordingly, in some embodiments canine positions are used to assess bite class rather than molar positions. Alternatively, molar positions may be used if a lateral image showing the molar positions of the mandibular and maxillary molars is provided. As shown, the lateral distance 715 between the tip of the maxillary canine and a boundary between the mandibular canine and premolar can indicate a type and/or level of a bite class.

[0160] FIG. 7B illustrates an example bite-closed side view image 720 of a patient used by the method of FIG. 6, according to some embodiments of the present disclosure. Image 720 shows two reference shapes (e.g., virtual markers) drawn onto the image 720. The two reference shapes include an FACC line 722 projected onto the image from a 3D model and a boundary line 724 drawn between the mandibular canine and premolar of the patient in the image. Also shown is a virtual marker 726 representing a shortest distance between the tip of the FACC 722 and the boundary line 724.

[0161] The determined distance between the tip of the FACC line 722 and the boundary line 724 may be compared to one or more distance thresholds in embodiments. The different distance thresholds may be associated with different malocclusion classes and/or severity levels. For healthy dentition, a tip of the FACC line of a maxillary canine is aligned with the interproximal boundary between the corresponding mandibular canine and a mandibular premolar.

[0162] For a class I malocclusion, the bite is normal in terms of molar relationship and/or canine/premolar relationship (e.g., the mesiobuccal cusp of the upper first molar fits into the buccal groove of the lower first molar), but there is crowding, spacing, or other tooth alignment problems. A first threshold distance between the tip of the FACC line 722 and the boundary line 724 may be used for assessing class I malocclusion. The first distance threshold may be, for example, 3 mm, 2 mm, or less. A patient may have a class I malocclusion with mild crowding (e.g., less than 3 mm of crowding), moderate crowding (e.g., 3-6 mm of crowding) or severe crowding (e.g., greater than 6 mm of crowding), with crossbite, arch asymmetry, with overbite, and/or with underbite in embodiments.

[0163] For class II malocclusion, the upper first molar is positioned ahead (anterior) of the lower first molar (e.g., the upper jaw is protruded and/or the lower jaw is retruded). Class II malocclusion may include an overjet and/or overbite. A class II malocclusion can be further divided into class II, division 1 , in which the molars are classified as class II, but the maxillary central incisors are normally inclined or proclined (e.g. upper front teeth are protruded), or class II, division 2, in which the molars are class II and the maxillary central incisors are retroclined (e.g., upper front teeth are tilted backward). To make such assessments of the specific class II division, measurements may be made of the molars and incisors. If the distance between the tip of the FACC line 722 and the boundary line 724 is greater than the first distance threshold and the upper jaw protrudes, then a class II malocclusion may be identified. A severity of the class II malocclusion may be determined based on the distance using one or more additional distance thresholds. For example, if the distance is greater than the first threshold and less than a second distance threshold of 6 mm, then a mild overjet may be identified. If the distance is greater than the second distance threshold but less than a third distance threshold of 9 mm, then a moderate overjet may be identified. If the distance is greater than the third distance threshold, then a severe overjet may be identified.

[0164] For class III malocclusion, the upper first molar is positioned behind (posterior to) the lower first molar (e.g., the lower jaw protrudes or the upper jaw is retruded). This is commonly known as an underbite. If the distance between the tip of the FACC line 722 and the boundary line 724 is greater than zero and the lower jaw protrudes, then a class III malocclusion may be identified. A severity of the class III malocclusion may be determined based on the distance using one or more additional distance thresholds. For example, if the distance is greater than zero and less than a fourth distance threshold of 3 mm, then a mild underbite may be identified. If the distance is greater than the fourth distance threshold but less than a fifth distance threshold of 5 mm, then a moderate underbite may be identified. If the distance is greater than the fifth distance threshold, then a severe underbite may be identified. [0165] The distance between the tip of the FACC line 722 and the boundary line 724 may be periodically measured during orthodontic treatment based on provided images (e.g., patient provided 2D images). For example, a patient may periodically (e.g., once a week) take photos of the patient’s teeth (e.g., using a smartphone or other camera device). These photos may then be analyzed as described herein to determine current distance measurements. Current distance measurements may be compared to prior distance measurements and/or target distance measurements that are part of a treatment plan to determine how a patient’s malocclusion is progressing. If the rate of change of the distance measurement is too slow (e.g., little to no change is detected), then a doctor may determine to change a patient’s treatment plan.

[0166] FIG. 8 illustrates a flow diagram of an example method 800 for determining a bite classification for a patient from image data of the patient, in accordance with some embodiments of the present disclosure. At block 802 of method 800, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Alternatively, method 800 may be performed prior to commencement of a treatment plan. For example, method 800 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 802 are omitted.

[0167] At block 804, processing logic may receive image data. The image data may include one or more images (e.g., two-dimensional (2D) images) of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed. The image data may have been generated by an imaging device of the person (e.g., of a patient), by an imaging device of a third party, and/or generated by an imaging device of a dental practice. In the example of image data captured by a patient during orthodontic treatment, the image data may have been captured by the patient (or a friend or family member of the patient) outside of a doctor office. For example, the patient may have generated one or more selfimages during treatment. The images may have been uploaded to a server (e.g., of a virtual dental care system), which may execute method 800 in some embodiments. In embodiments, the image data comprises one or more lateral bite-closed images of the patient. For example, a lateral bite-closed image showing at least the mandibular canine and premolar and the maxillary canine for one side of the patient’s mouth may be included in the image data.

[0168] At block 805, processing logic processes the image data to identify images that fail to satisfy one or more image criteria, and removes those images that fail to satisfy the image criteria. [0169] At block 806, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth.

[0170] At block 807, processing logic registers the image data (e.g., one or more 2D lateral bite- closed images) to a first 3D model of the upper dental arch of the patient and to a second 3D model of the lower dental arch of the patient. In one embodiment, registering the image data to the 3D models includes finding optimal camera parameters and tooth pose parameters to line up teeth of the 2D image with the same teeth in the 3D model(s). Based on the registration of the first and second 3D models to the image, the relative position and orientation of the first 3D model to the second 3D model may be determined.

[0171] At block 808, processing logic identifies a first reference point on the first 3D model. The first reference point may be, for example, a tip on a maxillary canine. In one embodiment, processing logic determines an FACC line of the maxillary caning, and determines a reference point at the tip of the FACC line.

[0172] At block 810, processing logic identifies a second reference point on the second 3D model. The second reference point may be, for example, a point on a boundary plane between two teeth of the lower dental arch. In one embodiment, processing logic determines a boundary line between the mandibular canine and mandibular premolar in the 2D image. Processing logic then projects the boundary line onto the 3D model of the lower dental arch as a plane extending normal to the image plane of the 2D image. Processing logic may then measure a shortest distance between the first reference point and the plane representing the boundary between the mandibular canine and the mandibular premolar. Alternatively, processing logic may determine a line at an intersection of the plane with the mandibular canine and/or mandibular premolar, and may determine a point that on the line that has a shortest distance to the first reference point.

[0173] At block 814, processing logic may measure a distance between the first reference point and the second reference point. Measurements in the 3D model space of the first and second 3D models may be in units of physical measurement (e.g., mm). At block 816, processing logic may determine a bite classification and/or a level or severity of a malocclusion based on the physical measurement. For example, the bite classification may be determined based on comparison of the measurement(s) to one or more threshold values as described above to determine a malocclusion class and/or severity. Additionally, a severity level of overbite (e.g., measured quantitatively by the amount of vertical overlap of the front teeth) may be determined based on measurements described above. A vertical distance between the tip of an upper tooth (e.g., upper canine) and an opposing lower tooth (e.g., lower canine) may be measured and compared to one or more distance thresholds. For example, if the vertical distance is less than a first vertical distance threshold of 2 mm, then no overbite may be identified. If the vertical distance is greater than the first vertical distance threshold, then a deep bite may be identified. In some embodiments, an overlap percentage threshold is used instead of or in addition to a vertical distance threshold for determining overbite. If a measured overlap percentage is less than a first overlap percentage threshold (e.g., 20% or 30%), then a normal overbite may be identified. If a measured overlap percentage is greater than the first overlap percentage threshold but less than a second overlap percentage (e.g., 50%), then a minor overbite may be identified. If the measured overlap percentage is greater than the second overlap percentage threshold, then a deep bite may be identified. If there is no vertical overlap (i.e., vertical distance is 0 or a negative value), then an open bite may be identified.

[0174] In some embodiments, processing logic may compare the determined bite classification and/or the measurement(s) to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of the malocclusion may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments. Additionally, or alternatively, processing logic may compare the determined bite classification and/or measurements to one or more prior bite classifications and/or measurements for the patient to determine an amount of change that has occurred and/or a rate of change.

[0175] Processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on.

[0176] FIG. 9 illustrates a flow diagram of an example method 900 for determining a bite classification for a patient from image data of the patient using one or more trained machine learning model, in accordance with some embodiments of the present disclosure. In some embodiments, a trained machine learning model may be trained to receive an input image (e.g., which may or may not be a segmented image), and to output an estimation of a bite class, a level of malocclusion between opposing teeth of the upper and lower jaws, a severity of an estimated bite class, and so on.

[0177] At block 902 of method 900, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Alternatively, method 900 may be performed prior to commencement of a treatment plan. For example, method 900 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 902 are omitted.

[0178] At block 904, processing logic may receive image data. The image data may include one or more images (e.g., two-dimensional (2D) images) of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed. The image data may have been generated by an imaging device of the person (e.g., of a patient), by an imaging device of a third party, and/or generated by an imaging device of a dental practice. In the example of image data captured by a patient during orthodontic treatment, the image data may have been captured by the patient (or a friend or family member of the patient) outside of a doctor office. For example, the patient may have generated one or more selfimages during treatment. The images may have been uploaded to a server (e.g., of a virtual dental care system), which may execute method 900 in some embodiments. In embodiments, the image data comprises one or more lateral bite-closed images of the patient. For example, a lateral bite-closed image showing at least the mandibular canine and premolar and the maxillary canine for one side of the patient’s mouth may be included in the image data.

[0179] Processing logic may process the image data to identify images that fail to satisfy one or more image criteria, and may remove those images that fail to satisfy the image criteria. In some instances, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth. Processing logic may additionally perform segmentation of one or more 3D models of the patient’s dental arch(es). For example, processing logic may identify 3D models of the patient’s upper and lower dental arches for a current stage of treatment, and perform segmentation on those 3D models. Alternatively, a treatment plan may include pre-segmented 3D model(s), and processing logic may retrieve such pre-segmented 3D model(s) associated with a current stage of treatment or otherwise associated with the treatment plan and/or the patient. In embodiments, the image may include a maxillary canine and/or molar and a mandibular canine, premolar and first molar for a side of the patient’s mouth.

[0180] At block 906, an input comprising the image (which may or may not include segmentation information of one or more teeth) and/or information from a treatment plan (e.g., 3D model(s) of a current stage of treatment) into a trained machine learning model. The machine learning model may be, for example, a neural network such as a deep neural network, a CNN, and so on. The machine learning model may output an estimate of one or more physical measurements corresponding to distances between two or more reference points. For example, processing logic may output an estimate of a physical measurement corresponding to a distance between a first reference point on the patient’s upper dental arch and a second reference point on the patient’s lower dental arch. In another example, processing logic may output an estimate of an amount of overbite or underbite. Alternatively, processing logic may output an estimate of a bite classification and/or a severity of a malocclusion without providing a physical measurement estimate. In another example, processing logic may output an estimate of a crowding severity for one or more teeth and/or for the upper and/or lower jaw.

[0181] At block 910, processing logic may determine a bite classification (or level of malocclusion between upper and lower teeth) based on the output of the machine learning model. Processing logic may also determine a level of crowding. Processing logic may present the determined information and/or provide suggestions for one or more actions to be performed, such as updates to a treatment plan, a recommendation to undergo orthodontic and/or palatal expansion treatment, a recommendation for surgery, and so on.

[0182] FIG. 10 illustrates a flow diagram of an example method 1000 for determining an amount of posterior crossbite for a patient from image data of the patient, in accordance with some embodiments of the present disclosure. At block 1002 of method 1000, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Alternatively, method 1000 may be performed prior to commencement of a treatment plan. For example, method 1000 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 1002 are omitted.

[0183] At block 1004, processing logic may receive image data. The image data may include one or more images (e.g., two-dimensional (2D) images) of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed. The image data may have been generated by an imaging device of the person (e.g., of a patient), by an imaging device of a third party, and/or generated by an imaging device of a dental practice. In the example of image data captured by a patient during orthodontic treatment, the image data may have been captured by the patient (or a friend or family member of the patient) outside of a doctor office. For example, the patient may have generated one or more selfimages during treatment. The images may have been uploaded to a server (e.g., of a virtual dental care system), which may execute method 1000 in some embodiments. In embodiments, the image data comprises one or more lateral bite-closed images of the patient. For example, a lateral bite-closed image showing at least a mandibular molar and/or premolar and a maxillary molar and/or premolar for one side of the patient’s mouth may be included in the image data.

[0184] At block 1005, processing logic may process the image data to identify images that fail to satisfy one or more image criteria, and may remove those images that fail to satisfy the image criteria. [0185] At block 1006, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth. Processing logic may additionally perform segmentation of one or more 3D models of the patient’s dental arch(es). For example, processing logic may identify 3D models of the patient’s upper and lower dental arches for a current stage of treatment, and perform segmentation on those 3D models. Alternatively, a treatment plan may include pre-segmented 3D model(s), and processing logic may retrieve such pre-segmented 3D model(s) associated with a current stage of treatment or otherwise associated with the treatment plan and/or the patient. In embodiments, the image may include a maxillary canine and/or molar and a mandibular canine, premolar and first molar for a side of the patient’s mouth.

[0186] At block 1008, processing logic measures a tooth height of a maxillary tooth and a tooth height of a mandibular tooth from the segmented image. The tooth heights may be measured in units of digital measurement (e.g., in pixels) in embodiments. In an example, processing logic may measure a tooth height of one or more maxillary molars or premolars and a tooth height of one or more mandibular molars or premolars.

[0187] At block 1010, processing logic determines a first ratio between the maxillary tooth height and the mandibular tooth height.

[0188] At block 1012, processing logic determines a tooth height of the maxillary tooth and a tooth height of the mandibular tooth from one or more 3D models of the treatment plan. These tooth heights may be measured in units of physical measurement, such as mm, in embodiments, and may represent a true ratio of tooth heights between the maxillary and mandibular teeth. Processing logic then determines a second ratio between the maxillary tooth height and the mandibular tooth height from the 3D models of the upper and lower dental arches.

[0189] At block 1014, processing logic compares the first ratio to the second ratio. An existence and/or amount of posterior crossbite may then be determined based on a difference between the ratios. If the first ratio (e.g., image ratio) is equivalent to the second ratio (e.g., actual ratio), this indicates that the image shows the full heights of both teeth - both mandibular and maxillary molars are fully visible in the bite-closed image. If the first ratio is less than the second ratio, then a portion of the maxillary molars has been covered by the mandibular molars in the image, and the patient is exhibiting posterior crossbite. When the first ratio is greater than the second ratio, this indicates the patient does not exhibit posterior crossbite because the maxillary molars are covering a part of the mandibular molars. [0190] In some embodiments, processing logic may compare the determined bite classification and/or the measurement(s) to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of the malocclusion may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments.

[0191] Processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on.

[0192] FIG. 11 illustrates a flow diagram of an example method 1100 for determining an amount of crossbite (e.g., posterior crossbite) for a patient from lateral images of a patient, in accordance with some embodiments of the present disclosure. Method 1100 is described with reference to determining posterior crossbite, but may also be applied to determine anterior crossbite and/or single-tooth crossbite in embodiments. At block 1102 of method 1100, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Alternatively, method 1100 may be performed prior to commencement of a treatment plan. For example, method 1100 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 1102 are omitted.

[0193] At block 1104, processing logic may receive image data. The image data may include at least a lateral bite-open image and a lateral bite-closed image of a person’s dentition. The image data may reflect a person’s dentition prior to orthodontic treatment, during a stage of orthodontic treatment, or after orthodontic treatment is completed.

[0194] Processing logic may process the image data to identify images that fail to satisfy one or more image criteria, and may remove those images that fail to satisfy the image criteria.

[0195] At block 1105, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth. [0196] At block 1106, processing logic measures a first tooth height of a maxillary tooth and a first tooth height of a mandibular tooth from the segmented bite-open image. The tooth heights may be measured in units of digital measurement (e.g., in pixels) in embodiments. In an example, processing logic may measure a tooth height of one or more maxillary molars or premolars and a tooth height of one or more mandibular molars or premolars.

[0197] At block 1108, processing logic determines a first ratio between the first maxillary tooth height and the first mandibular tooth height.

[0198] At block 1110, processing logic measures a second tooth height of the maxillary tooth and a second tooth height of the mandibular tooth from the segmented bite-closed image. The tooth heights may be measured in units of digital measurement (e.g., in pixels) in embodiments. In an example, processing logic may measure a tooth height of one or more maxillary molars or premolars and a tooth height of one or more mandibular molars or premolars.

[0199] At block 1112, processing logic determines a second ratio between the first maxillary tooth height and the first mandibular tooth height.

[0200] At block 1114, processing logic compares the first ratio to the second ratio. An existence and/or amount of posterior crossbite may then be determined based on a difference between the ratios. If the first ratio (e.g., bite-open ratio) is equivalent to the second ratio (e.g., bite-closed ratio), this indicates that the bite-closed image shows the full heights of both teeth - both mandibular and maxillary molars are fully visible in the bite-closed image. If the second ratio is less than the first ratio, then a portion of the maxillary molars has been covered by the mandibular molars in the bite-closed image, and the patient is exhibiting posterior crossbite. When the second ratio is greater than the first ratio, this indicates the patient does not exhibit posterior crossbite because the maxillary molars are covering a part of the mandibular molars in the bite-closed image.

[0201] In some embodiments, processing logic may compare the determined bite classification and/or the measurement(s) to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of the malocclusion may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments.

[0202] Processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on. [0203] Processing logic may perform any of the aforementioned operations to determine a bite classification, level of malocclusion, etc. at multiple points in time during treatment of a patient (e.g., during orthodontic treatment). For each stage of treatment, and/or for each time that measurements are generated, values/measurements for bite classification, level of malocclusion, etc. may be generated. These values/measurements may be tracked and optionally compared to a treatment plan to determine how a patient’s treatment is progressing over time through the course of the treatment. In embodiments, an orthodontic treatment plan may be updated based on a tracked progress of orthodontic treatment.

[0204] FIG. 12 illustrates a flow diagram of an example method 1250 for determining an amount of crossbite (e.g., posterior crossbite) for a patient from an anterior image of a patient, in accordance with some embodiments of the present disclosure. Method 1200 is described with reference to determining posterior crossbite, but may also be applied to determine anterior crossbite and/or single-tooth crossbite in embodiments. At block 1202 of method 1250, processing logic may access a treatment plan of a patient. The treatment plan may be a staged orthodontic treatment plan that includes multiple stages of treatment. Alternatively, method 1250 may be performed prior to commencement of a treatment plan. For example, method 1250 may be performed to identify and/or assess a malocclusion, and may be used to determine whether orthodontic treatment is warranted. In some embodiments, the operations of block 1252 are omitted.

[0205] At block 1254, processing logic may receive image data. The image data may include an anterior image (e.g., an anterior bite-open image or an anterior bite-closed image) of the patient’s dentition.

[0206] At block 1255, processing logic may process the image data to identify images that fail to satisfy one or more image criteria, and may remove those images that fail to satisfy the image criteria. [0207] At block 1256, processing logic processes the received image data using a trained machine learning model that performs image segmentation to segment the image(s) into a plurality of oral structures, which may include teeth, gingiva, and so on. In some embodiments, processing logic further identifies tooth numbers of teeth in the image(s) and assigns the tooth numbers to the teeth. [0208] At block 1258, processing logic identifies a first reference point on the image. The first reference point may be, for example, buccal edge of a maxillary molar.

[0209] At block 1260, processing logic identifies a second reference point on the image. The second reference point may be, for example, buccal edge of a mandibular molar that opposes the mandibular molar. [0210] At block 1262, processing logic may measure a distance between the first reference point and the second reference point. Measurements may be in units of digital measurement (e.g., in pixels) in embodiments.

[0211] At block 1264, processing logic may determine whether the patient has a posterior crossbite based on the digital measurement. Where the mandibular molars are more buccal than the maxillary molars, the patient is exhibiting posterior crossbite. When the maxillary molars are more buccal than the mandibular molars, the crossbite has been corrected.

[0212] In some embodiments, processing logic may register the segmented image to a 3D model of the upper dental arch and/or to a 3D model of the lower dental arch. The registration may be performed after segmenting the 3D model(s) into individual teeth in some embodiments. Based on the registration, a conversion factor for converting the digital measurement into a physical measurement may be determined. The conversion factor may be applied to the digital measurement to determine a physical measurement (e.g., in mm) for the distance between the buccal edges of the mandibular and maxillary molars. The physical measurement may be used to assess a severity of a posterior crossbite in embodiments.

[0213] In some embodiments, processing logic may compare the determined posterior crossbite to a current treatment stage to determine whether an amount of correction of the malocclusion for the current stage of treatment is on track with the treatment plan. Additionally, or alternatively, the level or amount of the posterior crossbite may be compared to a target final dentition to determine a percentage of a total amount of planned correction that has been achieved thus far. In some embodiments, new malocclusions may unexpectedly occur during treatment of other malocclusions. Such newly occurring malocclusions may be flagged in embodiments.

[0214] Processing logic may determine suggestions for one or more actions to be performed. If no treatment has been performed, then the actions may include generating a treatment plan for treating one or more identified malocclusions. If treatment is underway but treatment progress is not tracking a current treatment plan, then one or more adjustments may be made to the treatment plan, such as adding additional stages, removing stages, modifying one or more stages, changing an amount of treatment time associated with one or more stages, modifying a target final dentition arrangement, and so on.

[0215] FIG. 13 illustrates an example process 1300 for characterizing a dental occlusion using, for example, generative techniques, according to some embodiments of the present disclosure. Process 1300 of FIG. 13A is a process that produces data for characterizing crowding, an occlusion (e.g., a level of malocclusion between opposing teeth of the upper and lower jaws), bite and/or other metrics for a dentition of a patient, based on image data of the patient. FIGS. 18A- 18D show some examples of how generative techniques, using synthetic representations, may be used to characterize crowding, occlusion, and/or bite of a patient’s dentition.

[0216] At operation 13.1 of process 1300, image/representation generation, processing logic can intake image data 1302 (e.g., one or more 2D images of a patient’s dentition or teeth), and generate new non-visual representations of teeth (e.g., numerical and/or textual representations such as coordinates for one or more tooth features, tooth dimensions, tooth angles, etc.) and/or new visual representations, such as new images of teeth. The image data (e.g., input image(s)) may include one or more anterior images (e.g., a front facing or front view image of the patient’s dentition) with the patient’s jaws in a bite-open configuration (e.g., teeth not in occlusion) and/or with the patient’s jaws in a bite-closed configuration (e.g., teeth in occlusion). The image data may additionally or alternatively include one or more side-views or lateral images (e.g., left and/or right views) of the patient’s dentition with the patient’s jaws in a bite-open configuration and/or a bite-closed configuration. Different views may be usable to identify different types of malocclusion in embodiments. The image data may have been generated by a device of a patient (e.g., a camera device such as a phone/tablet, a scanning device) and/or of a dental practice (e.g., a camera device such as a phone/tablet, a clinical scanner) in embodiments.

[0217] In embodiments, processing logic includes a trained Al model that processes the input image(s) to generate the new representation(s) of the teeth. In embodiments, the Al model is a generative model, such as a diffusion model or a generator of a generative adversarial network (GAN). In embodiments, the received image(s) of the patient’s dentition comprise a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth. Because of the occlusion, one or more of the measurements described herein may not be immediately possible because one or more features used for the measurements may not be shown. However, the Al model may have been trained on a large sample of training data of prior patients, and may process the input image(s) to generate output images (and/or other new representations) of the teeth that show the regions of the one or more second teeth that were occluded in the input image(s) with a high degree of accuracy. An output of operation 13.1 may be synthetic image data and/or new representations (e.g., new visual representations included in one or more generated images).

[0218] The generated image data 1303 may include one or more anterior images (e.g., a front facing or front view image of the patient’s dentition) with the patient’s jaws in a bite-open configuration (e.g., teeth not in occlusion) and/or with the patient’s jaws in a bite-closed configuration (e.g., teeth in occlusion). The image data may additionally or alternatively include one or more side-views or lateral images (e.g., left and/or right views) of the patient’s dentition with the patient’s jaws in a bite-open configuration and/or a bite-closed configuration. In some embodiments, the generated image data 1303 includes one or more anterior images and/or posterior images with the patient’s jaws in a bite- closed configuration, and with additional contours of the one or more second teeth that are occluded by the one or more first teeth in the bite-closed configuration. These contours (which would ordinarily be hidden in the bite-closed configuration) may be shown using a different visualization than other tooth contours to show that they are hidden contours of the one or more second teeth in embodiments. In some embodiments, the synthetic image data 1303 includes one or more generated images of just those one or more second teeth that were occluded in the original image data 1302. For example, the synthetic image data may include an image that lacks a representation of the one or more first teeth. [0219] In some embodiments, the Al model outputs an image of just the one or more second teeth, showing the new contours of the one or more second teeth. Processing logic may then perform image processing to overlay the synthetic image of the one or more second teeth on an original input image in which the one or more second teeth were occluded. A new combined image may be generated that includes the original contours of the first image and the new contours of the one or more second teeth from the generated second image.

[0220] In some embodiments, the synthetic image data 1303 includes some or all of the above identified types of generated images (e.g., an image of just the one or more second teeth that were occluded in the input image and another image that essentially matches the input image but with the addition of new contours of the one or more second teeth that were occluded in the input image data 1302).

[0221] In some embodiments, prior to image/representation generation 13.1 , processing logic may assess the image data 1302 to determine whether images of the image data 1302 satisfy one or more image criteria. Image criteria may include a blurriness criterion, a criterion that at least a threshold amount of teeth are showing in the image, a sharpness criterion, a criterion that the image data includes one or more particular views of the patient’s dentition (e.g., an anterior bite-open view, an anterior bite-closed view, a lateral bite-open view, a lateral bite-closed view, etc.), and so on. Those images or sets of images that fail to satisfy image criteria may be filtered out. In some embodiments, images may be scored based on sharpness, amount of teeth visible, etc. In some embodiments, one or more highest scoring images (e.g., images that have a score exceeding a threshold value) of one or more views may be selected for analysis. For example, one or more highest scoring anterior bite-open views, one or more highest scoring anterior bite-closed views, one or more highest scoring lateral biteopen views, and/or one or more highest scoring lateral bite-closed views may be selected in embodiments. In some embodiments, if the image data 1302 does not include images that meet image criteria for one or more views, processing logic may output a recommendation to obtain additional images of the one or more views. Processing logic may indicate what the deficiencies of the existing images are to enable an individual to generate images that satisfy the image criteria.

[0222] In some embodiments, the segmentation may be performed on the original image data 1302 before it is input into the Al model. The segmentation may be performed using a trained Al model in embodiments, as described above. The segmentation may be performed, for example, to identify oral structures, such as teeth, gingiva, etc. within the image.

[0223] In some embodiments, at operation 13.2 of process 1300, image segmentation, processing logic can intake synthetic image data 1303 (e.g., one or more 2D generated images of a patient’s dentition or teeth), and segment the image data 1303 to identify oral structures, such as teeth, gingiva, etc. within the image(s), using the techniques described above. For example, image segmentation 13.2 may be performed by one or more trained machine learning (ML) models, such as artificial neural networks (e.g., deep neural networks, convolutional neural networks (CNNs), etc.). During image segmentation 13.2, image data 1302 (e.g., 2D images of a patient’s teeth) can be processed by one or more trained ML models to output segmentation information. In one embodiment, the ML model generates one or more pixel-level segmentation masks of the patient’s teeth. The pixel-level segmentation mask(s) may separately identify each tooth, or may provide a single pixel-level identification of teeth, without separately calling out individual teeth. Accordingly, the ML model may perform semantic segmentation of the image data or instance segmentation of the image data in embodiments. In embodiments, the ML model may output a tooth identification (e.g., a tooth number) for each of the identified teeth. Operation 13.2 can output the segmented image data (segmented data 1304) to operation 13.6. In some embodiments, image segmentation, operation 13.2, is not performed on the synthetic image data 1303. For example, segmented original image data 1302 may be input into the generative Al model that generates the synthetic image data 1303, and the generated synthetic image data 1303 may already include segmentation information in some embodiments.

[0224] At operation 13.3, digital measuring, processing logic can produce digital measurements 1306 from the segmented data 1304 (or from the synthetic image data and/or other new data representations 1303. Notably, the digital measurements may be oral diagnostics measurements that are generated at least in part using information of the at least one region of the one or more second teeth that was occluded in the original image data 1302. For example, the measurements may be made using generated contours of the one or more second teeth, where the generated contours are shown in the synthetic image data 1302 but not in the original image data 1302. Digital measurements 1306 can include pixel distances of features within the segmented data 1304 and/or synthetic image data 1303 in embodiments. For instance, digital measurements 1306 can include the pixel distance measurements of a tooth, or feature, visible within the segmented image data, or segmented data 1304. The digital measurements that are generated may depend on the view of the image (e.g., anterior view, left side view, right side view, etc.) and/or the type of malocclusion to be assessed. In embodiments, multiple different measurements may be made of the segmented image data to assess multiple different classes of malocclusion.

[0225] In an example, the digital measurements may include measurements of tooth heights of one or more exposed teeth and/or portions of teeth. In another example, the digital measurements may include measurements (e.g., vertical and/or horizontal measurements) between two or more features or reference points identified on one or more teeth and/or between one or more teeth. For example, a digital measurement may include a shortest distance between a point on a maxillary canine (e.g., tip of a facial axis of a clinical crown (FACC)) and a boundary line between a mandibular canine and a mandibular first premolar. In another example, a digital measurement may be a measurement of an amount of tooth on a jaw occluded by another tooth (e.g., an adjacent tooth) on that jaw to determine crowding. For example, a horizontal distance measurement may be made between one or more edges of a first tooth that occludes another tooth and one or more edges on the occluded second tooth that are occluded by the first tooth (where those edges are shown in the synthetic image data 1303 or otherwise represented in the new representation(s). In another example, a horizontal distance measurement may be made between one or more contours of a side of an occluded second tooth that are shown in the original image data 1302 (and that are also shown in the synthetic image data 1303) and one or more contours on the side of the occluded second tooth that are occluded in the original image data 1302 but are shown in the synthetic image data 1303. Multiple other digital measurements may also be generated. Any of the aforementioned oral diagnostics measurements may be made, such as to determine crowding, overbite, overjet, underbite, deep bite, open bite, cross bite, malocclusion class and/or severity, and so on.

[0226] At operation 13.4, conversion, processing logic may convert the digital measurements 1306 (i.e., digital oral diagnostics measurements) into physical measurements 1308 (i.e., physical oral diagnostics measurements), using any of the techniques described in detail above.

[0227] In some embodiments, at operation 13.5, correction, a correction is performed to correct for any error introduced by one or more angles between a camera that generated the original image data 1302 and the imaged dentition of the patient. For example, to introduce further accuracy to measurements taken from synthetic image data 1303 and/or segmented image data 1304, the angle at which an image sensor (e.g., a camera) is placed with respect to the teeth can be accounted for. For instance, in addition to, or alternatively to, conversion factor FCON, a correction factor FCOR can be computed to account for the changes in physical size attributable to an angle that the capturing image sensor (e.g., a camera) was held at. The correction factor may be a perspective correction factor determined based on an estimated inclination angle. In embodiments, this correction factor FCOR can be used for improving the conversion factor FCON, the pixel size estimate, and/or any of the digital measurements and/or physical measurements produced from the image data. In embodiments, the correction factor FCOR can be computed to correct for inaccuracies caused by a pitch or camera inclination of the camera (e.g., rotation about the y axis, which may be the left to right axis in the coordinate system of the patient’s dentition) and/or to correct for inaccuracies caused by a yaw of the camera (e.g., rotation about the z axis, which may be the vertical axis in the coordinate system of the patient’s dentition).

[0228] Once oral diagnostics measurements are generated, these measurements may be used to perform one or more assessments of the patient’s dentition, such as to determine a malocclusion class, a malocclusion severity, a level of crowding, cross bite, overbite, open bite, deep bite, and so on. Additionally, crowding measurements may be aggregated across teeth on a jaw to determine an aggregate crowding assessment for that jaw. One or more treatment recommendations may be made based on the determined malocclusions, such as recommendations for palatal expansion treatment and/or orthodontic treatment in embodiments.

[0229] FIG. 14 illustrates a flow diagram of an example method 1400 for performing oral diagnostics measurements of a patient’s dentition, in accordance with some embodiments of the present disclosure. At block 1404 of method 1400, processing logic receives one or more first images of a patient’s dentition. The images may be received from a remote computing device in some embodiments. For example, the images may be images generated by a mobile computing device (e.g., a mobile phone) of a user or by a device of a doctor. Of the one or more first images, at least one first image comprises a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth. In some instances, the occluded regions of the one or more second teeth may not be shown in any other provided image. This can make it difficult to perform oral diagnostics measurements that rely on contours of the one or more second teeth that are hidden.

[0230] Accordingly, at block 1406 processing logic processes the first image(s) to generate one or more second representations of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the at least one first image. This may include, at block 1407, processing the first image(s) to generate one or more second images of the patient’s dentition, the second image(s) comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the at least one first image. Additionally, or alternatively, this may include processing the first image(s) to generate one or more non-visual representations of the one or more second teeth. The non-visual representations may include, for example, numerical and/or textual representations, such as a matrix of values that includes values for the first and/or second teeth. For example, the non-visual representations may include at least one of dimensions or coordinate locations of one or more features of the one or more second teeth. In embodiments, the first image(s) are processed using a trained Al model, such as a generative model. The generative model may be trained to receive input dentition images with occluded teeth and to output updated dentition images that show the occluded teeth (e.g., that show contours of the occluded teeth that were not shown in the input images).

[0231] In one embodiment, the generative Al model is a diffusion model. A diffusion model is a type of generative model that creates new images by starting with random noise, and then gradually removing the noise step-by-step to reveal a structured, realistic image. A diffusion model has two stages, including a forward process, and a reverse process. For the forward process, a real image is introduced, and gradually small amounts of random noise are added to the real image over many steps. After enough steps of adding noise are completed, the image becomes pure random noise. During the reverse process, starting from the pure random noise, the model recreates a new image.

[0232] In one embodiment, the generative Al model outputs an image of just the one or more second teeth, showing the new contours of the one or more second teeth. Processing logic may then perform image processing to overlay the synthetic image of the one or more second teeth on an original input image in which the one or more second teeth were occluded. A new combined image may be generated that includes the original contours of the first image and the new contours of the one or more second teeth from the generated second image. Alternatively, or additionally, the generative Al model may directly output the second image that includes the original contours of the first image and the new contours of the one or more second teeth. For example, the generative Al model may output a third image of the patient’s dentition based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

[0233] In some embodiments, at block 1408 processing logic performs segmentation of the second image(s). The segmentation may be performed to provide contours of different features in the second image(s), which may be used for measurements. In some embodiments, the generated second image(s) already include segmentation information. In some embodiments, the segmentation is performed on the one or more first images before they are processed using the Al model. The Al model may then output segmented image data. [0234] At block 1410, processing logic performs one or more oral diagnostics measurements of the patient’s dentition using the second representation(s) of the one or more second teeth. In other words, one or more oral diagnostics measurements may be performed using the information of the at least one region of the one or more second teeth that is occluded in the first image. For example, processing logic may perform oral diagnostics measurements of the second image(s). This may include comparing original contours of the one or more second teeth to the new contours of the one or more second teeth in some embodiments. The second image(s) may include an image showing both original contours of the one or more second teeth and new contours showing regions of the one or more second teeth that were occluded in the first image(s). Measurements may be made between points on such contours in embodiments.

[0235] FIG. 15A illustrates a flow diagram of an example oral diagnostics measurement, in accordance with some embodiments of the present disclosure. At block 1502 of FIG. 15A, processing logic identifies a first reference point on the one or more first teeth or on one or more original contours of the one or more second teeth. The first reference point may be, for example, a top-most or bottommost point on a tooth, or a left-most or right-most point on the tooth. At block 1504, processing logic identifies a second reference point on the one or new contours of the one or more second teeth. The second reference point may be, for example, a top-most or bottom-most point on a tooth as depicted in the new contours, or a left-most or right-most point on the tooth as shown in the new contours. At block 1506, processing logic measures a distance (e.g., horizontal distance and/or vertical distance) between the first reference point and the second reference point. In some embodiments, the distance may be in units of digital measurement, and may be converted to units of physical measurement as described herein.

[0236] In one embodiment, the one or more first teeth and the one or more second teeth are on a same jaw of the patient. In such an instance, performing one or more oral diagnostics measurements of the patient’s dentition may include measuring a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth. This information may be used in embodiments to determine a crowding level based on the horizontal distance below at block 1414.

[0237] Returning to FIG. 14, in some embodiments, the generated measurements are digital measurements, which may be converted to physical measurements using the techniques described herein. In some embodiments, the generated measurements are physical measurements.

[0238] At block 1412, processing logic may perform a dentition assessment based on the one or more oral diagnostics measurements. This may include identifying whether crowding exists and/or a crowding severity for individual teeth or groups of teeth and/or for a full jaw. This may additionally or alternatively include identifying overbite and/or a severity thereof, overjet and/or a severity thereof, underbite and/or a severity thereof, malocclusion classification, cross bite and/or a severity thereof, and so on. In some instances, the one or more first teeth are on a first jaw of the patient and the one or more second teeth are on a second jaw of the patient that opposes the first jaw. In such instances, processing logic may characterize a level of malocclusion between opposing teeth of the first jaw and the second jaw of the patient based at least in part on the one or more oral diagnostics measurements. Any one or more of the techniques for identifying, and/or determining a severity of, malocclusions described above may be applied in embodiments.

[0239] At block 1414, processing logic outputs a result of the one or more oral diagnostics measurements. The result may show the oral diagnostics measurements in units of digital measurement and/or in units of physical measurement. The processing logic may also output assessment results, such as identified types and/or severity levels of malocclusion. In one embodiment, the first image(s) are output to a display, and the one or more oral diagnostics measurements are output as an overlay on the at least one first image(s). In one embodiment, the second image(s) are output to a display, and the one or more oral diagnostics measurements are overlaid over the one or more second image(s). For example, lines may be drawn on the one or more second image(s) showing features or points from which measurements are made, and showing lines drawn between those features with numerical representation of the measurements in digital units of measurement and/or physical units of measurement.

[0240] In some embodiments, method 1400 may be executed for a patient who is already undergoing treatment (e.g., palatal expansion treatment and/or orthodontic treatment). In such cases, the determined identified crowding, overbite, overjet, underbite, crossbite, etc. values may be compared to past values to determine an amount of improvement and/or a rate of improvement of the patient (e.g., by dividing a difference between a current level from a past level over the amount of time that has passed between when the current level and past level were determined). In some instances, an amount of crowding, overbite, overjet, underbite, crossbite, etc. may be specified for a current stage of treatment. In such instances, processing logic may compare the determined crowding, overbite, overjet, underbite, crossbite, etc. levels to the planned or target crowding, overbite, overjet, underbite, crossbite, etc. levels for the current stage of treatment to determine whether treatment is progressing as planned. In some instances, an amount of crowding, overbite, overjet, underbite, crossbite, etc. to be achieved by the treatment may be specified for a final stage of treatment. In such instances, processing logic may compare the determined crowding, overbite, overjet, underbite, crossbite, etc. levels to the planned or target final crowding, overbite, overjet, underbite, crossbite, etc. levels for the final stage of treatment to determine how much additional treatment is still required to achieve the target. Based on one or more of these comparisons, processing logic may determine whether to adjust a current treatment plan. For example, processing logic may adjust a final target of the treatment plan, may increase or reduce a number of stages of the treatment plan, and so on.

[0241] FIG. 15B illustrates a flow diagram of an example bite class assessment, in accordance with some embodiments of the present disclosure. At block 1522 of FIG. 15B, processing logic determines a level of dental occlusion between the one or more first teeth and the one or more second teeth. The level of dental occlusion may be determined based on a measured horizontal distance between an edge on the one or more first teeth and an edge on the new contours of the one or more second teeth and/or on a measured vertical distance between an edge on the one or more first teeth and an edge on the new contours of the one or more second teeth. At block 1524, processing logic may determine a bite classification based on the level of occlusion.

[0242] Returning to FIG. 14, in one embodiment, at block 1416 processing logic recommends a dental treatment product based on the results of the oral diagnostics measurements. For example, if crowding, crossbite and/or underbite were identified, then palatal expansion treatment may be recommended. In another example, if severe crowding is identified, then a first product and/or treatment such as palatal expansion or oral surgery may be recommended, followed by orthodontic treatment. If moderate or minor crowding is identified, then orthodontic treatment may be recommended without palatal expansion treatment. Techniques for assessing whether palatal expansion treatment or other treatment is recommended are described below with reference to FIGS. 16A-B and 18.

Additionally, in some embodiments processing logic predicts a length of dental treatment (e.g., a length of palatal expansion treatment and/or orthodontic treatment) based on the result of the oral diagnostics measurements. For example, a longer treatment time may be predicted for a more severe malocclusion than for a less severe malocclusion.

[0243] At block 1418, the output results and/or recommendations may be stored in a data store, output to a display, and/or transmitted to a remote device for display and/or storage thereon. For example, the results may be generated by a server computing device, and may be transmitted to a computing device of a doctor and/or patient.

[0244] FIG. 16A illustrates a flow diagram of an example method 1600 for assessing tooth crowding and recommending dental treatment, in accordance with some embodiments of the present disclosure. In some embodiments, operations 1608-1616 of method 1600 may be performed at blocks 1410 and/or 1412 of method 1400.

[0245] At block 1602 of method 1600, processing logic receives one or more first images of a patient’s dentition. The images may be received from a remote computing device in some embodiments. For example, the images may be images generated by a mobile computing device (e.g., a mobile phone) of a user or by a device of a doctor. Of the one or more first images, at least one first image comprises a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth. In some instances, the occluded regions of the one or more second teeth may not be shown in any other provided image. This can make it difficult to perform oral diagnostics measurements that rely on contours of the one or more second teeth that are hidden.

[0246] At block 1604, processing logic processes the first image(s) to generate one or more second representations of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the at least one first image. This may include at block 1606 processing the first image(s) to generate one or more second images of the patient’s dentition, the second image(s) comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the at least one first image. Additionally, or alternatively, this may include processing the first image(s) to generate one or more non-visual representations of the one or more second teeth. In embodiments, the first image(s) are processed using a trained Al model, such as a generative model, that outputs the second representation(s).

[0247] At block 1608, processing logic selects a set of two or more adjacent teeth that are included in the one or more second representations. At block 1610, processing logic measures a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth. An example of this measurement is shown in FIG. 18C. At block 1612, processing logic determines a crowding level of the set of teeth based on the measured horizontal distance. In some embodiments, the crowding level for the set of teeth may be compared to one or more thresholds to determine a severity of the crowding level. If the horizontal distance is greater than a first threshold, then minor crowding may be identified. If the horizontal distance is greater than a higher second threshold, then moderate crowding may be identified. If the horizontal distance is greater than an even higher third threshold, then severe crowding may be identified for the set of teeth.

[0248] At block 1614, processing logic may determine whether crowding measurements/levels have been determined for of the teeth that are included in the representations of teeth in the image data (e.g., in the first image(s) and/or second image(s). If crowding measurements have not been performed for some teeth, the method may return to block 1608, at which one or more additional teeth may be selected for measurement. This process may be repeated until measurements have been made for all sets of adjacent teeth. If at block 1614 a determination is made that all visible teeth have been assessed, the method continues to block 1616.

[0249] At block 1616, processing logic determines an aggregate crowding level for a jaw of the patient based on the crowding levels for the multiple sets/pairs of adjacent teeth. In embodiments, a separate aggregate crowding level may be determined for the upper jaw and the lower jaw.

[0250] In some embodiments, at block 1617 processing logic performs one or more additional oral diagnostics measurements as described herein. These measurements may include, for example, horizontal measurements and/or vertical measurements between features on one or more teeth on the same jaw and/or on opposing jaws. Processing logic may then determine levels of other types of malocclusion, such as levels of overbite, underbite, crossbite (including different types of crossbite), and/or overjet.

[0251] At block 1618, processing logic may recommend a dental treatment product (e.g., palatal expansion treatment and/or orthodontic treatment) based on the aggregate crowding level and/or on individual crowding levels and/or based on the one or more other determined levels of malocclusion (e.g., based on a level of overbite, overjet, underbite, crossbite, and so on). For example, if a child is determined to have an underbite, crossbite and/or crowding, then palatal expansion treatment may be recommended. Processing logic may additionally predict a length of treatment necessary to address the determined level of crowding and/or other levels of malocclusion. In some embodiments, the length of treatment is predicted via a lookup table that relates levels of crowding and/or other malocclusion levels to treatment times. In some embodiments, the length of treatment is predicted by inputting the aggregate crowding level(s), individual crowding level(s), and/or other malocclusion levels into a trained Al model (e.g., a convolutional neural network (CNN)), which outputs the predicted length of dental treatment. In some embodiments, the length of dental treatment includes a length of palatal expansion treatment and/or a length of orthodontic treatment.

[0252] FIG. 16B illustrates a flow diagram of an example method 1620 for determining whether to recommend palatal expansion treatment, in accordance with some embodiments of the present disclosure. Method 1620 may be particularly useful for determining whether to recommend palatal expansion treatment for youths. At block 1622 of method 1620, processing logic receives one or more first images of a patient’s dentition. Of the one or more first images, at least one first image comprises a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth.

[0253] At block 1632, processing logic processes the first image(s) to generate one or more second representations of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the at least one first image. This may include processing the first image(s) to generate one or more second images of the patient’s dentition, the second image(s) comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the at least one first image. Additionally, or alternatively, this may include processing the first image(s) to generate one or more non-visual representations of the one or more second teeth. In embodiments, the first image(s) are processed using a trained Al model, such as a generative model, that outputs the second representation(s).

[0254] At block 1640, processing logic performs one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth (e.g., using the new contours of the at least one region in the second image(s)).

[0255] At block 1642, processing logic determines an available amount of space on an upper jaw based on the oral diagnostics measurements. In some embodiments, the available amount of space is determined by measuring an arch width and/or a length along a perimeter of the arch. In some embodiments, horizontal gap measurements may be made between adjacent teeth using the measurement techniques described herein. The horizontal gap measurements for all of the pairs of teeth may be aggregated to determine an aggregate gap level. In some embodiments, crowding measurements are represented as a negative gap measurement, and an aggregate crowding level of the teeth includes a combination of any gap values and any crowding values. For example, if some teeth are crowded together with an overlap value of 1 mm and other teeth on the same jaw are spaced apart and have a gap value of 1 mm, then the overall combined aggregate crowding and gap level would be 0. An available amount of space may be separately determined for the upper jaw and lower jaw in some embodiments.

[0256] At block 1644, processing logic predicts an amount of space needed for the upper jaw based at least in part on the one or more oral diagnostics measurements. For example, processing logic may determine an aggregate crowding level as described with reference to FIG. 16A. Processing logic may then compare the aggregate crowding level to a value associated with the available amount of space. In some embodiments, an available amount of space is determined based on measuring gaps as set forth above, and an aggregate gap level is computed and compared to the aggregate crowding level.

[0257] An available amount of space may be at least in part dependent on whether a patient has any unerupted or erupting permanent teeth. Permanent teeth are generally larger than primary teeth, and thus take up more space on a dental arch. Accordingly, in some embodiments at block 1646 processing logic identifies any primary teeth, erupting teeth and/or unerupted teeth in the upper jaw and/or in the lower jaw. An identification of a primary tooth may correspond to an identification of an unerupted tooth in some embodiments. In some embodiments, erupting, unerupted and/or primary teeth are identified using a trained Al model (e.g., such as a CNN). The trained Al model may receive as an input one or more images (e.g., the first image(s) and/or second image(s)), and may output an indication of a number of primary, uninterrupted and/or erupting teeth in the image. IN some embodiments, the Al model performs segmentation and generated segmentation masks identifying any erupting, unerupted and/or primary teeth.

[0258] At block 1648, processing logic may compute a statistical distribution of tooth size and/or a lateral space used by teeth of the patient. This may be computed based on performing oral diagnostics measurements of teeth in embodiments to measure a width of the teeth (e.g., mesial-distal tooth width). The measured tooth sizes may then be averaged to determine an average tooth size. In embodiments, teeth that are identified as primary teeth are not included in the average. In some embodiments, different tooth size averages are determined for different types of teeth and/or regions on the dental arch. For example, a different average size may be determined for the molars than for the incisors. In some embodiments, average tooth size is determined at least in part from other patient data (e.g., pooled patient data from a large body of patients). In some embodiments, a logistic regression is performed based on the sizes of the existing teeth to estimate the sizes of the not yet erupted permanent teeth. In embodiments, an amount of lateral space used by teeth of the patient may be determined based on measuring lateral spaces between existing teeth of the patient. Alternatively, or additionally, lateral space between teeth may be determined from pooled patient data. In one embodiment, one or more tooth dimensions and/or a tooth shape (e.g., optionally including an amount of lateral space used by one or more teeth) are generated according to U.S. Patent No. 9,744,001 , issued August 29, 2017, which is incorporated by reference herein in its entirety.

[0259] At block 1650, processing logic predicts tooth sizes of the erupting and/or unerupted teeth and/or lateral space to be used by the erupting/unerupted teeth based on the statistics distribution computed at block 1648. Processing logic may determine a tooth size and amount of lateral space needed for each permanent tooth associated with an identified primary tooth, erupting tooth and/or unerupted tooth. This information may be aggregated across all of the erupting/unerupted teeth, and may be added to the space used by the already present permanent teeth.

[0260] At block 1652, processing logic compares the amount of available space to the predicted amount of space. If the amount of available space is less than the predicted amount of space, then the method may proceed to block 1654 and palatal expansion treatment may be recommended. If the amount of available space is greater than or equal to the predicted necessary amount to space, then the method may proceed to block 1652 and no palatal expansion treatment may be recommended. [0261] FIG. 17 illustrates a flow diagram of an example method 1700 for determining whether to recommend palatal expansion treatment, in accordance with some embodiments of the present disclosure. At block 1704 of method 1700, processing logic receives one or more first images of a patient’s dentition. Of the one or more first images, at least one first image comprises a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth.

[0262] At block 1706, processing logic processes the one or more first images to perform one or more oral diagnostics measurements of the patient’s dentition. The oral diagnostics measurements may be performed using any of the techniques described herein (e.g., those that include use of a generative model to generate a second representation of the first image(s), those that use an open bite and closed bite image to make measurements, and so on).

[0263] In one embodiment, at block 1708 processing logic processes the first image(s) to generate one or more second representations of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the at least one first image. This may include processing the first image(s) to generate one or more second images of the patient’s dentition, the second image(s) comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the at least one first image. Additionally, or alternatively, this may include processing the first image(s) to generate one or more non-visual representations of the one or more second teeth. In embodiments, the first image(s) are processed using a trained Al model, such as a generative model, that outputs the second representation(s).

[0264] At block 1710, processing logic performs one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth (e.g., using the new contours of the at least one region in the second image(s)).

[0265] At block 1712, processing logic processes the first image(s) and/or second image(s) to identify primary teeth vs. permanent teeth, and/or to identify erupting and/or unerupted teeth using the techniques described above. At block 1714, processing logic measures an available amount of space in the upper jaw as described above. At block 1715, processing logic predicts an amount of space needed in the upper jaw as described above. At block 1716, processing logic determines whether the available space is sufficient for replacement of the primary teeth in the upper jaw with permanent teeth.

[0266] At block 1730, processing logic determines whether to recommend palatal expansion treatment based on the one or more oral diagnostics measurements. If at block 1716 processing logic determined that the available space is not sufficient for replacement of the primary teeth with permanent teeth, then palatal expansion treatment may be recommended. If processing logic determined that the available space is sufficient for replacement of the primary teeth with permanent teeth, then palatal expansion treatment may not be recommended. In some embodiments, at block 1731 processing logic may estimate a length of a recommended palatal expansion treatment based on a difference between the available space and the predicted needed space on the upper jaw. The greater the difference between the available space and the predicted needed space, the more palatal expansion that is called for and the longer the treatment.

[0267] At block 1732, processing logic may output a recommendation with respect to palatal expansion treatment. For example, processing logic may output a recommendation to undergo palatal expansion treatment or to not undergo palatal expansion treatment based on the determination at block 1730. At block 1734, processing logic may store a result of the measurements and/or the recommendation in a data store. Additionally or alternatively, processing logic may transmit the results and/or recommendations to a remote computing device (e.g., a device of a doctor and/or patient).

[0268] FIG. 18A illustrates an example overbite measurement process using a single input image, according to some embodiments of the present disclosure. As shown in FIG. 18A, a closed bite image 1805 may be input into a generative model 1810A. In the closed bite image, teeth 8 and 9 may at least partially occlude teeth 24 and 25. The generative model 1810A may process closed bite image 1805 and output synthetic image 1815 in one embodiment. Synthetic image 1815 may include just a representation of tooth 24 and tooth 25, and may not show tooth 8 and tooth 9. As shown, synthetic image 1815 includes new contours of tooth 24 and of tooth 25 that are hidden in the closed bite image 1805. Synthetic image 1815 may then be merged with closed bite image 1805 to generate image 1820, which may be a combined image that shows the original contours of the teeth 8, 9, 24, 25 as well as the new contours of teeth 8, 9. As shown, the new contours 1822 of teeth 24, 25 may be represented using a different visualization than the original contours. For example, the new contours 1822 may be shown with a dashed line and the original contours may be shown with a solid line. In some embodiments, generative model 1810A outputs image 1820 directly based on processing of closed bite image 1805. In such an embodiment, generative model 1810A may optionally also output image 1815. As shown, an oral diagnostics measurement may be made between one or more features (e.g., a tooth edge) on the teeth 8, 9 and one or more features (e.g., a tooth edge) on the new contours for teeth 8, 9. In the illustrated example, the measurement is an overbite measurement 1825. However, the measurement may also be a crossbite measurement, an underbite measurement, and so on when different images showing different teeth are used.

[0269] FIG. 18B illustrates an example overbite measurement process using pair of input images, according to some embodiments of the present disclosure. As shown in FIG. 18B, a closed bite image 1805 and an open bite image 1830 may be input into a generative model 1810B. In the closed bite image 1805, teeth 8 and 9 may at least partially occlude teeth 24 and 25. In the open bite image 1830, the teeth 24, 25 may not be occluded by the teeth 8, 9. The generative model 181 OB may process closed bite image 1805 and open bite image 1830 and output synthetic image 1815 in one embodiment. Synthetic image 1815 may then be merged with closed bite image 1805 to generate image 1820, which may be a combined image that shows the original contours of the teeth 8, 9, 24, 25 as well as the new contours 1822 of teeth 8, 9. In some embodiments, generative model 1810B outputs image 1820 directly based on processing of closed bite image 1805 and open bite image 1830. In such an embodiment, generative model 1810B may optionally also output image 1815. As shown, an oral diagnostics measurement may be made between one or more features (e.g., a tooth edge) on the teeth 8, 9 and one or more features (e.g., a tooth edge) on the new contours for teeth 8, 9. In the illustrated example, the measurement is an overbite measurement 1825. However, the measurement may also be a crossbite measurement, an underbite measurement, and so on when different images showing different teeth are used.

[0270] FIG. 18C illustrates an example crowding measurement process using a single input image, according to some embodiments of the present disclosure. As shown in FIG. 18C, an image 1840 showing multiple teeth on a same jaw may be input into a generative model 1810C. In the image 1840, tooth 8 may at least partially occlude tooth 9. The generative model 1810C may process image 1840 and output synthetic image 1845 (e.g., of tooth 8) and/or synthetic image 1850 of tooth 9 in one embodiment. The synthetic image 1850 of tooth 9 may include new contours 1857 for tooth 9 that are hidden by tooth 8 in image 1840. Synthetic image 1850 may then be merged with synthetic image 1845 and/or input image 1840 to generate image 1860, which may be a combined image that shows the original contours of the teeth 8, 9 as well as the new contours 1857 of tooth 9. In some embodiments, generative model 1810C outputs image 1860 directly based on processing of image 1840. In such an embodiment, generative model 1810C may optionally also output image 1845 and/or image 1850. As shown, an oral diagnostics measurement may be made between one or more features (e.g., a tooth edge) on tooth 8 and one or more features (e.g., a tooth edge) on the new contours for tooth 9. In the illustrated example, the measurement is a crowding measurement 1860. However, the measurement may also another type of measurement.

[0271] FIG. 18D illustrates an example measurement of a synthetic image 1862, according to some embodiments of the present disclosure. Similar to what is shown in FIGS. 18A-C, synthetic image 1862 may have been generated by a generative model based on an input of an image in which one or more teeth were occluded by one or more other teeth. The synthetic image 1862 may include new contours 1880 for the one or more occluded teeth as well as original contours 1879 for one or more of the teeth. As shown, an oral diagnostics measurement may be made between one or more features (e.g., a tip of the upper canine 1865) that were visible in the input image and one or more features (e.g., a tip of the lower canine 1870) on the new contours that were not visible in the input image. In the illustrated example, the measurement may be used to measure overjet or a malocclusion class and/or severity level.

[0272] FIG. 19 illustrates a flow diagram of an example method for training an Al model to generate modified images of dentition usable for oral diagnostics measurements, in accordance with some embodiments of the present disclosure. At block 1902 of method 1900, processing logic identifies one or more 3D models of a patient’s dentition. The 3D model(s) may be retrieved from a data store containing millions of prior patients. The data store may contain 3D model(s) of the patient’s dentition and/or images (e.g., 2D images) of the patient’s dentition. The images may include images taken by the patients themselves and/or by a dental practitioner.

[0273] At block 1904, processing logic may receive (e.g., retrieve) a first image of the patient’s dentition from the images stored in the data store. Alternatively, the first image may be generated by projecting the 3D model(s) onto a 2D image plane. In either case, the first image may include a representation of one or more first teeth of a patient and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth. The first teeth and the second teeth may be on the same jaw or on opposing jaws.

[0274] In one embodiment, at block 1906 processing logic determines a jaw pose and camera parameters to be used for a generated image. The jaw pose may include a position and orientation of an upper jaw and/or of a lower jaw. The jaw pose may include, for example, a side view jaw pose, a front view jaw pose, an open bite, a closed bite pose, and so on. The camera parameters may include a distance and/or orientation of a camera relative to the 3D model(s), a field of view of the camera, and so on.

[0275] At block 1908, processing logic may perturb one or more teeth on one or both of the 3D models (e.g., on an upper jaw model and/or a lower jaw model). Perturbing the teeth may include rotating one or more teeth about up to 3 axes of rotation and/or moving the one or more teeth along up to three axes. In some instances, the perturbations are randomly performed. In some instances, the perturbations are manually performed by a technician.

[0276] At block 1910, processing logic projects the 3D model(s) (optionally having been updated based on the perturbations performed at block 1908) onto an image plane using the jaw pose and the camera parameters. This may generate the first image from the 3D model(s). In the generated image, the one or more second teeth are at least partially occluded by the one or more first teeth.

[0277] At block 1914, processing logic projects the one or more 3D models onto an image plane defined by the first image to generate a second image. The second image comprises a second representation of at least the one or more second teeth. In one embodiment, the second image comprises the contours of the one or more first teeth from the first image as well as the contours of the one or more second teeth from the first image. Additionally, the second image comprises new contours of the one or more second teeth that are occluded by the one or more first teeth in the first image. The new contours of the one or more second teeth may be shown using a different visualization than original contours to enable differentiation between the original contours and the new contours.

[0278] In some embodiments, processing logic generates the second image as well as a third image. The third image may include representations of the one or more second teeth but not of the one or more first teeth. One or more other images may also be generated, such as an image comprising a representation of the one or more first teeth without the one or more second teeth.

[0279] In some embodiments, rather than generating a second and/or third image, processing logic generates one or more non-visual representations of the original and/or new contours of the one or more second teeth. For example, processing logic may generate a numerical and/or textual representation of the one or more second teeth. In an example, a generated representation may include comprises at least one of dimensions or coordinate locations of one or more features of the one or more second teeth.

[0280] In one embodiment, to generate the second and/or third image at block 1916 processing logic performs a 2D to 3D registration between the first image and the 3D model(s) to align teeth in the first image with those teeth in the 3D model(s). At block 1919, processing logic may then determine camera parameters and a jaw pose for the first image based on the registration. The alignment may be performed when the first image is a received image rather than an image generated from the 3D model(s). If the first image was generated from the 3D model(s), then the jaw pose and camera parameters that were used to generate the first image from the 3D model(s) would already be known. [0281] At block 1920, processing logic then uses the camera parameters and the jaw pose to render the second image. The second image may be rendered in such a way that the new contours of the one or more second teeth that are not shown in the first image are shown in the second image. These contours may be known from the 3D model(s). A set of an input image and one or more target output images may be a training data item.

[0282] In some embodiments, processing logic additionally uses the camera parameters and the jaw pose to render a third image and/or one or more additional images. The third image may show just the teeth that were occluded in the first image, or may show just one tooth from the first image.

[0283] In some embodiments, processing logic additionally determines at least one of tooth locations, tooth orientations, tooth sizes or tooth geometry from the one or more 3D models. This information may be added to a training data item as additional input data. [0284] The operations of blocks 1902-1920 may be performed many times to generate many different input images and corresponding output images that may be used to form a distinct training data item. The training data items may be combined to form a training dataset.

[0285] At block 1932, processing logic trains an Al model using the first image as an input (and optionally the additional input data for the data item) and the corresponding second image (and/or third or additional image) of a training data item as a target output. The Al model may be trained to process input images comprising representations of dentition including occluded teeth and to generate output images showing contours of the occluded teeth that are occluded in the input images. The Al model may additionally be trained to output one or more additional images based on processing of the input image. For example, a first output image may include the original contours of the occluding teeth and the occluded teeth, along with new contours of the occluded teeth that were now shown in the input image. A second output image may include just the teeth that were occluded in the input image, with new contours of the previously occluded teeth. A third and/or further output image may include just one occluded or occluding tooth from the input image.

[0286] In embodiments, the Al model can be a generative Al model, such as a diffusion model or a GAN. In one embodiment, such an Al model can include one or more artificial neural networks (also referred to simply as a neural network). The artificial neural network can be, for example, a convolutional neural network (CNN) or a deep neural network.

[0287] In some embodiments, the artificial neural network(s) can generally include a feature representation component with a classifier or regression layers that map features to a target output space. A convolutional neural network (CNN), for example, can host multiple layers of convolutional filters. Pooling can be performed, and non-linearities can be addressed, at lower layers, on top of which a multi-layer perceptron is commonly appended, mapping top layer features extracted by the convolutional layers to decisions (e.g., classification outputs). The neural network can be a deep network with multiple hidden layers or a shallow network with zero or a few (e.g., 1-2) hidden layers. Deep learning is a class of machine learning algorithms that use a cascade of multiple layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. Neural networks can learn in a supervised (e.g., classification) and/or unsupervised (e.g., pattern analysis) manner. Some neural networks (e.g., such as certain deep neural networks) can include a hierarchy of layers, where the different layers learn different levels of representations that correspond to different levels of abstraction. In deep learning using such networks, each level can learn to transform its input data into a slightly more abstract and composite representation. In embodiments of such neural networks, such layers may not be hierarchically arranged (e.g., such neural networks can include structures that differ from a traditional layer-by-layer approach).

[0288] As indicated above, the Al model can include one or more generative Al models, allowing for the generation of new and original content, such a generative Al model can include aspects of a transformer architecture, or a generative adversarial network (GAN) architecture. Such a generative Al model can use other machine learning models including an encoder-decoder architecture including one or more self-attention mechanisms, and one or more feed-forward mechanisms. In some embodiments, the generative Al model can include an encoder that can encode input data into a vector space representation; and a decoder that can reconstruct the data from the vector space, generating outputs with increased novelty and uniqueness. Further detail regarding Al models and the training thereof is provided below.

[0289] Training of an Al model such as a neural network may be achieved in a supervised learning manner, which involves feeding a training dataset consisting of labeled inputs through the network, observing its outputs, defining an error (by measuring the difference between the outputs and the label values), and using techniques such as deep gradient descent and backpropagation to tune the weights of the network across all its layers and nodes such that the error is minimized. In many applications, repeating this process across the many labeled inputs in the training dataset yields a network that can produce correct output when presented with inputs that are different than the ones present in the training dataset. In high-dimensional settings, such as large images, this generalization is achieved when a sufficiently large and diverse training dataset is made available.

[0290] For the model training, a training dataset containing hundreds, thousands, tens of thousands, hundreds of thousands or more input and output image pairs should be used to form a training dataset. In embodiments, up to millions of cases of patient dentition that may have underwent a prosthodontic procedure and/or an orthodontic procedure may be available for forming a training dataset, where each case may include various labels of one or more types of useful information.

[0291] To effectuate training, processing logic inputs the training dataset(s) into one or more untrained machine learning models. Prior to inputting a first input into a machine learning model, the machine learning model may be initialized. Processing logic trains the untrained machine learning model(s) based on the training dataset(s) to generate one or more trained machine learning models that perform various operations as set forth above.

[0292] Training may be performed by inputting one or more of the images into the machine learning model one at a time. Each input may include data from an image in a training data item from the training dataset. The training data item may include, for example, an image that includes one or more first teeth that are fully shown and one or more second teeth that are partially or fully obscured by the one or more first teeth. The data that is input into the machine learning model may include a single layer (e.g., just a single image) or multiple layers (e.g., an open bite and closed bite image, or one or more images plus additional information about a patient’s teeth).

[0293] The machine learning model processes the input to generate an output. An artificial neural network includes an input layer that consists of values in a data point (e.g., intensity values and/or height values of pixels in a height map). The next layer is called a hidden layer, and nodes at the hidden layer each receive one or more of the input values. Each node contains parameters (e.g., weights) to apply to the input values. Each node therefore essentially inputs the input values into a multivariate function (e.g., a non-linear mathematical transformation) to produce an output value. A next layer may be another hidden layer or an output layer. In either case, the nodes at the next layer receive the output values from the nodes at the previous layer, and each node applies weights to those values and then generates its own output value. This may be performed at each layer. A final layer is the output layer. In a generative model, the output layer may be a final representation, which may be a visual representation (e.g., a new image that shows the one or more contours of the one or more second teeth that were obscured in the input image and/or a non-visual representation of the one or more second teeth).

[0294] Processing logic may then compare the generated output to the known target (e.g., the target image or images of the training data item). Processing logic determines an error (i.e., a classification error) based on the differences between the output image and the provided target image. Processing logic adjusts weights of one or more nodes in the machine learning model based on the error. An error term or delta may be determined for each node in the Al model. Based on this error, the Al model adjusts one or more of its parameters for one or more of its nodes (the weights for one or more inputs of a node). Parameters may be updated in a back propagation manner, such that nodes at a highest layer are updated first, followed by nodes at a next layer, and so on. An artificial neural network contains multiple layers of “neurons”, where each layer receives as input values from neurons at a previous layer. The parameters for each neuron include weights associated with the values that are received from each of the neurons at a previous layer. Accordingly, adjusting the parameters may include adjusting the weights assigned to each of the inputs for one or more neurons at one or more layers in the artificial neural network.

[0295] Once the model parameters have been optimized, model validation may be performed to determine whether the model has improved and to determine a current accuracy of the model. After one or more rounds of training, processing logic may determine whether a stopping criterion has been met. A stopping criterion may be a target level of accuracy, a target number of processed images from the training dataset, a target amount of change to parameters over one or more previous data points, a combination thereof and/or other criteria. In one embodiment, the stopping criteria is met when at least a minimum number of data points have been processed and at least a threshold accuracy is achieved. The threshold accuracy may be, for example, 70%, 80% or 90% accuracy. In one embodiment, the stopping criteria is met if accuracy of the machine learning model has stopped improving. If the stopping criterion has not been met, further training is performed. If the stopping criterion has been met, training may be complete. Once the machine learning model is trained, a reserved portion of the training dataset may be used to test the model.

[0296] At block 1934, processing logic may store a trained Al model. The trained Al model may then be implemented in a server or local computing device to perform the methods described herein. [0297] FIG. 20A illustrates a work flow 2000 for generating training data usable to train an Al model to generate modified images of dentition, in accordance with some embodiments of the present disclosure. As shown an image 2005 of a patient’s dentition and one or more 3D model(s) 2010 of the patient’s dentition may be registered together at block 2015 using a 2D to 3D registration process. From the registration, camera parameters 2025 and jaw pose 2030 for the image 2005 may be determined. Additionally, in some instances one or more perturbations 2020 may be made to teeth of the 3D model(s) to generate updated 3D model(s). The determined camera parameters 2025 and jaw pose 2030 may be used to render a synthetic full tooth mask 2042 from the 3D model(s) 2010, such as by projecting the 3D model(s) onto an image plane defined by the camera parameters and/or jaw pose. In some instances, the synthetic full tooth mask 2042 includes representations of all of the teeth from image 2025. In some instances, the rendering 2040 is performed using the new 3D model(s) optionally with updated teeth in view of the perturbations represented in the new 3D model(s). In some instances, the rendering 2040 is performed to generate a synthetic partial tooth mask 2044. The synthetic partial tooth mask 2044 may include just the one or more teeth that are occluded in the image 2005. In some instances, the rendering 2040 is performed to generate a synthetic full jaw (or jaw pair) rendering 2046 that includes both the occluding teeth from the image 2005 and the occluded teeth from the image 2005, including the new contours of the occluded teeth not shown in image 2005. The new contours may be shown using a different visualization than original contours in the synthetic full jaw rendering 2046.

[0298] FIG. 20B illustrates another work flow 2050 for generating training data usable to train an Al model to generate modified images of dentition, in accordance with some embodiments of the present disclosure. As shown no starting image of a patient’s dentition is provided for work flow 2050. Instead, just one or more 3D model(s) 2010 of the patient’s dentition are provided. Optionally, one or more teeth in the 3D model(s) may be perturbed 2020 to updated 3D model(s) 2010. A random or pseudorandom process 2052 may be performed to generate camera parameters 2025 and jaw pose 2030 to be used for rending an image. At block 2040, one or more rendering may be generated using the 3D model(s) 2010, the camera parameters 2025, and the jaw pose 2030. The determined camera parameters 2025 and jaw pose 2030 may be used to render a synthetic full tooth mask 2042 from the 3D model(s) 2010, such as by projecting the 3D model(s) onto an image plane defined by the camera parameters and/or jaw pose. In some instances, the synthetic full tooth mask 2042 includes representations of all of the teeth visible within the camera parameters and jaw pose. The synthetic full tooth mask 2042 may be used as an input image for training an Al model. In some instances, the rendering 2040 is performed using the new 3D model(s) optionally with updated teeth in view of the perturbations represented in the new 3D model(s). In some instances, the rendering 2040 is performed to generate a synthetic partial tooth mask 2044. The synthetic partial tooth mask 2044 may include just the one or more teeth that are occluded in the image 2005. The partial tooth mask 2042 may show new contours of the one or more teeth that are not shown in the synthetic full tooth mask 2042 in embodiments. In some instances, the rendering 2040 is performed to generate a synthetic full jaw (or jaw pair) rendering 2046 that includes both the occluding teeth from the image 2005 and the occluded teeth from the image 2005, including the new contours of the occluded teeth not shown in synthetic full tooth mask 2042.

[0299] FIG. 21 illustrates a flow diagram of an example method 2100 for modifying an orthodontic treatment plan, in accordance with some embodiments of the present disclosure.

[0300] At block 2110, method 2100 may include receiving patient data. In some embodiments, receiving patient data may include receiving patient data comprising one or more progress images associated with an orthodontic treatment plan.

[0301] As illustrated with callout block 2112, the processing device performing method 2100 may receive patient data that includes image data.

[0302] At block 2120, method 2100 may include processing the patient data. In some embodiments, processing the patient data may include processing the patient data to determine a level of progression associated with the orthodontic treatment plan based on processing of the image data as discussed with reference to any of FIGS. 1A-20B.

[0303] At block 2130, method 2100 may include modifying an orthodontic treatment plan and/or palatal expansion treatment plan. In some embodiments, an orthodontic treatment plan and/or palatal expansion treatment plan may be modified in response to the determined level of progression.

[0304] In some embodiments, the processing device performing method 2100 may modify an orthodontic treatment plan and/or palatal expansion treatment plan to advance a patient to a next treatment stage earlier than planned, or retain a patient in a current treatment stage longer than planned. In some cases, modifying an orthodontic treatment plan to advance a patient to a treatment stage, or retain a patient in a treatment stage may include modifying the orthodontic treatment plan in response to the determined level of progression comprises advancing a patient associated with the patient data to a subsequent stage of the orthodontic treatment plan or retaining the patient within a current stage of the orthodontic treatment plan. In some embodiments, processing logic may adjust the treatment plan, such as by adding one or more additional stages of treatment, removing one or more stages of treatment, modifying a tooth arrangement for one or more stages of treatment, modifying attachment placement for one or more stages of treatment, and so on.

[0305] At block 2140, method 2100 may include generating a notification. In some embodiments, generating a notification may include generating a notification of the modified orthodontic treatment plan.

[0306] In some embodiments, orthodontic treatment and/or palatal expansion treatment may have already been completed, and a patient may be wearing a retainer to prevent the patient’s teeth from reverting towards their original state. In such instances, the oral diagnostics measurements may be used to determine whether a patient’s teeth are regressing by comparing one or more calculated occlusion metric values (e.g., overbite level, overjet level, crowding level, etc.) with one or more target occlusion metric values. If the metric values worsen over time, then a patient may be recommended to return to their doctor for possible follow-up treatment. Additionally, or alternatively, a new retainer may be recommended to accommodate the patient’s regressed dentition (e.g., if the original retainer no longer fits comfortably.

[0307] FIG. 22 illustrates an example system architecture 2200 capable of supporting an occlusion monitor for monitoring an occlusion and/or patient bite of a patient during implementation of an orthodontic treatment plan, in accordance with embodiments of the present disclosure.

[0308] The treatment system architecture 2200 can correspond to system 2300 of FIG. 23, in embodiments. For example, system architecture 2300 can include a dental consumer/patient system 1590, a dental professional system 2380 and a virtual dental care system 2370. In embodiments, virtual dental care system 2370 includes a treatment plan coordination platform 2220, an occlusion monitor 2250 and/or a storage platform 2240. In embodiments, dental consumer/patient system 2390 includes a client device 2230B. In embodiments, dental professional system 2380 includes a client device 2230A. In some embodiments, any of the treatment plan coordination platform 2220, the client device(s) 2230A-C, the storage platform 2240, and/or the occlusion monitor 2250, can include, can be, or can otherwise be connected to one or more computing devices (such as a rackmount server, a router computer, a server computer, a personal computer, a mainframe computer, a laptop computer, a tablet computer, a desktop computer, etc.), one or more storage devices (e.g., hard disks, memories, databases), networks, software components, and/or hardware components capable of connecting to system 2200. In embodiments a platform can support any number of discrete software or hardware, or combination of such, portions, which can be referred to as modules.

[0309] In some embodiments, network 2201 can connect the various platforms and/or devices, which can include a public network (e.g., the Internet), a private network (e.g., a local area network (LAN) or wide area network (WAN)), a wired network (e.g., Ethernet network), a wireless network (e.g., an 802.11 network or a Wi-Fi network), a cellular network (e.g., a Long Term Evolution (LTE) network), routers, hubs, switches, server computers, and/or a combination thereof.

[0310] In some embodiments, the treatment plan coordination platform 2220 can facilitate or host services for coordinating HCP-patient communications relating to an on-going treatment plan for a patient. In embodiments, the treatment plan coordination platform 2220 can host, leverage, and/or include several modules for supporting such system functionalities. For instance, in embodiments, platform 2220 can support and/or integrate a control module (not shown in FIG. 22), for performing overall control of modules and devices associated with the platform, and a user-interface (Ul) control module (not shown in FIG. 22), for performing generation, and other processes associated with a Ul that will be presented through associated client devices. Platform 2220 can support a data processing module (not shown in FIG. 22), that can gather and manage data from storage and modules (such as patient data and/or plan data gathered from storage device 2244 and/or storage platform 2240).

Platform 2220 can also process, transmit, and/or receive incoming and outgoing data from client device 2230A and/or client device 2230B. Such modules can work collaboratively, and communicate internally or externally (e.g., to further systems and/or through APIs), to facilitate virtual meeting or communication capabilities for users across a range of client devices. Each module can include hardware, firmware, and/or software configured to provide a described functionality.

[0311] In some embodiments, platform 2220 (or an integrated control module) can orchestrate the overall functioning of the treatment coordination platform 2220. In some cases, platform 2220 can include algorithms and processes to direct the setup, data transfer, and processing for providing and receiving data associated with a treatment plan from connected devices (e.g., the client device 2230A- B). For example, when a user initiates engagement with the treatment plan coordination system 2200, the platform 2220 can initiate and manage the associated processes, including allocating resources, determining routing pathways for data and data streams, managing permissions, and so forth to interact with client devices to establish and maintain reliable connections and data transfer.

[0312] Platform 2220 can include a Ul controller, and can perform user-display functionalities of the system such as generating, modifying, and monitoring the individual Ul (s) and associated components that are presented to users of the platform 2220 through a client device. For example, a Ul control module can generate the Ul(s) (e.g., Ills 2234A-B of client devices 2230A-B) that users interact with while engaging with the treatment coordination system.

[0313] A III can include many interactive (and/or non-interactive) visual elements for display to a user. Such visual elements can occupy space within a III and can be visual elements such as windows displaying video streams, windows displaying images, chat panels, file sharing options, participant lists, and/or control buttons for controlling functions such as client application navigation, file upload and transfer, controlling communications functions such as muting audio, disabling video, screen sharing, etc. The III control module can work to generate such a III, including generating, monitoring, and updating the spatial arrangement and presentation of such visual elements, as well as working to maintain functions and manage user interactions, together with the platform 2220. Additionally, the III control module can adapt a user-interface based on the capabilities of client devices. In such a way the III control module can provide a fluid and responsive interactive experience for users of the treatment coordination platform.

[0314] In some embodiments, a data processing module can be responsible for storage and management of data. This can include gathering and directing data from client devices. In embodiments, the data processing module can communicate and store data, including to and/or from storage platforms and storage devices (e.g., such as storage device 2244), etc. For instance, once an initial treatment plan (e.g., initial treatment plan 2260) has been established, platform 2220 can perform tasks such as gathering and directing such data to the storage platform 2240, and/or to client devices 2230A-B.

[0315] In embodiments, data that is transmitted, managed, and/or manipulated by the system can include any kind of data associated with a treatment plan, including (e.g., treatment plan schedules, dates, times, etc.), patient data (e.g., such as images, values, sensor data, etc.), and so on.

[0316] In embodiments, the system 2200 can leverage an occlusion monitor 2250 for performing processes associated with data collected by the client devices. In embodiments, the occlusion monitor 2250 can include a dataset generator 2256 and an analysis module 2258. Occlusion monitor 2250 can intake collected data 2262 (e.g., collected data from a patient’s client device) and process such data to generate observations 2264, or progress indicators, extracted from the collected data 2262, and generate responses 2266. In a non-limiting example, occlusion monitor 2250 can intake collected data 2262 that can include image data of a patient’s oral cavity. Occlusion monitor 2250 can then extract observations 2264 from the image data and/or sensor data, such as an observation of a class I malocclusion, a class II malocclusion, a class III malocclusion, a deep bite, an underbite, a posterior crossbite, crowding, and so on. The occlusion monitor 2250 can intake this data and effect an appropriate response 2266 (e.g., such as generate a notification for a patient or HCP, recommend a treatment, and so on). In embodiments, responses 2266 generated by the occlusion monitor 2250 can include updates to the initial treatment plan 2260, notifications sent to any of the client devices or platforms or modules associated with the system, recommendations for orthodontic treatment and/or palatal expansion treatment to be performed, and so on. Occlusion monitor 2250 may store data associated with the observation and response and/or transmit the data to client devices 2230A-B and/or other devices.

[0317] Dataset generator 2256 can collect and organize collected data 2262 from one or more patients, observations 2264 produced by occlusion monitor 2250, and/or responses 2266 as produced by the occlusion monitor 2250. In embodiments, dataset generator 2256 can store data, or generate a dataset, with discretized segments corresponding to individual patient profiles. Analysis module 2258 can then analyze the collected data to identify significant trends, characterizations corresponding to specific treatment plans, associated data segments, and insights within the data.

[0318] In some embodiments, one or more client devices (e.g., client devices 2230A-B) can be connected to the system 2200. In embodiments, the client device(s) can each include computing devices such as personal computers (PCs), laptops, mobile phones, smart phones, tablet computers, netbook computers, notebook computers, network-connected televisions, etc. In some embodiments, client device(s) can also be referred to as “user devices.”

[0319] In embodiments, client devices (e.g., client devices 2230A-B) connected to the system can each include a client application (e.g., client application 2232A-B). In some embodiments, the client application can be an application that provides the user interface (Ul) (e.g., client application 2234A-B) and manages transmissions, inputs, and data to and from platform 2220. In some embodiments, the client application that provides the Ul can be, or can include, a web browser, a mobile application, a desktop application, etc.

[0320] Client devices, under direction by the treatment coordination platform when connected, can present or display a Ul (e.g., Ul 2234A-B) to a user of the respective client device. In embodiments, the Ul can be generated locally at the client device, e.g., through client applications 2232A-B. A Ul can include various visual elements and regions, and can be the primary mechanism by which the user interfaces with the client application, the treatment plan coordination platform, and the system at large. In some embodiments, the Ul(s) of the client device(s) can include multiple visual elements and regions that enable presentation of information, for decision-making, content delivery, etc. to a user of the device. In some embodiments, the Ul can be referred to as a graphical user interface (GUI).

[0321] In some embodiments, the system (or any associated platforms), can transmit any data, including audio, video, image, and textual data, to the client device to be interpreted by client application 2232A-B, and displayed via the Ul of the respective client device. Such data that can be transmitted to the client device through client applications 2232A-B can include, for example, III information, textual information, video, or audio streaming data associated with the HCP-patient communications, control, or navigation data, etc. In some embodiments, a client application 2232A-B (e.g., a dedicated application) incorporated within the client devices 2230A-B and can perform function associated with the end-user interface.

[0322] In embodiments, connected client devices 2230A-B can also collect input from users through input features. Input features can include III features, software features, and/or requisite hardware features (e.g., mouse and keyboard, touch screens, etc.) for inputting user requests, and/or data to the treatment plan coordination system. Input features of client devices 2230A-B can include space, regions, or elements of the III 2234A-B that accept user inputs. For example, input features can be visual elements such as buttons, text-entry spaces, selection lists, drop-down lists, control panels, etc.

[0323] In embodiments, connected client devices 2230A-B can also collect input from an associated media system 2236A-B e.g., a camera, microphone, and/or similar elements of a client device, to transmit or intake further user-inputs. In embodiments, the media system of the client device can include at least a display, a microphone, speakers, and a camera, etc., together with other media elements as well. Such elements (e.g., speakers, or a display) can further be used to output data, as well as intake data or inputs.

[0324] In embodiments, a client application (e.g., client application 2232A-B) can execute a series of protocols to access and control media system hardware resources, in some cases accessing devicelevel APIs or drivers that interact with the underlying hardware of a media system. Through such, or similar, protocols, client applications can utilize any of the components of a client device media system for specific functionalities within the context of virtual dental care. For instance, in embodiments, a display of the media system can be employed by the client application (under direction from the treatment coordination platform 2220) to render the III. In embodiments, graphical elements can be presented or displayed to the user via the display and the III. The client application of a device can direct rendering commands to the display to update the screen with relevant visual information.

Similarly, and/or simultaneously, in embodiments, a camera or imaging sensor of the media system can capture image and/or video input from the user to transmit. In embodiments, the client application can process, encode, and transmit such data from the client device, over the network, to the treatment plan coordination platform 2220.

[0325] As will be discussed further below, in embodiments, a client application 2232B associated with a patient client device 2230B can transfer patient data (including captured audio and/or image data) associated with the treatment plan to treatment plan coordination platform 2220, which can

-n- forward, process and/or store such data. In embodiments, such data can be forwarded from a first client device to a second client device.

[0326] As will be further discussed below, in embodiments, data collected from a patient client device 2230B can be stored in storage device 2244 as collected data 2262. Such collected data 2262 can include collected data 2262 associated with a single patient, and a single patient dental treatment plan. Alternatively, data associated with multiple patients and/or multiple dental treatment plans and separate procedures can be stored as individual data segments of collected data 2262.

[0327] In embodiments, a first client device 2230B can gather data and inputs from a patient, to be transmitted and displayed to an HCP at a second client device 2230A. For instance, in embodiments, client device 2230B can belong to a patient, while client device 2230A can belong to an HCP. In embodiments, such a pairing and configuration can facilitate communication and data transfer between both parties. For example collected patient data from client device 2230B can be transmitted and displayed to an HCP at client device 2230A, which can then transmit instructions, guidance, or any other kinds of data back to the patient client device 2230B. In embodiments, such data can include updates to a treatment plan.

[0328] lin embodiments, data can be gathered from an integrated camera of the media system 2236B of client device 2230B. For example, in some embodiments, client device 2230B can be a personal phone or similar device. In some embodiments, media system 2236B can access, include, or be a part of an image sensor or scanner for obtaining two-dimensional (2D) data of a dental site in a patient’s oral cavity (or another imaging device including a camera) and can be operatively connected to a personal client device (e.g., client device 2230B). In some embodiments, more than two, including any number of, client devices can be used to gather and monitor oral health data from the patient.

[0329] In embodiments herein, patient data collected by a client device of a user, as described with respect to devices 2230A-B, can be holistically referenced as collected data 2262, which can be ultimately stored within storage device 2244.

[0330] In some embodiments, the system can include storage platform 2240, which can host and manage storage device 2244. In some embodiments, platform 2240 can be a dedicated server for supporting storage device 2244 accessible via network 2201.

[0331] In embodiments, collected data 2262 can include any data that has been collected from client devices associated with the system. In embodiments, the collected data 2262 can be data collected from one or more patient’s before, after, or during a dental treatment plan. In embodiments, such data can be accessible and displayable via any of the connected client devices.

[0332] In embodiments, collected data 2262 can include the oral health data acquired through multiple sources and/or at different times. In embodiments, collected data 2262 can be any kind of data associated with a patient’s oral health, and/or data that is relevant for a treatment plan, such as 2D images of the patient’s dentition.

[0333] In embodiments, such collected data can include spatial positioning data, including 2D or 3D patient data. In embodiments, collected data can include image data which can be used to generate a virtual model (e.g., a virtual 2D model or virtual 3D model) of the real-time conditions of the patient’s oral features and/or dentition (e.g., conditions of a tooth, or a dental arch, etc., can be modeled).

[0334] In embodiments, storage device 2244 can further include an initial treatment plan 2260, and observations 2264 and responses 2266, as produced by occlusion monitor 2250 (and/or analysis module 2258).

[0335] In embodiments, the initial treatment plan 2260 can function as, or be an initial, pre-defined treatment plan that consists of scheduled stages designed to sequentially correct and improve aspects of a patient's health. The initial treatment plan may include palatal expansion treatment and/or orthodontic treatment in embodiments. In some embodiments, the initial treatment plan can be a plan for improving aspects of a patient’s oral health. In some cases, the plan can be an initial plan determined by an HCP, and based on portions of collected data 2262, such as tests, documentation, medical history, etc. For example, in embodiments the initial treatment plan can be a multi-stage palatal expansion treatment plan initially been generated by an HCP (e.g., an orthodontist) after performing a scan of an initial pre-treatment condition of the patient’s dental arch. In some embodiments, the initial treatment plan can begin at home (e.g., be based on a patient scan of his- or her-self) or at a scanning center. In embodiments, the initial treatment plan might be created automatically and/or by a professional (including an orthodontist) in a remote service center.

[0336] In embodiments, the initial dental treatment plan can be an orthodontic treatment plan based on intraoral scan data providing surface topography data for the patient's intraoral cavity (including teeth, gingival tissues, etc.). Such surface topography data can be generated by directly scanning the intraoral cavity, a physical model (positive or negative) of the intraoral cavity, or an impression of the intraoral cavity, using a suitable scanning device (e.g., a handheld scanner, desktop scanner, etc.), as was previously described. One of ordinary skill in the art, having the benefit of this disclosure, will appreciate that numerous methods, mechanisms, and strategies for generating an initial dental treatment plan exist, and that the discussed methods represent exemplary methods, mechanisms, and strategies for generating an initial dental treatment plan associated with the system. [0337] An orthodontic procedure can refer to, inter alia, any procedure involving the oral cavity and directed to the design, manufacture, or installation of orthodontic elements at a dental site within the oral cavity, or a real or virtual model thereof, or directed to the design and preparation of the dental site to receive such orthodontic elements. Such elements can be appliances including but not limited to brackets and wires, retainers, aligners, or functional appliances. In embodiments, various orthodontic aligner and/or palatal expansion devices can be formed, or one device can be modified, for each treatment stage to provide forces to move the patient’s teeth or jaw. The shape of such device(s) can unique and customized for a particular patient and a particular treatment stage.

[0338] In embodiments, one or more stages of the dental and/or orthodontic treatment plan can correspond to a dental appliance (e.g., orthodontic aligner) that the patient must wear for a predetermined period. In some embodiments, such a period, or time interval, can range from one day to three weeks. For example, in some cases, the treatment can begin with the first aligner, tailored to fit a patient's current dental configuration. Such an initial aligner can apply targeted pressure on regions of the patient’s teeth, initiating the process of gradual tooth repositioning. Once the patient has worn this initial aligner for the duration specified in the first stage of the initial treatment plan 2260, the patient can transition to subsequent stage (e.g., the subsequent stage in a sequence of stages). This can involve replacing the initial aligner with a new one, designed to continue the process of tooth repositioning. Subsequent stages can introduce a new orthodontic aligner, manufactured to incrementally move teeth closer to the desired final position.

[0339] Such an initial treatment plan 2260 can include checkpoints or assessment periods, where HCPs and/or dental professionals assess the progress of the treatment. During such checkpoints, digital scans, images, molds, etc., can be taken to ensure that the palate is expanding according to the planned trajectory. In embodiments, such checkpoints, or assessments, can occur during or in between stages of a given dental treatment plan. In embodiments, the dental treatment plan can prescribe, or outline specific time intervals between checkpoints. In some embodiments, any of the previously discussed collected data types can be collected during such checkpoints.

[0340] In some embodiments, any, or all of data within storage device 2244 can be accessed and modified by treatment coordination platform 2220 (or other modules and platforms of the system), for further processing.

[0341] As was discussed, in some embodiments, analysis module 2258 can include an Al model to analyze collected data 2262, and produce treatment plans or updates.

[0342] In embodiments, such an Al model can be one or more of decision trees (e.g., random forests), support vector machines, logistic regression, K-nearest neighbor (K NN), or other types of machine learning models, for example. In one embodiment, such an Al model can be one or more artificial neural networks (also referred to simply as a neural network). The artificial neural network can be, for example, a convolutional neural network (CNN) or a deep neural network. 10343] In one embodiment, processing logic performs supervised machine learning to train the neural network.

[0344] In some embodiments, the artificial neural network(s) can generally include a feature representation component with a classifier or regression layers that map features to a target output space. A convolutional neural network (CNN), for example, can host multiple layers of convolutional filters. Pooling can be performed, and non-linearities can be addressed, at lower layers, on top of which a multi-layer perceptron is commonly appended, mapping top layer features extracted by the convolutional layers to decisions (e.g., classification outputs). The neural network can be a deep network with multiple hidden layers or a shallow network with zero or a few (e.g., 1-2) hidden layers. Deep learning is a class of machine learning algorithms that use a cascade of multiple layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. Neural networks can learn in a supervised (e.g., classification) and/or unsupervised (e.g., pattern analysis) manner. Some neural networks (e.g., such as certain deep neural networks) can include a hierarchy of layers, where the different layers learn different levels of representations that correspond to different levels of abstraction. In deep learning using such networks, each level can learn to transform its input data into a slightly more abstract and composite representation. In embodiments of such neural networks, such layers may not be hierarchically arranged (e.g., such neural networks can include structures that differ from a traditional layer-by-layer approach).

[0345] In some embodiments, such an Al model can be one or more recurrent neural networks (RNNs). An RNN is a type of neural network that includes a memory to enable the neural network to capture temporal dependencies. An RNN is able to learn input-output mappings that depend on both a current input and past inputs. The RNN will address past and future measurements and make predictions based on this continuous measurement information. One type of RNN that can be used is a long short-term memory (LSTM) neural network.

[0346] As indicated above, such an Al model can include one or more generative Al models, allowing for the generation of new and original content, such a generative Al model can include aspects of a transformer architecture, or a generative adversarial network (GAN) architecture. Such a generative Al model can use other machine learning models including an encoder-decoder architecture including one or more self-attention mechanisms, and one or more feed-forward mechanisms. In some embodiments, the generative Al model can include an encoder that can encode input textual data into a vector space representation; and a decoder that can reconstruct the data from the vector space, generating outputs with increased novelty and uniqueness. The self-attention mechanism can compute the importance of phrases or words within a text data with respect to all of the text data. A generative Al

-SI- model can also utilize the previously discussed deep learning techniques, including recurrent neural networks (RNNs), convolutional neural networks (CNNs), or transformer networks. Further details regarding generative Al models are provided herein.

[0347] In some embodiments, analysis module 2258 includes one or more modules for identifying reference points in images and/or 3D models and generating one or more measurements based on such reference points. For example, analysis module 2258 may perform the operations discussed with reference to any of FIGS. 1A-20B.

[0348] In some embodiments, storage device 2244 can be hosted by one or more storage devices, such as main memory, magnetic or optical storage-based disks, tapes or hard drives, network- attached storage (NAS), storage area network (SAN), and so forth. In some embodiments, storage device 2244 can be a network-attached file server, while in other embodiments, storage device 2244 can be or can host some other type of persistent storage such as an object-oriented database, a relational database, and so forth.

[0349] In some embodiments, storage device(s) 2244 can be hosted by any of the platforms or device associated with system 2200 (e.g. treatment plan coordination platform 2220). In other embodiments, storage device 2244 can be on or hosted by one or more different machines coupled to the treatment coordination platform via network 2201 . In some cases, the storage device 2244 can store portions of audio, video, image, or text data received from the client devices (e.g., client device 2230A-B) and/or any platform and any of its associated modules.

[0350] In some embodiments, any one of the associated platforms (e.g., treatment plan coordination platform 2220) can temporarily accumulate and store data until it is transferred to storage devices 2244 for permanent storage.

[0351] It is appreciated that in some implementations, the functions of platforms 2220 and/or 2240 can be provided by a fewer number of machines. For example, in some implementations, functionalities of platforms 2220 and/or 2240 can be integrated into a single machine, while in other implementations, functionalities of platforms 2220 and/or 2240 can be integrated into multiple, or more, machines. In addition, in some implementations, only some platforms of the system can be integrated into a combined platform.

[0352] While the modules of each platform are described separately, it should be understood that the functionalities can be divided differently or integrated in various ways within the platform while still applying similar functionality for the system. Furthermore, each platform and associated modules can be implemented in various forms, such as standalone applications, web-based platforms, integrated systems within larger software suites, or dedicated hardware devices, just to name a few possible forms. [0353] In general, functions described in embodiments as being performed by platforms 2220, 2240, and/or occlusion monitor 2250 can also be performed by client devices (e.g., client device 2230A, client device 2230B). In addition, the functionality attributed to a particular component can be performed by different or multiple components operating together. Platforms 2220, 2240, and/or occlusion monitor 2250 can also be accessed as a service provided to other systems or devices through appropriate application programming interfaces, and thus is not limited to use in websites. [0354] It is appreciated that in some implementations, platforms 2220, 2240, and/or occlusion monitor 2250 or client devices of the system (e.g., client device 2230A, client device 2230B) and/or storage device 2244, can each include an associated API, or mechanism for communicating with APIs. In such a way, any of the components of system 2200 can support instructions and/or communication mechanisms that can be used to communicate data requests and formats of data to and from any other component of system 2200, in addition to communicating with APIs external to the system (e.g., not shown in FIG. 22).

[0355] In some embodiments of the disclosure, a “user” can be represented as a single individual. However, other implementations of the disclosure encompass a “user” being an entity controlled by a set of users and/or an automated source. For example, a set of individual users federated as a community in a social network can be considered a “user.” In another example, an automated consumer can be an automated ingestion pipeline, such as a topic channel.

[0356] In situations in which the systems, or components therein, discussed here collect personal information about users, or can make use of personal information, the users can be provided with an opportunity to control whether the system or components collect user information (e.g., information about a user’s social network, social actions or activities, profession, a user’s preferences, or a user’s current location), or to control whether and/or how to receive content from the system or components that can be more relevant to the user. In addition, certain data can be treated in one or more ways before it is stored or used, so that personally identifiable information is removed. For example, a user’s identity can be treated so that no personally identifiable information can be determined for the user, or a user’s geographic location can be generalized where location information is obtained (such as to a city, ZIP code, or state level), so that a particular location of a user cannot be determined. Thus, the user can have control over how information is collected about the user and used by the system and components.

[0357] FIG. 23 shows a block diagram of an example system for virtual dental care for orthodontic treatment and/or palatal expansion treatment, in accordance with some embodiments. As shown in FIG. 23, system 2300 can include a dental consumer/patient system 2302, a dental professional system 2350, a virtual dental care system 2306, and a network 2304. The dental consumer/patient system 2302, dental professional system 2350, and virtual dental care system 2306 can communicate to one another over the network 2304, which can include one or more local area networks (LANs), public wide area networks (e.g., such as the Internet) and/or private wide area networks (e.g., Intranets), one or more personal Area Networks (PAN), one or more cellular networks (e.g., a Global System for Mobile Communications (GSM) network), and/or any other suitable network.

[0358] Dental consumer/patient system 2302 generally represents any type or form of computing device capable of reading computer-executable instructions. Dental consumer/patient system 2302 can be, for example, a desktop computer, a tablet computing device, a laptop, a smartphone, an augmented reality device, or other consumer device. Additional examples of dental consumer/patient system 2302 include, without limitation, laptops, tablets, desktops, servers, cellular phones, Personal Digital Assistants (PDAs), multimedia players, embedded systems, wearable devices (e.g., smart watches, smart glasses, etc.), smart vehicles, smart packaging (e.g., active or intelligent packaging), gaming consoles, I nternet-of-Things devices (e.g., smart appliances, etc.), variations or combinations of one or more of the same, and/or any other suitable computing device. The dental consumer/patient system 2302 need not be or include a clinical scanner (e.g., an intraoral scanner), though it is contemplated that in some implementations the functionalities described herein in relation to the dental consumer/patient system 2302 can be incorporated into a clinical scanner. As an example of various implementations, a camera 2332 of the dental consumer/patient system 2302 can comprise an ordinary camera that captures 2D images of the patient's dentition and does not capture height-map and/or other data (e.g., three-dimensional (3D) data) that is used to stitch a mesh of a 3D surface. In some examples, the dental consumer/patient system 2302 can include an at-home intraoral scanner.

[0359] In some implementations, the dental consumer/patient system 2302 is configured to interface with a dental consumer and/or dental patient. A “dental consumer,” as used herein, can include a person seeking assessment, diagnosis, and/or treatment for a dental condition (general dental condition, orthodontic condition, endodontic condition, condition requiring restorative dentistry, etc.). A dental consumer can, but need not, have agreed to and/or started treatment for a dental condition. A “dental patient,” as used herein, can include a person who has agreed to diagnosis and/or treatment for a dental condition. A dental consumer and/or a dental patient, can, for instance, be interested in and/or have started orthodontic treatment, such as treatment using one or more (e.g., a sequence of) aligners (e.g., polymeric appliances having a plurality of tooth-receiving cavities shaped to successively reposition a person's teeth from an initial arrangement toward a target arrangement). In various implementations, the dental consumer/patient system 2302 provides a dental consumer/dental patient with software (e.g., one or more webpages, standalone applications, mobile applications, etc.) that allows the dental consumer/patient to capture images of their dentition, interact with dental professionals (e.g., users of the dental professional system 2350), manage treatment plans (e.g., those from the virtual dental care system 2306 and/or the dental professional system 2350), and/or communicate with dental professionals (e.g., users of the dental professional system 2280).

[0360] Dental professional system 2350 generally represents any type or form of computing device capable of reading computer-executable instructions. Dental professional system 2350 can be, for example, a desktop computer, a tablet computing device, a laptop, a smartphone, an augmented reality device, or other consumer device. Additional examples of dental professional system 2350 include, without limitation, laptops, tablets, desktops, servers, cellular phones, Personal Digital Assistants (PDAs), multimedia players, embedded systems, wearable devices (e.g., smart watches, smart glasses, etc.), smart vehicles, smart packaging (e.g., active or intelligent packaging), gaming consoles, I nternet-of-Things devices (e.g., smart appliances, etc.), variations or combinations of one or more of the same, and/or any other suitable computing device.

[0361] In various implementations, the dental professional system 2350 is configured to interface with a dental professional. A “dental professional” (used interchangeably with dentist, orthodontist, and doctor herein) as used herein, can include any person with specialized training in the field of dentistry, and can include, without limitation, general practice dentists, orthodontists, dental technicians, dental hygienists, etc. A dental professional can include a person who can assess, diagnose, and/or treat a dental condition. “Assessment” of a dental condition, as used herein, can include an estimation of the existence of a dental condition. An assessment of a dental condition need not be a clinical diagnosis of the dental condition. In some embodiments, an “assessment” of a dental condition can include an “image-based assessment,” that is an assessment of a dental condition based in part or on whole on photos and/or images (e.g., images that are not used to stitch a mesh or form the basis of a clinical scan) taken of the dental condition. A “diagnosis” of a dental condition, as used herein, can include a clinical identification of the nature of an illness or other problem by examination of the symptoms.

“Treatment” of a dental condition, as used herein, can include prescription and/or administration of care to address the dental conditions. In particular, embodiments are directed to prescription and/or administration of treatment with respect to palatal expansion. The dental professional system 2350 can provide to a user software (e.g., one or more webpages, standalone applications (e.g., dedicated treatment planning and/or treatment visualization applications), mobile applications, etc.) that allows the user to interact with users (e.g., users of the dental consumer/patient system 2302, other dental professionals, etc.), create/modify/manage treatment plans (e.g., those from the virtual dental care system 2306 and/or those generated at the dental professional system 2350), etc.

[0362] Virtual dental care system 2306 generally represents any type or form of computing device that is capable of storing and analyzing data. Virtual dental care system 2306 can include a backend database server for storing patient data and treatment data. Additional examples of virtual dental care system 2306 include, without limitation, security servers, application servers, web servers, storage servers, and/or database servers configured to run certain software applications and/or provide various security, web, storage, and/or database services. Although illustrated as a single entity in FIG. 23, virtual dental care system 2306 can include and/or represent a plurality of servers that work and/or operate in conjunction with one another.

[0363] As illustrated in FIG. 23, dental consumer/patient system 2302, virtual dental care system, 2306, and/or dental professional system 2350 can include one or more memory devices, such as memory 2340. Memory 2340 generally represents any type or form of volatile or non-volatile storage device or medium capable of storing data and/or computer-readable instructions. In one example, memory 2340 can store, load, execute in conjunction with physical processor(s) 2330, and/or maintain one or more of virtual dental care modules 2308. Examples of memory 2340 include, without limitation, Random Access Memory (RAM), Read Only Memory (ROM), flash memory, Hard Disk Drives (HDDs), Solid-State Drives (SSDs), optical disk drives, caches, variations, or combinations of one or more of the same, and/or any other suitable storage memory.

[0364] As illustrated in FIG. 23, dental consumer/patient system 2302, dental professional system 2350, and/or virtual dental care system 2306 can also include one or more physical processors, such as physical processor(s) 2330. Physical processor(s) 2330 generally represents any type or form of hardware-implemented processing unit capable of interpreting and/or executing computer-readable instructions. In one example, physical processor(s) 2330 can access and/or modify one or more of virtual dental care modules 2308 stored in memory 2340. Additionally or alternatively, physical processor 2330 can execute one or more of virtual dental care modules 2308 to facilitate preamble phrase. Examples of physical processor(s) 2330 include, without limitation, microprocessors, microcontrollers, Central Processing Units (CPUs), Field-Programmable Gate Arrays (FPGAs) that implement softcore processors, Application-Specific Integrated Circuits (ASICs), portions of one or more of the same, variations or combinations of one or more of the same, and/or any other suitable physical processor.

[0365] In some embodiments, dental consumer/patient system 2302 can include a camera 2332. Camera 2332 can comprise a camera, scanner, or other optical sensor. Camera 2332 can include one or more lenses or can, one or more camera devices, and/or one or more other optical sensors. In some examples, camera 2332 can include other sensors and/or devices which can aid in capturing optical data, such as one or more lights, depth sensors, etc. In various implementations, the camera 2332 is not a clinical scanner. [0366] Virtual dental care datastore(s) 2320 include one or more datastore configured to store any type or form of data that can be used for virtual dental care. In some embodiments, the virtual dental care datastore(s) 2320 include, without limitation, patient data 2336 and treatment data 2338. Patient data 2336 can include data collected from patients, such as patient dentition information, patient historical data, patient scans, patient information, etc. Treatment data 2338 can include data used for treating patients, such as treatment plans, state of treatment, success of treatment, changes to treatment, notes regarding treatment, etc.

[0367] As will be described in greater detail below, one or more of virtual dental care modules 2308 and/or the virtual dental care datastore(s) 2320 in FIG. 23 can, (when executed by at least one processor of dental consumer/patient system 2302, virtual dental care system 2306, and/or dental professional system 2350) enable dental consumer/patient system 2302, virtual dental care system 2306, and/or dental professional system 2350 to optimize palatal expansion treatment of an existing palatal expansion treatment plan and/or generate a new palatal expansion treatment plan.

[0368] Some embodiments provide patients with “Virtual dental care.” “Virtual dental care,” as used herein, can include computer-program instructions and/or software operative to provide remote dental services by a health professional (dentist, orthodontist, dental technician, etc.) to a patient, a potential consumer of dental services, and/or other individual. Virtual dental care can comprise computer-program instructions and/or software operative to provide dental services without a physical meeting and/or with only a limited physical meeting. As an example, virtual dental care can include software operative to providing dental care from the dental professional system 2350 and/or the virtual dental care system 2306 to the computing device 2302 over the network 2304 through e.g., written instructions, interactive applications that allow the health professional and patient/consumer to interact with one another, telephone, chat etc. Some embodiments provide patients with “Remote dental care.” “Remote dental care,” as used herein, can comprise computer-program instructions and/or software operative to provide a remote service in which a health professional provides a patient with dental health care solutions and/or services. In some embodiments, the virtual dental care facilitated by the elements of the system 2300 can include non-clinical dental services, such as dental administration services, dental training services, dental education services, etc.

[0369] In some embodiments, the elements of the system 2300 (e.g., the virtual dental care modules 2308 and/or the virtual dental care datastore(s) 2320) can be operative to provide intelligent photo guidance to a patient to take images relevant to virtual dental care using the camera 2332 on the computing device 2302.

[0370] FIG. 24 illustrates a block diagram of an example processing device 2400 operating in accordance with one or more aspects of the present disclosure. In one implementation, the processing device 2400 can be a part of any computing device of FIG. 22, or any combination thereof. Example processing device 2400 can be connected to other processing devices in a LAN, an intranet, an extranet, and/or the Internet. The processing device 2400 can be a personal computer (PC), a set-top box (STB), a server, a network router, switch or bridge, or any device capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by that device. Further, while only a single example processing device is illustrated, the term “processing device” shall also be taken to include any collection of processing devices (e.g., computers) that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methods discussed herein.

[0371] Example processing device 2400 can include a processor 2402 (e.g., a CPU), a main memory 2404 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM), etc.), a static memory 2406 (e.g., flash memory, static random access memory (SRAM), etc.), and a secondary memory (e.g., a data storage device 2418), which can communicate with each other via a bus 2430.

[0372] Processor 2402 represents one or more general-purpose processing devices such as a microprocessor, central processing unit, or the like. More particularly, processor 2402 can be a complex instruction set computing (CISC) microprocessor, reduced instruction set computing (RISC) microprocessor, very long instruction word (VLIW) microprocessor, processor implementing other instruction sets, or processors implementing a combination of instruction sets. Processor 2402 can also be one or more special-purpose processing devices such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), network processor, or the like. In accordance with one or more aspects of the present disclosure, processor 2402 can be configured to execute instructions (e.g. processing logic 2426 can implement the holistic monitor of FIG. 22).

[0373] Example processing device 2400 can further include a network interface device 2408, which can be communicatively coupled to a network 2420. Example processing device 2400 can further comprise a video display 2410 (e.g., a liquid crystal display (LCD), a touch screen, or a cathode ray tube (CRT)), an alphanumeric input device 2412 (e.g., a keyboard), an input control device 2414 (e.g., a cursor control device, a touch-screen control device, a mouse), and a signal generation device 2416 (e.g., an acoustic speaker).

[0374] Data storage device 2418 can include a computer-readable storage medium (or, more specifically, a non-transitory computer-readable storage medium) 2428 on which is stored one or more sets of executable instructions 2422. In accordance with one or more aspects of the present disclosure, executable instructions 2422 can comprise executable instructions (e.g. instructions for implementing the holistic monitor of FIG. 22). [0375] Executable instructions 2422 can also reside, completely or at least partially, within main memory 2404 and/or within processor 2402 during execution thereof by example processing device 2400, main memory 2404 and processor 2402 also constituting computer-readable storage media. Executable instructions 2422 can further be transmitted or received over a network via network interface device 2408.

[0376] While the computer-readable storage medium 2428 is shown in FIG. 24 as a single medium, the term “computer-readable storage medium” should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, and/or associated caches and servers) that store the one or more sets of operating instructions. The term “computer-readable storage medium” shall also be taken to include any medium that is capable of storing or encoding a set of instructions for execution by the machine that cause the machine to perform any one or more of the methods described herein. The term “computer-readable storage medium” shall accordingly be taken to include, but not be limited to, solid-state memories, and optical and magnetic media.

[0377] Multiple exemplary embodiments are now described.

[0378] Embodiment 1 : A method for characterizing a dental occlusion of a patient, the method comprising: receiving a first image of a patient’s dentition depicting upper and lower jaws of the patient in a bite-open arrangement and a second image of the patient’s dentition depicting the upper and lower jaws of the patient in a bite-closed arrangement; determining one or more first digital measurements of the patent’s dentition from the first image and one or more second digital measurements of the patient’s dentition from the second image; and characterizing a level of malocclusion between opposing teeth of the upper and lower jaws of the patient based at least in part on the one or more first digital measurements and the one or more second digital measurements.

[0379] Embodiment 2: The method of embodiment 1 , further comprising: determining one or more actions to be performed based on the determined characterization of the level of malocclusion between the opposing teeth of the upper and lower jaws of the patient.

[0380] Embodiment 3: The method of embodiment 1 or 2, further comprising: converting at least one of a) the one or more first digital measurements or b) the one or more second digital measurements into one or more corresponding physical measurements, wherein the one or more first digital measurements and the one or more second digital measurements comprise pixel measurements and the one or more physical measurements comprise length measurements.

[0381] Embodiment 4: The method of embodiment 3, further comprising: processing at least one of the first image or the second image to determine one or more conversion factors for converting the at least one of a) the one or more first digital measurements or b) the one or more second digital measurements to the one or more corresponding physical measurements, wherein the converting is performed using the one or more conversion factors.

[0382] Embodiment 5: The method of embodiment 3 or 4, further comprising: estimating, from at least one of the first image or the second image, an inclination angle of an image sensor relative to the patient’s dentition during capture of at least one of the first image or the second image; computing a perspective correction factor based on the estimated inclination angle; and modifying the one or more physical measurements based on the computed perspective correction factor.

[0383] Embodiment 6: The method of embodiments 1 -5, further comprising: determining a resolution of the first image; and determining a technique for characterizing the level of the dental occlusion between the opposing teeth of the upper and lower jaws based on the resolution of the first image.

[0384] Embodiment 7: The method of embodiment 6, further comprising: determining a low- resolution technique for characterizing the level of the dental occlusion between the opposing teeth of the upper and lower jaws responsive to determining that the resolution of the first image is below a resolution threshold, wherein the low-resolution technique comprises: determining an amount of one or more teeth of a dental arch that are covered by teeth of an opposing dental arch based on the one or more first digital measurements and the one or more second digital measurements; and converting the amount to a physical measurement.

[0385] Embodiment 8: The method of embodiment 6 or 7, further comprising: determining a high- resolution technique for characterizing the level of the dental occlusion between the opposing teeth of the upper and lower jaws responsive to determining that the resolution of the first image is at or above a resolution threshold, wherein the high-resolution technique comprises: converting the one or more first digital measurements into one or more first physical measurements; converting the one or more second digital measurements into one or more second physical measurements; and determining an amount of one or more teeth of a dental arch that are covered by teeth of an opposing dental arch based on the one or more first physical measurements and the one or more second physical measurements.

[0386] Embodiment 9: The method of embodiments 1 -8, wherein: the one or more first digital measurements comprise a first digital measurement of a dimension of one or more first teeth of a first dental arch in the first image and a second digital measurement of a dimension of one or more second teeth of a second dental arch in the first image; and the one or more second digital measurements comprise a third digital measurement of a dimension of the one or more first teeth in the second image and a fourth digital measurement of a dimension of the one or more second teeth in the second image, wherein a portion of the one or more second teeth is occluded in the second image. [0387] Embodiment 10: The method of embodiment 9, wherein characterizing the level of the dental occlusion between the opposing teeth of the upper and lower jaws of the patient, comprises: converting the first digital measurement into a first physical measurement, the second digital measurement into a second physical measurement, the third digital measurement into a third physical measurement, and the fourth digital measurement into a fourth physical measurement; and determining a difference between a) a sum of the first physical measurement and the second physical measurement and b) a sum of the third physical measurement and the fourth physical measurement.

[0388] Embodiment 11 : The method of embodiment 9 or 10, wherein characterizing the level of the dental occlusion between the opposing teeth of the upper and lower jaws of the patient comprises: determining a ratio of the fourth digital measurement to the third digital measurement; determining a fifth digital measurement of the portion of the one or more second teeth that is occluded in the second image based on the ratio; and converting the fifth digital measurement into a physical measurement. [0389] Embodiment 12: The method of embodiments 9-11 , wherein the one or more first teeth comprise one or more upper incisors, wherein the one or more second teeth comprise one or more lower incisors, and wherein the dimension comprises height.

[0390] Embodiment 13: The method of embodiments 1-12, further comprising: determining a bite classification based at least in part on the one or more first digital measurements and the one or more second digital measurements, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

[0391] Embodiment 14: The method of embodiments 1-13, wherein the first image and the second image are front-view images of the patient’s dentition.

[0392] Embodiment 15: The method of embodiments 1-14, wherein the first image and the second image are side-view images of the patient’s dentition.

[0393] Embodiment 16: The method of embodiments 1-15, wherein processing the first image to determine the one or more first digital measurements of the patent’s dentition and processing the second image to determine the one or more second digital measurements of the patient’s dentition comprises: segmenting the first image into a plurality of teeth; segmenting the second image into the plurality of teeth; measuring one or more dimensions of two or more teeth of the plurality of teeth in the first image; and measuring the one or more dimensions of the two or more teeth of the plurality of teeth in the second image.

[0394] Embodiment 17: The method of embodiments 1-16, wherein the first image and the second image were generated at a first time, the method further comprising: receiving a third image of upper and lower jaws of the patient in the bite-open arrangement and a fourth image of the upper and lower jaws of the patient in the bite-closed arrangement generated at a second time; processing the third image and the fourth image to determine at least one of a) one or more third digital measurements of the patent’s dentition from the third image or b) one or more fourth digital measurements of the patient’s dentition from the fourth image; characterizing an update to the level of the dental occlusion between the opposing teeth of the upper and lower jaws of the patient based on at least one of the one or more third digital measurements or the one or more fourth digital measurements; and tracking progress of an orthodontic treatment associated with the level of malocclusion between the opposing teeth of the upper and lower jaws.

[0395] Embodiment 18: The method of embodiment 17, further comprising: updating an orthodontic treatment plan based on the tracked progress of the orthodontic treatment.

[0396] Embodiment 19: A method for measuring a bite classification, the method comprising: receiving a bite-closed image of a patient’s dentition; registering one or more 3D models of the patient’s dentition to the bite-closed image; identifying a first reference point on the bite-closed image associated with one or more maxillary teeth of the patient; identifying a second reference point on the bite-closed image associated with one or more mandibular teeth of the patient; identifying, within an image space of the bite-closed image, a virtual marker corresponding to a distance between the first reference point and the second reference point; projecting the virtual marker from the image space to a 3D space of the one or more 3D models; determining a physical measurement corresponding to the projected virtual marker; and determining the bite classification for the patient based on the physical measurement.

[0397] Embodiment 20: The method of embodiment 19, further comprising: determining one or more actions to be performed based on the determined bite classification.

[0398] Embodiment 21 : The method of embodiment 19 or 20, further comprising: segmenting the bite-closed image into a plurality of teeth, wherein the first reference point and the second reference point are determined based on a result of the segmenting.

[0399] Embodiment 22: The method of embodiments 19-21 , wherein the bite-closed image was generated at a first time, the method further comprising: receiving a second image of the patient’s dentition generated at a second time; determining an updated bite classification based on the second image; and tracking progress of an orthodontic treatment associated with the bite classification.

[0400] Embodiment 23: The method of embodiment 22, further comprising: updating an orthodontic treatment plan based on the tracked progress of the orthodontic treatment.

[0401] Embodiment 24: The method of embodiments 19-23, wherein the bite-closed image comprises a lateral image of the patient’s dentition.

[0402] Embodiment 25: The method of embodiments 19-24, wherein the one or more maxillary teeth comprise a maxillary canine, and wherein identifying the first reference point comprises: projecting a facial axis of a clinical crown (FACC) line of the maxillary canine from the one or more 3D models to the bite-closed image.

[0403] Embodiment 26: The method of embodiment 25, wherein the second reference point comprises a point on a boundary line between a mandibular canine and a mandibular first premolar, and wherein the virtual marker comprises a minimum distance from a tip of the FACC line of the maxillary canine to the boundary line.

[0404] Embodiment 27: The method of embodiments 19-26, further comprising: receiving image data comprising the bite-closed image and a second image; determining that the second image of the image data fails to satisfy one or more criteria; and filtering out the second image of the image data. [0405] Embodiment 28: The method of embodiments 19-27, wherein the bite classification comprises a level of dental occlusion between the one or more maxillary teeth and the one or more mandibular teeth of the patient.

[0406] Embodiment 29: A method for measuring a bite classification, the method comprising: receiving, from an image sensor, image data comprising a bite-closed image of a patient’s upper dental arch and lower dental arch; registering a first 3D model of the patient’s upper dental arch to the bite- closed image; registering a second 3D model of the patient’s lower dental arch to the bite-closed image; identifying a first reference point on the first 3D model; identifying a second reference point on the second 3D model; determining a physical measurement corresponding to a distance between the first reference point and the second reference point; and determining the bite classification for the patient based on the physical measurement.

[0407] Embodiment 30: The method of embodiment 29, further comprising: determining one or more actions to be performed based on the determined bite classification of the patient.

[0408] Embodiment 31 : The method of embodiments 29-30, wherein the bite classification comprises a level of dental occlusion between one or more maxillary teeth and one or more mandibular teeth of the patient.

[0409] Embodiment 32: A method for measuring a bite classification, the method comprising: receiving, from an image sensor, image data comprising a bite-closed image of a patient’s upper dental arch and lower dental arch; processing the image data using a trained machine learning model, wherein the trained machine learning model outputs an estimate of a physical measurement corresponding to a distance between a first reference point on the patient’s upper dental arch and a second reference point on the patient’s lower dental arch; and determining the bite classification for the patient based on the physical measurement. [0410] Embodiment 33: The method of embodiment 32, wherein the bite classification comprises a level of dental occlusion between one or more maxillary teeth and one or more mandibular teeth of the patient.

[0411] Embodiment 34: A method for measuring an amount of posterior crossbite, the method comprising: receiving a bite-closed image of a patient’s upper dental arch and lower dental arch; segmenting the bite-closed image into a plurality of teeth; measuring a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient in the bite- closed image; determining a first ratio between the first tooth height and the second tooth height; determining a second ratio between a third tooth height of the maxillary tooth and a fourth tooth height of the mandibular tooth in one or more 3D models of the patient’s upper dental arch and lower dental arch; determining a difference between the first ratio and the second ratio; and determining the amount of posterior crossbite based on the difference.

[0412] Embodiment 35: A method for measuring an amount of posterior crossbite, the method comprising: receiving a bite-closed image and an bite-open image of a patient’s upper dental arch and lower dental arch; segmenting the bite-closed image and the bite-open image into a plurality of teeth; measuring a first tooth height of a maxillary tooth of the patient and a second tooth height of an opposing mandibular tooth of the patient in the bite-closed image; determining a first ratio between the first tooth height and the second tooth height; measuring a third tooth height of the maxillary tooth of the patient and a fourth tooth height of the opposing mandibular tooth of the patient in the bite-open image; determining a second ratio between the third tooth height and the fourth tooth height; determining a difference between the first ratio and the second ratio; and determining the amount of posterior crossbite based on the difference.

[0413] Embodiment 36: A method comprising: receiving a first image of a patient’s dentition, the first image comprising a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth; processing the first image of the patient’s dentition to generate a second representation of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the first image; performing one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth that is occluded in the first image; and outputting a result of the one or more oral diagnostics measurements. [0414] Embodiment 37: The method of embodiment 36, wherein the second representation of the one or more second teeth comprises at least one of dimensions or coordinate locations of one or more features of the one or more second teeth. [0415] Embodiment 38: The method of embodiments 36-37, wherein the second representation is a non-visual representation.

[0416] Embodiment 39: The method of embodiments 36-38, wherein: the second representation is a second visual representation of the one or more second teeth included in a second image that comprises new contours of the at least one region of the one or more second teeth that is occluded in the first image; and the one or more oral diagnostics measurements of the patient’s dentition are performed using the new contours of the at least one region of the one or more second teeth.

[0417] Embodiment 40: The method of embodiment 39, wherein performing the one or more oral diagnostics measurements comprises: comparing original contours of the one or more second teeth to the new contours of the one or more second teeth.

[0418] Embodiment 41 : The method of embodiment 39 or 40, further comprising: outputting the second image to a display, wherein the result of the one or more oral diagnostics measurements is overlaid on the second image.

[0419] Embodiment 42: The method of embodiments 39-41 , further comprising: transmitting the second image to a remote device that outputs the second image to a display of the remote device. [0420] Embodiment 43: The method of embodiments 39-42, wherein performing the one or more oral diagnostics measurements comprises: identifying a first reference point on the one or more first teeth; identifying a second reference point on one of the new contours of the one or more second teeth; and measuring a distance between the first reference point and the second reference point.

[0421] Embodiment 44: The methods of embodiment 39-43, wherein the second image lacks a representation of the one or more first teeth.

[0422] Embodiment 45: The method of embodiments 39-44, wherein a third image of the patient’s dentition is also generated based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

[0423] Embodiment 46: The method of embodiments 39-45, wherein the one or more first teeth and the one or more second teeth are on a same jaw of the patient, and wherein performing one or more oral diagnostics measurements of the patient’s dentition comprises: measuring a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth; and determining a crowding level based on the horizontal distance. [0424] Embodiment 47: The method of embodiment 46, further comprising: determining crowding levels for a plurality of pairs of adjacent teeth on the jaw of the patient; and determining an aggregate crowding level for the jaw based on the crowding levels for the plurality of pairs of adjacent teeth.

[0425] Embodiment 48: The method of embodiment 47, further comprising: predicting at least one of a length of dental treatment or a recommended product for dental treatment based on the aggregate crowding level.

[0426] Embodiment 49: The method of embodiment 47 or 48, further comprising: recommending palatal expansion treatment based on the aggregate crowding level.

[0427] Embodiment 50: The method of embodiments 39-49, further comprising: performing segmentation of the second image to generate a segmented version of second image, wherein the one or more oral diagnostics measurements of the patient’s dentition are performed from the segmented version of the second image.

[0428] Embodiment 51 : The method of embodiments 36-50, wherein performing the one or more oral diagnostics measurements comprises: determining a level of dental occlusion between the one or more first teeth and the one or more second teeth.

[0429] Embodiment 52: The method of embodiment 51 , further comprising: determining a bite classification for the patient based on the level of dental occlusion.

[0430] Embodiment 53: The method of embodiment 52, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

[0431] Embodiment 54: The method of embodiments 3-536, further comprising: segmenting the first image to generate segmentation information for the one or more first teeth and the one or more second teeth, wherein processing the first image comprises processing the segmentation information for the first image.

[0432] Embodiment 55: The method of embodiments 36-54, further comprising: determining a malocclusion based on the one or more oral diagnostics measurements, the determined malocclusion comprising at least one of an overbite level, a crowding level, or an underbite level; and outputting the determined malocclusion.

[0433] Embodiment 56: The method of embodiments 3-556, wherein the first image is a closed bite image, and wherein processing the first image comprises providing an input comprising the first image and a third image that is an open bite image into a trained artificial intelligence (Al) model, wherein the trained Al model processes the first image and the third image to output the second image of the patient’s dentition. [0434] Embodiment 57: The method of embodiments 36-56, further comprising: identifying at least one of crowding, overbite, underbite or crossbite based on the one or more oral diagnostics measurements; and recommending palatal expansion treatment based on at least one of the crowding, the overbite, the underbite, or the crossbite.

[0435] Embodiment 58: The method of embodiments 36-57, further comprising: determining an available amount of space on an upper jaw based on the one or more oral diagnostics measurements; predicting an amount of space needed for the upper jaw based at least in part on the one or more oral diagnostics measurements; and recommending palatal expansion treatment responsive to determining that the available amount of space is less than the predicted amount of space.

[0436] Embodiment 59: The method of embodiment 58, wherein predicting the amount of space needed for the upper jaw comprises: determining that the upper jaw comprises erupting or unerupted teeth; and predicting the amount of space needed for the upper jaw based at least in part on the erupting or unerupted teeth.

[0437] Embodiment 60: The method of embodiment 59, wherein predicting the amount of space needed for the upper jaw further comprises: computing a statistical distribution of tooth size and lateral space used by teeth of the patient; and predicting a tooth size and lateral space to be used by the erupting or unerupted teeth based on the statistical distribution, wherein the predicted amount of space needed for the upper jaw is determined based at least in part on the predicted tooth size and lateral space to be used by the erupting or unerupted teeth.

[0438] Embodiment 61 : The method of embodiment 60, further comprising: identifying one or more primary teeth of the patient; and determining that the upper jaw comprises the unerupted teeth based on the one more identified primary teeth of the patient.

[0439] Embodiment 62: The method of embodiments 36-61 , further comprising: comparing the result of the one or more oral diagnostics measurements to one or more predetermined values associated with a treatment plan; and determining whether to adjust the treatment plan based on a result of the comparing.

[0440] Embodiment 63: The method of embodiments 36-62, wherein the one or more first teeth are on a first jaw of the patient and the one or more second teeth are on a second jaw of the patient that opposes the first jaw, the method further comprising: characterizing a level of malocclusion between opposing teeth of the first jaw and the second jaw of the patient based at least in part on the one or more oral diagnostics measurements.

[0441] Embodiment 64: The method of embodiments 36-63, further comprising: outputting the first image to a display, wherein the result of the one or more oral diagnostics measurements is overlaid on the first image. [0442] Embodiment 65: A method comprising: receiving a first image of a patient’s dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; determining one or more three-dimensional (3D) models of the patient’s dentition; projecting the one or more 3D models onto an image plane defined by the first image to generate a second image comprising a second representation of at least the one or more second teeth, wherein the contours of the more second teeth that are occluded by the one or more first teeth in the first image are shown in the second image; and training an artificial intelligence (Al) model using the first image as an input and the second image as a target output, wherein the Al model is trained to process input images comprising representations of dentition including occluded teeth and to generate output images showing contours of the occluded teeth that are occluded in the input images.

[0443] Embodiment 66: The method of embodiment 65, further comprising: generating the first image from the one or more 3D models by projecting the one or more 3D models onto the image plane. [0444] Embodiment 67: The method of embodiment 66, further comprising: perturbing at least one tooth of the one or more first teeth and the one or more second teeth in the one or more 3D models prior to generating at least one of the first image or the second image from the one or more 3D models. [0445] Embodiment 68: The method of embodiment 66 or 67, wherein the first image and second image are two-dimensional (2D) images, the method further comprising: performing 2D to 3D registration to align the one or more first teeth and the one or more second teeth of the first image with the one or more first teeth and the one or more second teeth of the one or more 3D models; and determining camera parameters and a jaw pose based on a result of the 2D to 3D registration; wherein the camera parameters and the jaw pose are used to render the second image.

[0446] Embodiment 69: The method of embodiments 66-68, further comprising: generating the first image of the patient’s dentition based on the one or more 3D models of the patient’s dentition.

[0447] Embodiment 70: The method of embodiment 69, further comprising: perturbing at least one tooth of the one or more first teeth and the one or more second teeth in the one or more 3D models; and/or selecting at least one of a jaw pose or camera parameters for generation of the first image.

[0448] Embodiment 71 : The method of embodiments 65-69, further comprising: determining at least one of tooth locations, tooth orientations, tooth sizes or tooth geometry from the one or more 3D models; and using at least one of the tooth locations, the tooth orientations, the tooth sizes or the tooth geometry in addition to the first image as the input for training the Al model.

[0449] Embodiment 72: The method of embodiments 65-71 , further comprising: projecting the one or more 3D models onto the image plane defined by the first image to generate a third image comprising a third representation of at least the one or more second teeth, wherein the one or more first teeth are not included in the third image; wherein the Al model is trained to process the input images and generate the output images showing the contours of the occluded teeth and additional images showing only the occluded teeth.

[0450] Embodiment 73: A method comprising: receiving a first image of a patient’s dentition, the first image comprising a visual representation of one or more first teeth and a first visual representation of one or more second teeth that are at least partially occluded by the one or more first teeth; processing the first image of the patient’s dentition to generate a second representation of the one or more second teeth that includes information of at least one region of the one or more second teeth that is occluded in the first image; performing one or more oral diagnostics measurements of the patient’s dentition using the information of the at least one region of the one or more second teeth that is occluded in the first image; determining whether to recommend palatal expansion treatment based on the one or more oral diagnostics measurements of the patient’s dentition; and outputting a recommendation with respect to palatal expansion treatment.

[0451] Embodiment 74: The method of embodiment 73, wherein the second representation of the one or more second teeth comprises at least one of dimensions or coordinate locations of one or more features of the one or more second teeth.

[0452] Embodiment 75: The method of embodiments 73-74, wherein: the second representation is a second visual representation of the one or more second teeth included in a second image that comprises new contours of the at least one region of the one or more second teeth that is occluded in the first image; and the one or more oral diagnostics measurements of the patient’s dentition are performed using the new contours of the at least one region of the one or more second teeth.

[0453] Embodiment 76: The method of embodiments 73-75, wherein processing the first image comprises: identifying which teeth in the first image are primary teeth and which teeth in the first image are permanent teeth; measuring an available space in an upper jaw of the patient based on the one or more oral diagnostics measurements; and determining whether the available space is sufficient for replacement of the primary teeth on the upper jaw with permanent teeth; wherein the determining of whether to recommend palatal expansion treatment is based at least in part on whether the available space is sufficient for replacement of the primary teeth on the upper jaw with permanent teeth.

[0454] Embodiment 77: The method of embodiments 73-76, wherein the first image is processed using a trained artificial intelligence (Al) model to generate the second representation of the one or more second teeth.

[0455] Embodiment 78: The method of embodiments 73-77, wherein the one or more first teeth and the one or more second teeth are on a same jaw of the patient, and wherein performing one or more oral diagnostics measurements of the patient’s dentition comprises: measuring a horizontal distance between a first point on the one or more new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth; and determining a crowding level based on the horizontal distance.

[0456] Embodiment 79: The method of embodiment 78, further comprising: determining crowding levels for a plurality of pairs of adjacent teeth on the jaw of the patient; and determining an aggregate crowding level for the jaw based on the crowding levels for the plurality of pairs of adjacent teeth.

[0457] Embodiment 80: The method of embodiment 79, further comprising: predicting a length of palatal expansion treatment based at least in part on the aggregate crowding level.

[0458] Embodiment 81 : The method of embodiments 79-80, further comprising: recommending palatal expansion treatment based on the aggregate crowding level.

[0459] Embodiment 82: The method of embodiments 73-81 , further comprising: identifying at least one of crowding, overbite, underbite or crossbite based on the one or more oral diagnostics measurements; and recommending palatal expansion treatment based on at least one of the crowding, the overbite, the underbite, or the crossbite.

[0460] Embodiment 83: The method of embodiments 73-82, further comprising: determining an available amount of space on an upper jaw based on the one or more oral diagnostics measurements; predicting a necessary amount of space on the upper jaw; and recommending palatal expansion treatment responsive to determining that the available amount of space is less than the predicted necessary amount of space.

[0461] Embodiment 84: The method of embodiment 83, wherein predicting the necessary amount of space on the upper jaw comprises: determining that the upper jaw comprises erupting or unerupted teeth; and predicting the necessary amount of space for the upper jaw based at least in part on the erupting or unerupted teeth.

[0462] Embodiment 85: The method of embodiment 84, wherein predicting the necessary amount of space for the upper jaw further comprises: computing a statistical distribution of tooth size and lateral space used by teeth of the patient; and predicting a tooth size and lateral space to be used by the erupting or unerupted teeth based on the statistical distribution, wherein the predicted necessary amount of space is determined based at least in part on the predicted tooth size and the predicted lateral space to be used by the erupting or unerupted teeth.

[0463] Embodiment 86: The method of embodiments 84-85, further comprising: identifying one or more primary teeth of the patient; and determining that the upper jaw comprises the unerupted teeth based on the one more identified primary teeth of the patient. [0464] Embodiment 87: A non-transitory computer readable medium comprising instructions that, when executed by a processing device, cause the processing device to perform the method of any of embodiments 1-86.

[0465] Embodiment 88: A computing device comprising: a memory configured to store instructions; and a processing device configured to execute the instructions from the memory to perform the method of any of embodiments 1-86.

[0466] Embodiment 89: A system comprising: a first computing device comprising a memory and one or more processors, wherein the first computing device is configured to: receive a first image of a patient’s dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; process the first image of the patient’s dentition to generate a second image of the patient’s dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image; perform one or more oral diagnostics measurements of the patient’s dentition using the new contours of the at least one region of the one or more second teeth; and output a result of the one or more oral diagnostics measurements.

[0467] Embodiment 90: The system of embodiment 89, wherein performing the one or more oral diagnostics measurements comprises: comparing original contours of the one or more second teeth to the new contours of the one or more second teeth.

[0468] Embodiment 91 : The system of embodiments 89-90, wherein the processing of the first image of the patient’s dentition is performed by a trained artificial intelligence (Al) model that outputs the second image of the patient’s dentition.

[0469] Embodiment 92: The system of embodiments 89-91 , wherein the first computing device is further configured to: output at least one of the first image or the second image to a display, wherein the result of the one or more oral diagnostics measurements is overlaid on at least one of the first image or the second image.

[0470] Embodiment 93: The system of embodiments 89-92, further comprising: a second computing device comprising a display, the second computing device configured to: receive the one or more oral diagnostics measurements and at least one of the first image or the second image from the first computing device; and output the one or more oral diagnostics measurements and at least one of the first image or the second image to the display.

[0471] Embodiment 94: The system of embodiments 89-93, wherein performing the one or more oral diagnostics measurements comprises: determining a level of dental occlusion between the one or more first teeth and the one or more second teeth; and determining a bite classification for the patient based on the level of dental occlusion, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

[0472] Embodiment 95: The system of embodiments 89-94, wherein performing the one or more oral diagnostics measurements comprises: identifying a first reference point on the one or more first teeth; identifying a second reference point on one of the new contours of the one or more second teeth; and measuring a distance between the first reference point and the second reference point.

[0473] Embodiment 96: The system of embodiments 89-95, wherein a third image of the patient’s dentition is also generated based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

[0474] Embodiment 97: The system of embodiments 89-96, wherein the one or more first teeth and the one or more second teeth are on a same jaw of the patient, and wherein performing one or more oral diagnostics measurements of the patient’s dentition comprises: measuring a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth; and determining a crowding level based on the horizontal distance.

[0475] Embodiment 98: The system of embodiment 97, wherein the first computing device is further configured to: determine crowding levels for a plurality of pairs of adjacent teeth on the jaw of the patient; and determine an aggregate crowding level for the jaw based on the crowding levels for the plurality of pairs of adjacent teeth.

[0476] Embodiment 99: The system of embodiments 89-98, wherein the first computing device is further configured to: identify at least one of crowding, overbite, underbite or crossbite based on the one or more oral diagnostics measurements; and recommend palatal expansion treatment based on at least one of the crowding, the overbite, the underbite, or the crossbite.

[0477] Embodiment 100: The system of embodiments 89-99, wherein the first computing device is further configured to: determine an available amount of space on an upper jaw based on the one or more oral diagnostics measurements; predict an amount of space needed for the upper jaw based at least in part on the one or more oral diagnostics measurements; and recommend palatal expansion treatment responsive to determining that the available amount of space is less than the predicted amount of space. [0478] Embodiment 101 : The system of embodiment 100, wherein predicting the amount of space needed for the upper jaw comprises: determining that the upper jaw comprises erupting or unerupted teeth; predicting the amount of space needed for the upper jaw based at least in part on the erupting or unerupted teeth; computing a statistical distribution of tooth size and lateral space used by teeth of the patient; and predicting a tooth size and lateral space to be used by the erupting or unerupted teeth based on the statistical distribution, wherein the predicted amount of space needed for the upper jaw is determined based at least in part on the predicted tooth size and lateral space to be used by the erupting or unerupted teeth.

[0479] Embodiment 102: The system of embodiments 89-101 , wherein the first computing device is further configured to: compare the result of the one or more oral diagnostics measurements to one or more predetermined values associated with a treatment plan; and determine whether to adjust the treatment plan based on a result of the comparing.

[0480] Embodiment 103: The system of embodiments 89-102, further comprising: a second computing device configured to: generate the first image of the patient’s dentition; and transmit the first image to the first computing device.

[0481] Embodiment 104: A method comprising: transmitting a first image of a patient’s dentition to a remote computing device, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; and receiving, from the remote computing device: a second image of the patient’s dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image, the remote computing device having processed the first image of the patient’s dentition to generate the second image of the patient’s dentition; and a result of one or more oral diagnostics measurements of the patient’s dentition determined by the remote computing device using the new contours of the at least one region of the one or more second teeth; and outputting at least one of the first image or the second image to a display; and outputting the result of the one or more oral diagnostics measurements. [0482] Embodiment 105: The method of embodiment 104, further comprising: outputting the result of the one or more oral diagnostics measurements overlaid on at least one of the first image or the second image.

[0483] Embodiment 106: The method of embodiments 104-105, wherein the second image was generated using a trained artificial intelligence (Al) model, and wherein the one or more oral diagnostics measurements were performed by comparing original contours of the one or more second teeth to the new contours of the one or more second teeth. [0484] Embodiment 107: The method of embodiments 104-106, wherein the result of the one or more oral diagnostics measurements comprises a) a level of dental occlusion between the one or more first teeth and the one or more second teeth and b) a bite classification determined based on the level of dental occlusion, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

[0485] Embodiment 108: The method of embodiments 104-107, further comprising: receiving a third image of the patient’s dentition from the remote computing device that was also generated based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

[0486] Embodiment 109: The method of embodiments 104-108, wherein the one or more oral diagnostics measurements comprises a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth that indicates an amount of the one or more second teeth occluded by the one or more first teeth, and wherein the result of the one or more oral diagnostics measurements comprises a crowding level determined based on the horizontal distance.

[0487] Any of the methods (including user interfaces) described herein can be implemented as software, hardware or firmware, and can be described as a non-transitory machine-readable storage medium storing a set of instructions capable of being executed by a processor (e.g., computer, tablet, smartphone, etc.), that when executed by the processor causes the processor to control perform any of the steps, including but not limited to: displaying, communicating with the user, analyzing, modifying parameters (including timing, frequency, intensity, etc.), determining, alerting, or the like. For example, computer models (e.g., for additive manufacturing) and instructions related to forming a dental device can be stored on a non-transitory machine-readable storage medium.

[0488] It should be understood that the above description is intended to be illustrative, and not restrictive. Many other embodiment examples will be apparent to those of skill in the art upon reading and understanding the above description. Although the present disclosure describes specific examples, it will be recognized that the systems and methods of the present disclosure are not limited to the examples described herein, but can be practiced with modifications within the scope of the appended claims. Accordingly, the specification and drawings are to be regarded in an illustrative sense rather than a restrictive sense. The scope of the present disclosure should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

[0489] The embodiments of methods, hardware, software, firmware, or code set forth above can be implemented via instructions or code stored on a machine-accessible, machine readable, computer accessible, or computer readable medium which are executable by a processing element. “Memory” includes any mechanism that provides (i.e., stores and/or transmits) information in a form readable by a machine, such as a computer or electronic system. For example, “memory” includes random-access memory (RAM), such as static RAM (SRAM) or dynamic RAM (DRAM); ROM; magnetic or optical storage medium; flash memory devices; electrical storage devices; optical storage devices; acoustical storage devices, and any type of tangible machine-readable medium suitable for storing or transmitting electronic instructions or information in a form readable by a machine (e.g., a computer).

[0490] Reference throughout this specification to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the disclosure. Thus, the appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0491] In the foregoing specification, a detailed description has been given with reference to specific exemplary embodiments. It will, however, be evident that various modifications and changes can be made thereto without departing from the broader spirit and scope of the disclosure as set forth in the appended claims. The specification and drawings are, accordingly, to be regarded in an illustrative sense rather than a restrictive sense. Furthermore, the foregoing use of embodiment, embodiment, and/or other exemplarily language does not necessarily refer to the same embodiment or the same example, but can refer to different and distinct embodiments, as well as potentially the same embodiment.

[0492] The words “example” or “exemplary” are used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “example’ or “exemplary” is not necessarily to be construed as preferred or advantageous over other aspects or designs. Rather, use of the words “example” or “exemplary” is intended to present concepts in a concrete fashion. As used in this application, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless specified otherwise, or clear from context, “X includes A or B” is intended to mean any of the natural inclusive permutations. That is, if X includes A; X includes B; or X includes both A and B, then “X includes A or B” is satisfied under any of the foregoing instances. In addition, the articles “a” and “an” as used in this application and the appended claims should generally be construed to mean “one or more” unless specified otherwise or clear from context to be directed to a singular form. Moreover, use of the term “an embodiment” or “one embodiment” or “an embodiment” or “one embodiment” throughout is not intended to mean the same embodiment or embodiment unless described as such. Also, the terms “first,” “second,” “third,” “fourth,” etc. as used herein are meant as labels to distinguish among different elements and can not necessarily have an ordinal meaning according to their numerical designation.

[0493] A digital computer program, which can also be referred to or described as a program, software, a software application, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a digital computing environment. The essential elements of a digital computer a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and digital data. The central processing unit and the memory can be supplemented by, or incorporated in, special purpose logic circuitry or quantum simulators. Generally, a digital computer will also include, or be operatively coupled to receive digital data from or transfer digital data to, or both, one or more mass storage devices for storing digital data, e.g., magnetic, magneto-optical disks, optical disks, or systems suitable for storing information. However, a digital computer need not have such devices.

[0494] Digital computer-readable media suitable for storing digital computer program instructions and digital data include all forms of non-volatile digital memory, media, and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; CD-ROM and DVD-ROM disks.

[0495] Control of the various systems described in this specification, or portions of them, can be implemented in a digital computer program product that includes instructions that are stored on one or more non-transitory machine-readable storage media, and that are executable on one or more digital processing devices. The systems described in this specification, or portions of them, can each be implemented as an apparatus, method, or system that can include one or more digital processing devices and memory to store executable instructions to perform the operations described in this specification.

[0496] While this specification contains many specific embodiment details, these should not be construed as limitations on the scope of what can be claimed, but rather as descriptions of features that can be specific to particular embodiments. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable sub-combination. Moreover, although features can be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination can be directed to a sub-combination or variation of a subcombination.

[0497] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing can be advantageous. Moreover, the separation of various system modules and components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0498] Particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. For example, the actions recited in the claims can be performed in a different order and still achieve desirable results. As one example, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some cases, multitasking and parallel processing can be advantageous.

Claims

CLAIMS: What is claimed is:

1. A system comprising: a first computing device comprising a memory and one or more processors, wherein the first computing device is configured to perform operations comprising: access a first image of a patient’s dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; process the first image of the patient’s dentition to generate a second image of the patient’s dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image; perform one or more oral diagnostics measurements of the patient’s dentition using the new contours of the at least one region of the one or more second teeth; and output a result of the one or more oral diagnostics measurements.

2. The system of claim 1 , wherein performing the one or more oral diagnostics measurements comprises: comparing original contours of the one or more second teeth to the new contours of the one or more second teeth.

3. The system of claim 1 or claim 2, further comprising: a second computing device configured to perform operations comprising: generate the first image of the patient’s dentition; and transmit the first image to the first computing device.

4. The system of any of claims 1 -3, wherein the first computing device is configured to perform further operations comprising: output at least one of the first image or the second image to a display, wherein the result of the one or more oral diagnostics measurements is overlaid on at least one of the first image or the second image.

5. The system of any of claims 1 -4, further comprising: a second computing device comprising a display, the second computing device configured to perform operations comprising: receive the one or more oral diagnostics measurements and at least one of the first image or the second image from the first computing device; and output the one or more oral diagnostics measurements and at least one of the first image or the second image to the display.

6. The system of any of claims 1 -5, wherein performing the one or more oral diagnostics measurements comprises: determining a level of dental occlusion between the one or more first teeth and the one or more second teeth; and determining a bite classification for the patient based on the level of dental occlusion, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

7. The system of any of claims 1 -6, wherein performing the one or more oral diagnostics measurements comprises: identifying a first reference point on the one or more first teeth; identifying a second reference point on one of the new contours of the one or more second teeth; and measuring a distance between the first reference point and the second reference point.

8. The system of any of claims 1 -7, wherein a third image of the patient’s dentition is also generated based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

9. The system of any of claims 1 -8, wherein the one or more first teeth and the one or more second teeth are on a same jaw of the patient, and wherein performing one or more oral diagnostics measurements of the patient’s dentition comprises: measuring a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth to determine an amount of the one or more second teeth occluded by the one or more first teeth; and determining a crowding level based on the horizontal distance.

10. The system of claim 9, wherein the first computing device is configured to perform further operations comprising: determine crowding levels for a plurality of pairs of adjacent teeth on the jaw of the patient; and determine an aggregate crowding level for the jaw based on the crowding levels for the plurality of pairs of adjacent teeth.

11 . The system of any of claims 1-10, wherein the first computing device is configured to perform further operations comprising: identify at least one of crowding, overbite, underbite or crossbite based on the one or more oral diagnostics measurements; and recommend palatal expansion treatment based on at least one of the crowding, the overbite, the underbite, or the crossbite.

12. The system of any of claims 1-11 , wherein the first computing device is configured to perform further operations comprising: determine an available amount of space on an upper jaw based on the one or more oral diagnostics measurements; predict an amount of space needed for the upper jaw based at least in part on the one or more oral diagnostics measurements; and recommend palatal expansion treatment responsive to determining that the available amount of space is less than the predicted amount of space.

13. The system of claim 12, wherein predicting the amount of space needed for the upper jaw comprises: determining that the upper jaw comprises erupting or unerupted teeth; predicting the amount of space needed for the upper jaw based at least in part on the erupting or unerupted teeth; computing a statistical distribution of tooth size and lateral space used by teeth of the patient; and predicting a tooth size and lateral space to be used by the erupting or unerupted teeth based on the statistical distribution, wherein the predicted amount of space needed for the upper jaw is determined based at least in part on the predicted tooth size and lateral space to be used by the erupting or unerupted teeth.

14. The system of any of claims 1-13, wherein the first computing device is configured to perform further operations comprising: compare the result of the one or more oral diagnostics measurements to one or more predetermined values associated with a treatment plan; and determine whether to adjust the treatment plan based on a result of the comparing.

15. A method of oral diagnostics measurements, performed by the system of any of claims 1-14, the method comprising: performing the operations that the system of any of claims 1-14 is configured to perform.

16. A method comprising: transmitting a first image of a patient’s dentition to a remote computing device, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; and receiving, from the remote computing device: a second image of the patient’s dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image, the remote computing device having processed the first image of the patient’s dentition to generate the second image of the patient’s dentition; and a result of one or more oral diagnostics measurements of the patient’s dentition determined by the remote computing device using the new contours of the at least one region of the one or more second teeth; and outputting at least one of the first image or the second image to a display; and outputting the result of the one or more oral diagnostics measurements.

17. The method of claim 16, further comprising: outputting the result of the one or more oral diagnostics measurements overlaid on at least one of the first image or the second image.

-I l l-

18. The method of claim 16 or claim 17, wherein the second image was generated using a trained artificial intelligence (Al) model, and wherein the one or more oral diagnostics measurements were performed by comparing original contours of the one or more second teeth to the new contours of the one or more second teeth.

19. The method of any of claims 16-18, wherein the result of the one or more oral diagnostics measurements comprises a) a level of dental occlusion between the one or more first teeth and the one or more second teeth and b) a bite classification determined based on the level of dental occlusion, wherein the bite classification comprises one of a deep bite classification, an underbite classification, an anterior crossbite classification, a single-tooth crossbite classification, or a posterior crossbite classification.

20. The method of any of claims 16-19, further comprising: receiving a third image of the patient’s dentition from the remote computing device that was also generated based on the processing of the first image, wherein the third image of the patient’s dentition comprises the representation of the one or more first teeth and a third representation of the one or more second teeth, wherein the new contours of the one or more second teeth are shown in the third representation using a different visualization than original contours of the one or more second teeth that are also shown in the first representation.

21 . The method of any of claims 16-20, wherein the one or more oral diagnostics measurements comprises a horizontal distance between a first point on the new contours of the one or more second teeth and a second point on a contour of the one or more first teeth that indicates an amount of the one or more second teeth occluded by the one or more first teeth, and wherein the result of the one or more oral diagnostics measurements comprises a crowding level determined based on the horizontal distance.

22. A computing device comprising: a memory; and one or more processors configured to execute instructions from the memory to perform the method of any of claims 16-21.

23. A non-transitory computer readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving a first image of a patient’s dentition, the first image comprising a representation of one or more first teeth and a first representation of one or more second teeth that are at least partially occluded by the one or more first teeth; processing the first image of the patient’s dentition to generate a second image of the patient’s dentition, the second image comprising a second representation of the one or more second teeth that includes new contours of at least one region of the one or more second teeth that is occluded in the first image; performing one or more oral diagnostics measurements of the patient’s dentition using the new contours of the at least one region of the one or more second teeth; and outputting a result of the one or more oral diagnostics measurements.