WO2025034749A2

WO2025034749A2 - Automated musculoskeletal disorder diagnosis and therapy recommendation

Info

Publication number: WO2025034749A2
Application number: PCT/US2024/041112
Authority: WO
Inventors: Shlomo MANDEL
Original assignee: Individual
Current assignee: Individual
Priority date: 2023-08-07
Filing date: 2024-08-06
Publication date: 2025-02-13
Anticipated expiration: 2026-02-07
Also published as: WO2025034749A3

Abstract

A method of facilitating patient therapy includes obtaining, via a computing device, symptom information from the patient indicating symptoms experienced by the patient; obtaining, via at least one camera, motion image data indicating at least one characteristic of movement of the patient; obtaining, via the computing device, anatomy image data indicating a condition of at least one portion of the patient's anatomy; processing, using the computing device, the symptom information, the motion image data, and the anatomy image data using a machine learning component to identify a musculoskeletal disorder affecting the patient; determining, using the computing device, a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder; and providing, via a communication module, the self-administrable therapy regimen to the patient.

Description

AUTOMATED MUSCULOSKELETAL DISORDER

DIAGNOSIS AND THERAPY RECOMMENDATION

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application Serial No. 63/531,039, filed August 7, 2023, the entire disclosure of which is incorporated herein by reference.

BACKGROUND

[0002] A significant percentage of people, which in the US is estimated at about 50%, will report musculoskeletal (MSK) symptoms. The incidence of MSK disorders increases with age, and they are a major cause of disability, loss of independence and chronic pain. The postpandemic increased sedentary lifestyle and psychological stress levels are accelerating the rate of MSK complaints.

[0003] MSK disorders have detrimental economic effects in addition to the pain they cause. The lost productivity and wages associated with MSK disorders is estimated to be at least one trillion dollars in the United States alone. MSK treatment has been the leading category of paid insurance claims. The introduction of new chemotherapy drugs recently made cancer care the leader in that regard but MSK disorder treatment remains a significant expense.

[0004] The MSK treatment journey is typically triggered by the patient’s subjective symptoms. Patients usually do not seek treatment for MSK symptoms until the pain is intolerable or their experienced loss of function or range of motion is unacceptable. Diagnosing MSK disorder typically includes interviewing a patient to obtain a symptom inventory and radiographic imaging. Treatment options for MSK disorders are well-established but limited.

Correctly matching treatment(s) to the presumptive diagnosis is challenging. Success is generally measured by decreased pain and improved function. SUMMARY

[0005] An illustrative example embodiment of a method of facilitating patient treatment includes obtaining symptom information from the patient indicating symptoms experienced by the patient; obtaining motion image data indicating at least one characteristic of movement of the patient; obtaining anatomy image data indicating a condition of at least one portion of the patient's anatomy; processing the symptom information, the motion image data, and the anatomy image data using a machine learning algorithm to identify a musculoskeletal disorder affecting the patient; determining a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder; and providing the self-administrable therapy regimen to the patient.

[0006] The various features and advantages of at least one disclosed example embodiment will become apparent to those skilled in the art from the following detailed description. The drawings that accompany the detailed description can be briefly described as follows.

BRIEF DESCRIPTION OF THE DRAWINGS

[0007] This disclosure is best understood from the following detailed description when read in conjunction with the accompanying drawings. It is emphasized that, according to common practice, the various features of the drawings are not to-scale. On the contrary, the dimensions of the various features are arbitrarily expanded or reduced for clarity.

[0008] Figure 1 is a block diagram illustrating an example of a system for musculoskeletal (MSK) disorder diagnosis and mitigation, in accordance with the present disclosure.

[0009] Figure 2 is a block diagram of an example internal configuration of a computing device of a computing device, in accordance with the present disclosure.

[0010] Figure 3 is a schematic block diagram illustrating selected portions of an example system for automating diagnosis of a physical condition, such as MSK disorder, and facilitating self-administered therapy for the diagnosed condition.

[0011] Figure 4 is a schematic block diagram illustrating an example process for MSK diagnosis, in accordance with the present disclosure.

[0012] Figure 5 is a flow chart that illustrates an example method, in accordance with the present disclosure. DETAILED DESCRIPTION

[0013] Musculoskeletal (MSK) symptoms present for a variety of reasons. Among them arc direct trauma, degeneration, malignancy, infection and non MSK disorders that mimic MSK symptoms. In addition to nociceptive mediated pain, influence from the central nervous system can play an important role. Unlike certain metabolic disorders, there is no diagnostic ‘ground truth,’ or recognized MSK biomarkers.

[0014] Generic treatment recommendations without diagnostic screening are typically not effective, and at times even dangerous. For example, an older person who is gardening might notice pain after lifting a bag of mulch. If the pain in this instance represents simple muscle strain, then physical therapy may help. But if the lifting resulted in a vertebral body fracture, physical therapy exercise can be harmful. Traditionally MSK symptoms are evaluated through a detailed history, physical examination and relevant imaging. Therapeutic digital and virtual MSK interventions, primarily physical therapy, are rapidly gaining popularity. However, while they reduce the financial burden by limiting in - person therapy visits, the process depends on a physician's diagnosis and referral. While the diagnosis of LBP symptoms can be complex, known machine learning symptom checkers are widely available, but are generally not provided for the diagnosis of spinal disorders or other MSK disorders.

[0015] The subjective nature of MSK evidence presents a unique challenge. While certain laboratory or imaging biologic findings are objective, and independent of reporting by the source, MSK features often are a function of subjective sensation. For example, patient-reported MSK signs may be determined by both physiological disorder and the sensation of pain. From an anatomic perspective, MSK diagnoses are often based on magnetic resonance imaging (MRI) findings. However, there is often a discord between MSK symptoms and MRI-based anatomic findings. Additionally, access to clinicians specializing in MSK disorders may be limited by factors such as cost, lack of transportation, and/or insufficient supply.

[0016] MSK evidence values can vary based on poorly defined conditions such as barometric weather, sleep deprivation, dietary intake, underlying disease and more. Demographic and historical data is self-reported. The information is scored and transformed based on published statistical analysis. So called ‘symptom checkers’ are well established screening, triage or diagnostic tools. However, their ability to correctly diagnose the symptomatic source is limited by the absence of physical exam, lab and imaging data. [0017] To address at least some of these and potentially other challenges, some implementations disclosed herein utilize machine learning to facilitate diagnosis of MSK disorders and may be used to provide therapeutic suggestions based, or contingent on, a clinical diagnosis. For example, some implementations may provide, based on a diagnosis, an anatomic and/or subjective pain-based mitigation plan. Some implementations utilize a smartphone to capture video motion data, which may then be analyzed by a machine learning component to identify an MSK disorder.

[0001] A machine learning (ML) component refers to software capable of performing machine learning. For example, an ML component may be or include an ML algorithm. An ML component may be implemented on any number of different hardware devices and may include one or more machine learning models. ML is a field of study that gives computers the ability to perform certain tasks without being explicitly programmed to perform those tasks. In traditional computing, a programmer would encode instructions (e.g., to solve a quadratic equation using the quadratic formula), and the computer would perform those exact instructions. In contrast, in ML, a computer can be provided with examples and be trained to perform a task such as prediction or classification, without the programmer encoding explicit instructions for the task. ML explores the study and construction of algorithms, also referred to herein as tools, models, and/or components, which may learn from existing data and make predictions about new data. Such ML tools operate by building a model from example training data in order to make data- driven predictions or decisions expressed as outputs or assessments. Although example embodiments are presented with respect to a few ML tools, the principles presented herein may be applied to other ML tools. In some example embodiments, different ML tools may be used. ML tools may include, for example, Logistic Regression (LR), Naive-Bayes, Random Forest (RF), neural networks (NN), matrix factorization, and/or Support Vector Machines (SVM) tools. [0002] Two common types of problems in ML are classification problems and regression problems. Classification problems, also referred to as categorization problems, aim at classifying items into one of several category values (for example, is this object an apple or an orange).

Regression algorithms aim at quantifying some items (for example, by providing a value that is a real number). The ML algorithms utilize the training data to find correlations among identified features that affect the outcome. The ML algorithms utilize features for analyzing the data to generate assessments. A feature is an individual measurable property of a phenomenon being observed. The concept of a feature is related to that of an explanatory variable used in statistical techniques such as linear regression. Choosing informative, discriminating, and independent features is important for effective operation of the MLP in pattern recognition, classification, and regression. Features may be of different types, such as numeric features, strings, and graphs. [0003] ML components utilize the training data to find correlations among the identified features that affect the outcome or assessment. In some example embodiments, the training data includes labeled data, which is known data for one or more identified features and one or more outcomes. With the training data and the identified features, the ML tool may be trained. The ML tool appraises the value of the features as they correlate to the training data. The result of the training is the trained ML component. When the ML component is used to perform an assessment, new data is provided as an input to the trained ML component, and the ML component generates an assessment as output.

[0004] ML techniques train models to accurately make predictions on data fed into the models (e.g., what was said by a user in a given utterance; whether a noun is a person, place, or thing; what the weather will be like tomorrow). During a learning phase, the models are developed against a training dataset of inputs to optimize the models to correctly predict the output for a given input. Generally, the learning phase may be supervised, semi-supervised, or unsupervised; indicating a decreasing level to which the “correct” outputs are provided in correspondence to the training inputs. In a supervised learning phase, all of the outputs are provided to the model and the model is directed to develop a general rule or algorithm that maps the input to the output. In contrast, in an unsupervised learning phase, the desired output is not provided for the inputs so that the model may develop its own rules to discover relationships within the training dataset. In a semi-supervised learning phase, an incompletely labeled training set is provided, with some of the outputs known and some unknown for the training dataset.

[0005] Models may be run against a training dataset for several epochs (e.g., iterations), in which the training dataset is repeatedly fed into the model to refine its results. For example, in a supervised learning phase, a model is developed to predict the output for a given set of inputs, and is evaluated over several epochs to more reliably provide the output that is specified as corresponding to the given input for the greatest number of inputs for the training dataset. In another example, for an unsupervised learning phase, a model is developed to cluster the dataset into n groups, and is evaluated over several epochs as to how consistently it places a given input into a given group and how reliably it produces the n desired clusters across each epoch.

[0006] Once an epoch is run, the models are evaluated and the values of their variables are adjusted to attempt to better refine the model in an iterative fashion. In various aspects, the evaluations are biased against false negatives, biased against false positives, or evenly biased with respect to the overall accuracy of the model. The values may be adjusted in several ways depending on the ML technique used. For example, in a genetic or evolutionary algorithm, the values for the models that are most successful in predicting the desired outputs are used to develop values for models to use during the subsequent epoch, which may include random variation/mutation to provide additional data points. One of ordinary skill in the art will be familiar with several other ML algorithms that may be applied with the present disclosure, including linear regression, random forests, decision tree learning, neural networks, deep neural networks, etc.

[0007] Each model develops a rule or algorithm over several epochs by varying the values of one or more variables affecting the inputs to more closely map to a desired result, but as the training dataset may be varied, and is preferably very large, perfect accuracy and precision may not be achievable. A number of epochs that make up a learning phase, therefore, may be set as a given number of trials or a fixed time/computing budget, or may be terminated before that number/budget is reached when the accuracy of a given model is high enough or low enough or an accuracy plateau has been reached. For example, if the training phase is designed to run n epochs and produce a model with at least 95% accuracy, and such a model is produced before the n^th epoch, the learning phase may end early and use the produced model satisfying the end-goal accuracy threshold. Similarly, if a given model is inaccurate enough to satisfy a random chance threshold (e.g., the model is only 55% accurate in determining true/false outputs for given inputs), the learning phase for that model may be terminated early, although other models in the learning phase may continue training. Similarly, when a given model continues to provide similar’ accuracy or vacillate in its results across multiple epochs - having reached a performance plateau - the learning phase for the given model may terminate before the epoch number/computing budget is reached.

[0008] Once the learning phase is complete, the models are finalized. In some example embodiments, models that are finalized are evaluated against testing criteria. In a first example, a testing dataset that includes known outputs for its inputs is fed into the finalized models to determine an accuracy of the model in handling data that it has not been trained on. In a second example, a false positive rate or false negative rate may be used to evaluate the models after finalization. In a third example, a delineation between data clusterings is used to select a model that produces the clearest bounds for its clusters of data.

[0018] For example, some implementations may include a supervised ML model trained on data from reputable published articles and clinical guidance by MSK medical specialists. The ML model may be configured to rank features based on their predictive value power, allowing for a more accurate diagnosis. Reported and/or motion-captured data may be converted to an ML feature and provided to the ML component. The data may be converted to an ML feature via a scoring scheme that may be used by the ML component. For example, motion image data may be obtained using a camera of a patient device. The patient device may include a computing device such as a smartphone, a laptop, a desktop computer, a computing workstation, and/or a tablet, among other examples. The motion image data may depict movement of a body part of the patient. One or more characteristics of the movement may be scored based on the scoring scheme to generate the ML feature. The scoring scheme may include mappings of characteristics of motion of one or more body pails of the patient to values. In some implementations, the scoring scheme also may map scores to any number of other features that may be taken, as input, by the ML component. The features may include, for example, subjective pain levels, physiological observations (e.g., via motion image data), anatomical image characteristics, patient health history, patient activity history, patient demographic and historical data, patient vital signs, height, weight, age, and/or answers to questions presented in a survey, among other examples. [0019] In an example, motion image data may show a straight leg raise test in which a patient raises his or her leg straight up from a first position. The observed movement, which may be extracted and analyzed by the ML component, may be classified, in accordance with a scoring scheme, as positive, equivocal, or negative. The classification may be based on the patient's pain response and/or range of motion. The classification may be further refined based on any number of additional features such as those described above.

[0020] Aspects of the present disclosure may democratize MSK symptom management without compromising quality, safety, security or functional outcomes. Implementations include a patient-centric business-to-consumer (e.g., patient, user, etc.) device. Implementations may facilitate convenient, fast, accessible, easy to use and affordable MSK symptom management options. Some implementations also include a feature for identifying "red flags," which arc conditions that require urgent medical attention, such as vertebral body fracture, active infection, or malignancy

[0021] In some implementations, collecting a medical history virtually from a patient may be fairly straightforward. For example, based on the chief complaint of the patient, a series of questions may be generated to identify the source of the symptoms. Smart chatbots can ask follow-up questions to further narrow the differential diagnoses.

[0022] Implementations described herein address a limitation in the diagnosis of MSK disorders by replacing traditional hands-on physical examinations with a technology-based approach utilizing cell phone video motion capture. The recorded digital images, coupled with patient-reported pain levels, closely mimic the observations made by clinicians during a physical examination. Notably, the digital images obtained through this method provide certain advantages over human visual assessment, particularly in terms of capturing detailed angles and positions of anatomical features. These recorded images or video sequences can be analyzed to extract specific features, such as motion velocity, range of motion, and angular displacement, among other parameters. The extracted features can subsequently be inputted into a machine learning model, which may then generate outputs including, but not limited to, one or more diagnoses, one or more treatment plans, and/or predictive analytics related to the patient’s condition.

[0023] The ability of the techniques described herein to measure motion velocity may facilitate accurate identification of underlying MSK disorders. Through a unique grading and indexing methodology, which incorporates variables derived from testing procedures (including pain response), the techniques may facilitate conversion of traditional analog assessments into precise digital data.

[0024] While the motion capture technology may not replicate certain aspects of a hands-on physical examination — such as palpation — it may effectively capture functional movements, such as those observed during a one-leg (stork) test. When used in conjunction with other assessments, such as the standing Romberg pose, the motion capture technology may be used for evaluation of sensation, proprioception, and/or coordination, thereby providing a comprehensive assessment of MSK function. [0025] In some implementations, the techniques and/or devices described herein may serve as a diagnostic aid for clinical assessment of MSK symptoms, facilitating the safe and effective direction of treatment strategies. The techniques and/or devices may provide users with rapid triage information concerning MSK symptoms, aiding in the prioritization of medical intervention. The techniques and/or devices may enable ongoing monitoring of symptomatic progression by updating the survey and motion capture processes over time. The techniques and/or devices may be capable of delivering therapeutic exercises and other treatment modalities directly to the patient. In some implementations, the techniques and/or devices may be utilized by clinical workers to triage patients, thereby routing the patients to an appropriate healthcare practitioner, department, and/or therapy. Some implementations include functionality to screen for and identify 'red flag' conditions that require immediate medical attention. Some implementations may be configured to measure and manage limitations in function or disability, providing actionable data for patient management.

[0026] Figure 1 is a block diagram illustrating an example of a system 100 for MSK disorder diagnosis and mitigation, in accordance with the present disclosure. The system 100 integrates various components to facilitate MSK analysis and diagnosis. The system 100 includes an MSK support platform 102, which incorporates a machine learning (ML) component 104. The ML component 104 utilizes machine learning algorithms to analyze data and derive diagnostic insights. The patient device 106 interfaces with the MSK support platform 102 via network 110. This network 110 enables the seamless transfer of data between the patient device 106 and the MSK support platform 102, ensuring real-time access to diagnostic information and recommendations .

[0027] The MSK support platform 102 may be implemented as a server. The MSK support platform 102 may be a distributed system of multiple computing devices. The MSK support platform 102 may be a cloud-based system. The MSK support platform 102 may be a combination of hardware and software components. The MSK support platform is configured to implement the ML component 104. The ML component 104 is an aspect of the MSK support platform 102 that utilizes machine learning algorithms to analyze data and derive diagnostic insights. The ML component 104 include any number of machine learning algorithms, such as neural networks, decision trees, clustering algorithms, and the like. The ML component 104 may be trained on a database of MSK conditions and their associated symptoms, motion patterns, and anatomical features. The ML component 104 analyzes the collected data to identify the most likely MSK disorder affecting the patient.

[0028] The patient device 106 captures motion image data using a camera, for example. The patient device 106 may be any computing device with a camera, such as a smartphone, tablet, laptop, or other computing device with a camera. The patient device 106 may be used by a patient to capture motion image data and anatomical image data. The patient device 106 may be used by a patient to capture motion image data and anatomical image data in response to a prompt or request from the MSK support platform 102. The motion image data may be captured by the patient to capture a motion of a body part of the patient.

[0029] In some implementations, the motion image data may be analyzed by the patient device 106. In some implementations, the motion image data may be analyzed by the MSK support platform 102. In some implementations, the motion image data may be analyzed by a combination of the patient device 106 and the MSK support platform 102. The motion image data may be analyzed by the ML component 104. In some implementations, the ML component 104 may be included in the patient device 106. In some implementations, the ML component 104 may be included in the MSK support platform 102. In some implementations, the ML component 104 may be a component of the patient device 106 and/or the MSK support platform 102.

[0030] The ML component 104 may be trained on a database of MSK conditions and their associated symptoms, motion patterns, and anatomical features. The ML component 104 analyzes the collected data to identify the most likely MSK disorder affecting the patient. The ML component 104 may output a diagnosis, a treatment plan, and/or a prediction associated with the diagnosis and/or treatment plan. The ML component 104 may take, as input, motion image data captured by the patient device 106, symptom information obtained from the patient (e.g., via a survey), and/or other information that may be used by the ML component 104 to identify the most likely MSK disorder affecting the patient.

[0031] Additionally, the information source 108 may provide relevant data into the MSK support platform 102, contributing to the analytical processes carried out by the ML component 104. This information source 108 may include historical medical data, diagnostic criteria, and other relevant datasets essential for the system's diagnostic accuracy and comprehensiveness. For example, the information source may include a database of MSK conditions and their associated symptoms, motion patterns, and anatomical features. The information source 108 may also include imaging data, such as MRI, CT, ultrasound, or other imaging data. The imaging data may be from a database of imaging data, such as a hospital's imaging records. Training data, for supervised or unsupervised learning, may be obtained from the information source 108. Anatomical image data may be obtained from the information source 108.

[0032] In some implementations, for example, the ML component 104 may include one or more neural networks. The ML component 104 may be trained on a database of MSK conditions and their associated symptoms, motion patterns, and anatomical features. For example, the ML component 104 may be trained using motion image data associated with any number of different MSK conditions and their associated symptoms, motion patterns, and anatomical features. The ML component 104 may be configured to extract features from the motion image data such as, for example, motion velocity, motion range, patterns of motion, and/or angle of motion. The ML component 104 may be trained to identify, based on the extracted features, one or more diagnoses, one or more treatment plans, and/or one or more predictions associated therewith.

[0033] The network 110 may 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, a Wi-Fi network, or wireless LAN (WLAN)), a cellular network (e.g., a Long Term Evolution (LTE) network), routers, hubs, switches, server computers, and/or a combination thereof.

[0034] The MSK support platform 102 may be in communication with the patient device 106 via the network 110. In some implementations, the patient device 106 may represent a plurality of patient devices. Similarly, the ML component 104 may represent a plurality of ML components and/or the information source 108 may represent a plurality of information sources.

[0035] Figure 2 is a block diagram of an example internal configuration of a computing device 200 of a computing device, in accordance with the present disclosure. In some implementations, the computing device 200 may implement one or more aspects of the system 100 shown in Figure 1 such as, for example, the MSK support platform 102, the ML component 104, the patient device 106, and/or the information source 108. In some implementations, the computing device 200 may implement one or more aspects of the system 300 shown in Figure 3, such as, for example, the server 302, the computing device 304, the memory 306, the communication module 308, the patient device 310, the patient device 312, and/or the patient imaging database 318. [0036] The computing device 200 includes components or units, such as a processor 202, a memory 204, a bus 206, a power source 208, peripherals 210, a user interface 212, a network interface 214, other suitable components, or a combination thereof. One or more of the memory 204, the power source 208, the peripherals 210, the user interface 212, or the network interface 214 can communicate with the processor 202 via the bus 206.

[0037] The processor 202 is a central processing unit, such as a microprocessor, and can include single or multiple processors having single or multiple processing cores. Alternatively, the processor 202 can include another type of device, or multiple devices, configured for manipulating or processing information. For example, the processor 202 can include multiple processors interconnected in one or more manners, including hardwired or networked. The operations of the processor 202 can be distributed across multiple devices or units that can be coupled directly or across a local area or other suitable type of network. The processor 202 can include a cache, or cache memory, for local storage of operating data or instructions.

[0038] The memory 204 includes one or more memory components, which may each be volatile memory or non-volatile memory. For example, the volatile memory can be random access memory (RAM) (e.g., a DRAM module, such as DDR SDRAM). In another example, the non-volatile memory of the memory 204 can be a disk drive, a solid state drive, flash memory, or phase-change memory. In some implementations, the memory 204 can be distributed across multiple devices. For example, the memory 204 can include network-based memory or memory in multiple clients or servers performing the operations of those multiple devices.

[0039] The memory 204 can include data for immediate access by the processor 202. For example, the memory 204 can include executable instructions 216, application data 218, and an operating system 220. The executable instructions 216 can include one or more application programs, which can be loaded or copied, in whole or in part, from non-volatile memory to volatile memory to be executed by the processor 202. For example, the executable instructions 216 can include instructions for performing some or all of the techniques of this disclosure. The application data 218 can include user data, database data (e.g., database catalogs or dictionaries), or the like. In some implementations, the application data 218 can include functional programs, such as a web browser, a web server, a database server, another program, or a combination thereof. The operating system 220 can be, for example, Microsoft Windows®, Mac OS X®, or Linux®; an operating system for a mobile device, such as a smartphone or tablet device; or an operating system for a non-mobile device, such as a mainframe computer.

[0040] The power source 208 provides power to the computing device 200. For example, the power source 208 can be an interface to an external power distribution system. In another example, the power source 208 can be a battery, such as where the computing device 200 is a mobile device or is otherwise configured to operate independently of an external power distribution system. In some implementations, the computing device 200 may include or otherwise use multiple power sources. In some such implementations, the power source 208 can be a backup battery.

[0041] The peripherals 210 includes one or more sensors, detectors, or other devices configured for monitoring the computing device 200 or the environment around the computing device 200. For example, the peripherals 210 can include a geolocation component, such as a global positioning system location unit. In another example, the peripherals can include a temperature sensor for measuring temperatures of components of the computing device 200, such as the processor 202. In some implementations, the computing device 200 can omit the peripherals 210.

[0042] The user interface 212 includes one or more input interfaces and/or output interfaces. An input interface may, for example, be a positional input device, such as a mouse, touchpad, touchscreen, or the like; a keyboard; or another suitable human or machine interface device. An output interface may, for example, be a display, such as a liquid crystal display, a cathode-ray tube, a light emitting diode display, or other suitable display.

[0043] The network interface 214 provides a connection or link to a network (e.g., the network 114 shown in FIG. 1). The network interface 214 can be a wired network interface or a wireless network interface. The computing device 200 can communicate with other devices via the network interface 214 using one or more network protocols, such as using Ethernet, transmission control protocol (TCP), internet protocol (IP), power line communication, an IEEE 802.X protocol (e.g., Wi-Fi, Bluetooth, or ZigBee), infrared, visible light, general packet radio service (GPRS), global system for mobile communications (GSM), code-division multiple access (CDMA), Z-Wave, another protocol, or a combination thereof.

[0044] Figure 3 is a schematic block diagram illustrating selected portions of an example system 300 for automating diagnosis of a physical condition, such as musculoskeletal (MSK) disorder, and facilitating self- administered therapy for the diagnosed condition. The system 300 includes a server 302 that comprises at least one computing device 304, memory 306 including a machine learning database, and a communication module 308.

[0045] The computing device 304 includes at least one processor that comprises hardware and software configured to run at least one machine learning algorithm stored in the memory 306 of the server 302. The processor may be a central processing unit, such as a microprocessor, and can include single or multiple processors having single or multiple processing cores. Alternatively, the processor may include another type of device, or multiple devices, configured for manipulating or processing information. For example, the processor may include multiple processors interconnected in one or more manners, including hardwired or networked. The operations of the processor may be distributed across multiple devices or units that can be coupled directly or across a local area or other suitable type of network. The processor may include a cache, or cache memory, for local storage of operating data or instructions.

[0046] The memory 306 may include one or more memory components, which may each be volatile memory or non-volatile memory. For example, the volatile memory may be random access memory (RAM) (e.g., a DRAM module, such as DDR SDRAM). In another example, the non-volatile memory of the memory may be a disk drive, a solid state drive, flash memory, or phase-change memory. In some implementations, the memory 306 may be distributed across multiple devices. For example, the memory 306 may include network-based memory or memory in multiple clients or servers performing the operations of those multiple devices. The memory may include a non-transitory computer-readable medium storing instructions executable by the one or more processors.

[0047] The memory 306 may include data for immediate access by the processor. The executable instructions may include one or more application programs, which may be loaded or copied, in whole or in part, from non-volatile memory to volatile memory to be executed by the processor. For example, the executable instructions may include instructions for performing some or all of the techniques of this disclosure.

[0048] The memory 306 also includes a machine learning database that includes information or data that allows the computing device 304 to run the machine learning algorithm to diagnose an individual patient's condition. The machine learning database includes, for example, example data indicating combinations of characteristics related to a variety of known conditions that an individual using the system 300 may have. The computing device 304 is configured to generate a recommended therapy regimen based on the condition diagnosed through the machine learning algorithm. The communication module 308 is configured so the server 302 can communicate with remotely located devices. Figure 1 shows two example user or patient devices 310 and 312. The patient device 310 is a portable or laptop computer that includes a camera 314 (or another image capture device). The patient device 312 is a smartphone that includes a camera 316.

[0049] The server 302 also communicates with remotely located databases, such as a patient imaging database 318, to obtain information regarding a patient. For example, the patient imaging database 318 may be maintained by a hospital or healthcare provider and includes imaging studies or records, such as x-rays, computerized tomography scans, and magnetic resonance imaging scans.

[0050] Figure 4 is a schematic block diagram illustrating an example process 400 for MSK diagnosis, in accordance with the present disclosure. As shown in Figure 4, an ML component 402 may take, as input, symptom information 404, patient information 406, and motion image data 408. The symptom information 404 may include information regarding the patient's symptoms, such as pain, stiffness, and/or other symptoms. The patient information 406 may include information regarding the patient, such as the patient's height, weight, age, medical history, and/or other information. The motion image data 408 may include motion image data captured by the patient device 106.

[0051] The ML component 402 may be trained on a database of MSK conditions and their associated symptoms, motion patterns, and anatomical features. The ML component 402 may be trained using motion image data associated with any number of different MSK conditions and their associated symptoms, motion patterns, and anatomical features.

[0052] The ML component 402 may be configured to extract features from the motion image data such as, for example, motion velocity, motion range, patterns of motion, and/or angle of motion. The ML component 402 may be trained to identify, based on the extracted features, one or more diagnoses, one or more treatment plans, and/or one or more predictions associated therewith.

[0053] The ML component 402 may provide, as output, a diagnosis indication (shown as “DX”) 410. The diagnosis indication 410 may include a diagnosis of an MSK disorder. The diagnosis indication 410 may include a diagnosis of a specific MSK disorder. The diagnosis indication 410 may include a treatment plan for the specific MSK disorder. The diagnosis indication 410 may include a prediction of a likelihood that the specific MSK disorder will occur in the future. In some implementations, the ML component 402 may determine the diagnosis indication 410 based on the symptom information 404, the patient information 406, and the motion image data 408.

[0054] The diagnosis indication 410 may be provided to a mitigation component 414, which may generate a treatment plan 416. The mitigation component 414 may be configured to generate a treatment plan 416 for the patient based on the diagnosis indication 410. In some implementations, the mitigation component 414 may be included in the ML component 402. In some implementations, the mitigation component 414 may be included in the patient device 106. In some implementations, the mitigation component 414 may be included in the MSK support platform 102. In some implementations, the mitigation component 414 may be a component of the patient device 106 and/or the MSK support platform 102. The treatment plan 416 may include a treatment regimen for the patient to follow to mitigate the MSK disorder.

[0055] To transform observed information such as motion image data into digital input for the ML component 402, the patient device 106 may include a camera that can be used to capture sequential still images or video while the patient performs specific exercises or movements. The captured images or motion data may then be processed to identify frames that accurately represent observed motion patterns.

[0056] Once the key frames or data points are identified, the motion image data 408 is converted into a digital format through various processing techniques. These techniques may include, but are not limited to, image segmentation to separate the subject (e.g., the patient's moving limb) from the background, and motion tracking algorithms to follow the movement of specific anatomical points over time. The processed image data may further be transformed into numerical datasets that represent attributes such as movement vectors, angles, and velocities. [0057] The digital data derived from the processed motion image data 408 may then be fed into the ML component 402. The ML component 402 may utilize these numerical datasets as input features. To handle these features, specific pre-processing steps may be employed, such as normalization or scaling of the motion attributes to ensure consistent input values before they are analyzed by the ML algorithms. The processed motion data may then be used to train the ML model, enabling it to learn patterns and correlations between the motion features and various MSK conditions. [0058] This transformation from observed motion image data to digital input ensures that the ML component 402 can effectively process and analyze the data, thereby improving the accuracy and reliability of the diagnosis indication 410 and subsequent treatment plan 416.

[0059] Figure 5 is a flow chart that illustrates an example method 500, in accordance with the present disclosure. The method 500 may be performed, for example, by one or more components of the system 300 shown in Figure 3, the computing device 200 shown in Figure 2, and/or the system 100 shown in Figure 1. At 502, a communication module (e.g., the communication module 308 shown in Figure 3) receives symptom information from a patient. The patient or someone authorized by the patient enters information using one of the devices 310 or 312 to describe the patient and the symptoms that the patient is experiencing. The symptom information may be collected through a user interface (e.g., provided via a website or a mobile application) designed to prompt detailed and structured responses. The user interface may dynamically generate questions based on answers to prior questions.

[0060] At 504, the communication module (e.g., the communication module 308 shown in Figure 3) receives motion image data showing at least one characteristic of movement of the patient. For example, the patient uses the camera 314 and/or the camera 316 to capture motion image data, which may be a series of still images or a video, such as, for example, a video showing the patient walking or bending over. The motion image data shows at least one characteristic of movement such as a range of motion, a type of motion, acceleration, or a combination of those. Depending on the particular situation, other characteristics of movement may be represented in the motion image data. Reported and/or captured (e.g., recorded) motion data may be transformed into a machine learning feature.

[0061] The feature may be scored based on subjective pain and physiological observations. For example, the straight leg raise (SLR) is central to the diagnosis of nerve root irritation, or sciatica. To convert the SLR response to an ML feature, the observed motion is classified as: a) POSITIVE (e.g., pain with knee extension in the gluteal region, possibly aggravated by foot dorsiflexion); b) Equivocal Non-specific pain (e.g., neither localized to gluteal region, no increased with knee extension); or c) NEGATIVE (e.g., No pain, full knee extension). Similarly, scoring may be applied for range of motion and power. Given that angled motion and sit-to-stand actions can be quantitatively assessed, the scoring methodology can be further refined.

[0062] An example of a spinal MSK diagnosis includes a diagnosis of spondyloarthropathy. Spondyloarthropathy is an inflammatory condition that frequently results in sacroiliac joint pain. For the diagnosis of this condition, some implementations may include the collection of data such as, for example, a history of morning stiffness with no radiation, a pain diagram indicating paraspinal pain, and/or an examination video demonstrating kyphotic posture, normal power, sensation, and pain with extension.

[0063] A function of diagnostic screening may be the identification of ‘red flag’ conditions requiring urgent medical intervention. These conditions may include: a) Vertebral body fracture; b) Active infection; and c) Malignancy. The diagnosis of these conditions is informed by demographic data, history of cancer, osteoporosis, systemic symptoms (fever, chills), persistent pain that is typically worse at night, a pain diagram indicating axial location, and limited or guarded motion.

[0064] At 506, the communication module (e.g., the communication module 308 shown in Figure 3) facilitates obtaining anatomy image data (e.g., from the patient imaging database 318 shown in Figure 3) based on the patient giving authorization for such records to be accessed. The anatomy image data provides information or an indication of a condition of the patient's anatomy involved in the symptoms and the motion image data. The anatomy image data may be obtained through advanced imaging techniques that provide high-resolution, multi-dimensional anatomical representations. For example, if the database 318 contains imaging study data from an x-ray, a computerized tomography scan, or a magnetic resonance imaging scan for the patient, that information is useful for diagnosing the patient's condition.

[0065] At 508, a computing device (e.g., the computing device 304 shown in Figure 3) uses a machine learning component to diagnose the patient's condition. The machine learning component takes the patient's symptom information, the patient's motion image data, and the patient's anatomy image data into account. The machine learning component identifies data from the machine learning database having corresponding characteristics to those indicated by the patient's data. The computing device determines whether there is a match, or at least close correspondence, between the combination of the patient's symptoms, motion image data and anatomy image data and corresponding data with a machine learning database a memory (e.g., the machine learning database in the memory 306 shown in Figure 3). When there is a sufficiently close correspondence between the patient's combined information and information associated with a known condition, the machine learning algorithm identifies that known condition as the condition of the patient.

[0066] Example conditions that can be diagnosed include soft tissue injury, joint arthrosis, neuropathy, paraspinal, inflammatory, infection, or malignancy. If the computing device is unable to diagnose the condition within one of such categories, the system will prompt the patient to provide additional information.

[0067] The machine learning algorithm may use any combination of techniques to detect or determine whether a patient's information corresponds to a known condition. For example, the symptom(s) described by the patient combined with the range of motion indicated by the motion image data may correspond to several possible conditions. If the anatomy image data does not correspond to image data related to at least some of those possible conditions within the machine learning database in the memory, then those conditions arc ruled out. The machine learning algorithm identifies the best combined match or correspondence for the information obtained from the patient to diagnose the patient's condition. In some aspects, the machine learning algorithm may include one or more machine learning models such as, for example, one or more classifiers and/or one or more neural networks. The machine learning algorithm may be trained on a curated dataset of clinically-validated cases to identify a musculoskeletal disorder affecting the patient.

[0068] Processing the patient's symptom information, the patient's motion image data, and/or the patient's anatomy image data may include, for example, applying a pre-processing step to normalize and filter the input data for noise reduction and consistency. A multi-layer neural network with optimized hyperparameters tailored for MSK disorder detection may be utilized to process the normalized and filtered input data. The neural network may generate diagnostic probabilities (e.g., identified candidate diagnoses with corresponding probability levels) and confidence scores for different MSK conditions. One or more diagnoses may be determined based on the diagnostic probabilities and confidence scores. For example, a diagnosis may be identified from among the diagnostic probabilities based on a corresponding confidence score satisfying a threshold.

[0069] Identifying corresponding motion image data may include, for example, known video processing techniques to locate a portion of a patient's body within the motion image data and to determine at least one characteristic of the motion represented in the data provided by the patient. The same technique may be used to process motion image data within the machine learning database. The motion image data may be directly compared to identify the same or similar motion within the patient's motion image data and the image data from the database in the memory. Alternatively, the patient motion image data may be processed and converted into a set of parameters or descriptive indicators. The database in the memory includes sets of parameters or descriptive indicators associated with known conditions. The machine algorithm in such embodiments identifies a best fit or match between the machine learning database and the patient's information.

[0070] In some aspects, the machine learning algorithm may be trained using historical data associated with the patient and/or with other patients. The machine learning algorithm may be configured to detect, analyze and/or otherwise extract at least one characteristic of movement from the motion image data. The at least one characteristic of movement may include any number of different motion features. Motion features may include any number of different characteristics of the patient's motion that are captured in the image data such as for example, motion speed (e.g., a speed and/or velocity of a movement of the patient or a body part thereof), motion pattern features (e.g., rhythms inherent in the motion, patterns of movement such as jittery motion or shaking, stopping and starting, motion indicative of hesitation, among other examples), and/or motion directions (e.g., a direction of motion relative to a coordinate system, an axis, and/or a body part), among other examples. For example, in some aspects, the machine learning algorithm may be configured to determine a velocity (speed and direction) of a motion of the patient or a body part thereof.

[0071] By utilizing a camera (e.g., the camera 34 and/or the camera 36) to capture patient motion image data, more subtle motion features than are observable by the human eye may be extracted and used by the computing device 24 to facilitate more accurate diagnosis, thereby increasing the likelihood of correctly diagnosing MSK disorders early enough to improve potential outcomes. In some aspects, motion feature measurements may be categorized according to any number of different grading and/or indexing schemes. For example, a motion of the patient (or a body pail thereof) may be graded as slow, medium, or fast. In another example, the motion may be categorized as smooth, halting, hesitating, or impeded. In another example, the motion may be indexed according to a relative speed value and/or direction value. The machine learning algorithm may be trained, using supervised and/or unsupervised learning, to determine one or more possible diagnoses based on the motion image data, a feature extracted therefrom, and/or any number of other types of information as described herein.

[0072] Data from reputable published articles and clinical guidance by MSK medical specialists may be used to produce a supervised machine learning model. Ranking of features based on the respective predictive value power may be implemented to further refine the machine learning model. Published medical literature and clinical guidance may be used to determine the diagnostic significance of selected features and/or independent variables.

[0073] The computing device processes anatomy image data in a similar manner as just described regarding the motion image data. That is, the machine learning algorithm may determine similarities between a patient's imaging study and at least one image in the database by directly comparing features or characteristics of the respective images. Alternatively, the images may be processed and converted into parameters or descriptive indicators that serve as a basis for comparison.

[0074] In some embodiments, when multiple imaging studies have been performed for a particular patient, a combination of information from such studies may facilitate the diagnosis. [0075] At 510, the computing device determines a therapy recommendation for providing to the patient to address the diagnosed condition. At 512, the computing device generates output indicative of a therapy regimen explanation for the patient and a communication module (e.g., the communication module 308 shown in Figure 3) provides that to a user device (e.g., the patient device 310 and/or the patient device 312 shown in Figure 3).

[0076] The output from the system may facilitate a patient self-administering the therapy to address or alleviate the symptoms caused by the diagnosed condition in many situations. The output from the system in some embodiments includes video tutorials of how to perform exercises or types of movement or situations to avoid, for example. The output may be customized based on the patient's information, such as showing a female of a selected age performing recommended exercises when the patient is female and within a corresponding age range. To the extent the diagnosed condition is not properly treated through a self-administered therapy regimen, the output provided to the patient may include recommendations or referrals for appropriate care from a medical professional.

[0077] The system may facilitate monitoring patient compliance with the recommended therapy regimen and tracking improvement. The patient may record video, for example, of the patient performing exercises. That additional motion image data may allow the computing device, through the machine learning algorithm, to recognize or identify improvement or decline in the patient's condition to make alternative recommendations or to refer the patient to a local specialist.

[0078] An example scenario including use of the example system for a patient experiencing lower back pain (LBP) follows. To use the service, the prospective patient or user is directed to a website for registration, consent, and a short survey. Alternatively, the user downloads an application (e.g., a phone app) to a patient device (e.g., the patient device 310 and/or the patient device 312 shown in Figure 3). Questions are focused on LBP symptoms such as pain level, function, and underlying psychological condition.

[0079] The patient questions in an example embodiment are presented as follows. a) Did your pain start after an injury yes no. If yes, please indicate the date b) Did you have x-ray or MRI done? yes no c) Are you disabled due to the back pain? yes no d) Are you taking prescription pain medications? yes no e) Do you have medications allergies? no yes: (list them) f) Are you seeing a doctor for cancer diabetes chronic pain g) Does your pain travel down the legs yes no h) Do you have numbness or weakness in the legs? If yes right left both i) Is your pain getting better staying the same worse j) Are you currently receiving treatment? Physical therapy chiropractic

> pain injections k) What’s your date of birth? l) How tall are you? m) How much do you weigh? n) Do you take over the counter medications? yes no. Does the medication help?

> yes, no

[0080] In some implementations, an artificial intelligence (Al) driven chatbot may be used to present follow-up questions such as, for example, “are your symptoms improving?” and “were you diagnosed with osteoporosis or risk for bone fractures?”. Some embodiments include presenting the patient with a survey such as the PHQ9 Patient Health Questionnaire or the GAD7 Generalized Anxiety Disorder Screener. [0081] Relevant imaging, such as x-ray, CT and MRI are pulled from the database 38 and catalogued. In a situation involving spine imaging data, the information is classified into at least one of the following:

1. No findings, negative

2. Arthrosis

- Spine

- Sacroiliac

3. Spondylosis

4. Incidental, paraspinal

- Renal

- Vascular-, AAA

- Abdominal

5. Bone

- Deformity

- Metastasis

- Fracture

- Ankylosing

6. Stenosis

- Central

- Foraminal

7. Compression of the thecal sac and/or nerve root (e.g., by finding nerve tension and/or nerve dysfunction)

[0082] Physical exam by a medical professional may not be needed because the patient may use a camera (e.g., the camera 314 and/or the camera 316 shown in Figure 3) to capture motion image data for motion tracking and analysis. Examination of LBP may include analysis of gait (heel/toe/tandem), sit-to-stand, get up and go, flexion/extension/rotation of the spine one leg standing balance and Rhomberg, straight leg raise, and/or hip motion, among other examples. In some implementations, the camera may be used to record the patient’s standing posture to be analyzed for diagnosis of kyphosis and/or scoliosis. The machine learning component may use the clinical history, imaging, and/or motion image data to determine an accurate LBP diagnosis, which is a type of MSK diagnosis. [0083] Example embodiments include, but are not limited to, using motion tracking to assess any of the following: a) Global limitation of spine motion

- Trauma

- Malignancy

- Symptom magnification

- Spondylosis, advanced

- Arthrosis, advanced

- Inflammatory b) Limited gait; include heel/toe and tandem

- Symptoms magnification

- Focal weakness, radiculopathy

- Upper motor neuron disorder

- Localize the spinal level

- Focal weakness

- Bilateral weakness

- Coordination and balance disturbance c) Limited flexion

- Soft tissue/acute injury

- Discogenic/spondylosis

- Inflammatory

- Symptoms exaggeration

- Nerve tension, irritation, or compression

- Mild

- Moderate

- Severe d) Limited extension

- Facet, arthritis joint

- SI pain

- Inflammatory

- Symptom magnification e) Straight leg raise (SLR) with neck flexion enhancement

- Nerve tension, direct or crossed

- Bilateral or unilateral

- Symptom exaggeration f) Inability to stand from seated position

- Weakness

- Symptom magnification

- Non spine disorder g) Specific joint angle or motion

- voice activation may be used to facilitate distinguishing between pain and anatomic mediated motion restriction

[0084] Once a diagnosis is established, educational material and appropriate treatment options may be offered. Example outputs from the system may include, but not be limited to, any one or more of: exercise instructions through the user online portal or smartphone application, over-the-counter medication recommendation, therapeutic tools recommendation, durable medical equipment prescription, pharmaceutical prescription, referral for in-person care, referral for physical therapy, referral for chiropractic care, order(s) for additional diagnostic testing, and consultation with a specialty provider. In addition, a phone application implemented in connection with the techniques described herein may assign and monitor the performance of therapeutic exercises. Patients may be encouraged to use the application for maintenance exercise, and investigation of new symptoms, should they arise. Treatment recommendations may be dynamic, reflecting the changing nature of MSK disorders.

[0085] Options for treatment include topical modalities, medications, psychological support, telemedicine visit, and/or monitored exercise, among other examples. Patients that do not respond may be offered interventional procedures or spine surgery.

[0086] If despite educated recommendations and patient participation, there’s no symptomatic improvement, or in the event the system detects a ‘red flag’, the patient may be directed to schedule a virtual appointment with an MSK clinician. More advanced interventions including updated spine imaging, referral to a primary care physician for evaluation of a non MSK disorder, or in-person encounter may be provided.

[0087] One feature of the example system is that it can be easily accessible, simple to use and affordable. Operation is autonomous and may help all patients equally and do no harm. Data may be cvidcncc-bascd, and protocols may be transparent. Some implementations may increase a likelihood for improved patient health beyond the direct symptoms of MSK. Therapeutic exercise and reduced pain may also help patients bring down blood pressure, compete for employment, and/or lower the risk of chronic disease, among other examples. Additionally, the system can benefit patients recovering from strokes, participating in cardiac rehabilitation, attempting to treat diabetes through weight loss, and/or restoring joint motion after orthopedic surgery, among other examples.

[0088] Figure 6 is a flow chart that illustrates an example method 600, in accordance with the present disclosure. The method 600 may be performed, for example, by one or more components of the system 300 shown in Figure 3, the computing device 200 shown in Figure 2, and/or the system 100 shown in Figure 1.

[0089] At 602, a computing device (e.g., the computing device 200 shown in Figure 2) obtains, via at least one camera of a patient device, motion image data indicating at least one characteristic of movement of a patient. In some implementations, the patient device may include a smartphone and/or a portable computer. The motion image data may include video data or still image data.

[0090] At 604, the computing device provides at least the motion image data for input to a machine learning component configured to determine a diagnosis corresponding to a musculoskeletal disorder affecting the patient based at least in part on motion image data. In some implementations, the computing device may obtain anatomy image data by accessing a database storing the anatomy image data. The anatomy image data may include data from a previously performed imaging study of a portion of the patient's anatomy associated with the at least one characteristic of movement. The at least one characteristic of movement may include at least one of a velocity, an acceleration, a range of motion, or a type of motion.

[0091] The machine learning component may be trained using historical data associated with the patient. The machine learning component may be trained using historical data associated with at least one other patient. The machine learning component may be configured to detect at least one motion feature from the motion image data. The at least one motion feature may include at least one of a motion speed, a motion pattern, or a motion direction. The at least one motion feature may include a motion direction relative to a coordinate system. The at least one motion feature may include a motion direction relative to a body part of the patient.

[0092] At 606, the computing device provides, for display via the patient device, an output indicative of treatment information associated with the musculoskeletal disorder. In some implementations, the output may include an indication of a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder.

[0093] Some embodiments are described as numbered examples (Example 1, 2, 3, etc.). These are provided as examples only and do not limit the technology disclosed herein.

[0094] Example 1 includes a method of facilitating patient therapy, the method comprising: obtaining, via a computing device, symptom information from the patient representing symptoms experienced by the patient; obtaining, via at least one camera, motion image data indicating at least one characteristic of movement of the patient; obtaining, via the computing device, anatomy image data indicating a condition of at least one portion of the patient's anatomy; processing, using the computing device, the symptom information, the motion image data, and the anatomy image data using a machine learning algorithm to identify a musculoskeletal disorder affecting the patient; determining, using the computing device, a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder; and providing, via a communication module, the self-administrable therapy regimen to the patient.

[0095] Example 2 includes the method of Example 1, wherein obtaining the symptom information and the motion image data includes receiving the symptom information and the motion image data from a device used by the patient, the device comprising the at least one camera.

[0096] Example 3 includes the method of Example 2, wherein the device used by the patient is a mobile station comprising at least one of a smartphone or a portable computer.

[0097] Example 4 includes the method of any of Examples 1-3, wherein obtaining the anatomy image data includes accessing a database storing the anatomy image data.

[0098] Example 5 includes the method of any of Examples 1-4, wherein the at least one characteristic of movement indicated by the motion image data comprises at least one of a velocity, an acceleration, a range of motion, or a type of motion.

[0099] Example 6 includes the method of any of Examples 1-5, wherein the anatomy image data comprises data from a previously performed imaging study of a portion of the patient's anatomy associated with the at least one characteristic of movement. [0100] Example 7 includes the method of any of Examples 1 -6, wherein the machine learning algorithm is trained using historical data associated with the patient.

[0101] Example 8 includes the method of any of Examples 1-7, wherein the machine learning algorithm is trained using historical data associated with at least one other patient. [0102] Example 9 includes the method of any of Examples 1-8, wherein the machine learning algorithm is configured to detect at least one motion feature from the motion image data.

[0103] Example 10 includes the method of Example 9, wherein the at least one motion feature comprises at least one of a motion speed, a motion pattern, or a motion direction.

[0104] Example 11 includes the method of Example 9, wherein the at least one motion feature comprises a motion direction relative to a coordinate system.

[0105] Example 12 includes the method of Example 9, wherein the at least one motion feature comprises a motion direction relative to a body part of the patient.

[0106] Example 13 includes the method of any of Examples 1-12, wherein the motion image data comprises video data or still image data.

[0107] Example 14 includes a non-transitory computer-readable medium storing instructions that, when executed by a computing device, cause the computing device to perform the method of any of Examples 1-13.

[0108] Example 15 includes a device for facilitating patient therapy, the device comprising: a processor; and a memory coupled with the processor and including instructions that, when executed by the processor, are configured to cause the processors to cause the device to perform the method of any of Examples 1-13.

[0109] Example 16 includes a method of facilitating patient therapy, the method comprising: obtaining, via at least one camera of a patient device, motion image data indicating at least one characteristic of movement of a patient; providing at least the motion image data for input to a machine learning component configured to determine a diagnosis corresponding to a musculoskeletal disorder affecting the patient based at least in part on motion image data; and providing, for display via the patient device, an output indicative of treatment information associated with the musculoskeletal disorder.

[0110] Example 17 includes the method of Example 16, wherein the output comprises an indication of a self-administrable therapy regimen for the patient based on the musculoskeletal disorder.

[0111] Example 18 includes the method of either of Examples 16 or 17, wherein the patient device comprises at least one of a smartphone or a portable computer.

[0112] Example 19 includes the method of any of Examples 16-18, further comprising obtaining anatomy image data by accessing a database storing the anatomy image data.

[0113] Example 20 includes the method of Example 19, wherein the anatomy image data comprises data from a previously performed imaging study of a portion of an anatomy of the patient associated with the at least one characteristic of movement.

[0114] Example 21 includes the method of any of Examples 16-20, wherein the at least one characteristic of movement comprises at least one of a velocity, an acceleration, a range of motion, or a type of motion.

[0115] Example 22 includes the method of any of Examples 16-21, wherein the machine learning component is trained using historical data associated with the patient.

[0116] Example 23 includes the method of any of Examples 16-22, wherein the machine learning component is trained using historical data associated with at least one other patient.

[0117] Example 24 includes the method of any of Examples 16-23, wherein the machine learning component is configured to detect at least one motion feature from the motion image data.

[0118] Example 25 includes the method of Example 24, wherein the at least one motion feature comprises at least one of a motion speed, a motion pattern, or a motion direction.

[0119] Example 26 includes the method of either of Examples 24 or 25, wherein the at least one motion feature comprises a motion direction relative to a coordinate system.

[0120] Example 27 includes the method of any of Examples 24-26, wherein the at least one motion feature comprises a motion direction relative to a body part of the patient.

[0121] Example 28 includes the method of any of Examples 16-27, wherein the motion image data comprises video data or still image data.

[0122] Example 29 includes a non-transitory computer-readable medium storing instructions that, when executed by a computing device, cause the computing device to perform the method of any of Examples 16-28.

[0123] Example 30 includes a device for facilitating patient therapy, the device comprising: a processor; and a memory coupled with the processor and including instructions that, when executed by the processor, are configured to cause the processors to cause the device to perform the method of any of Examples 16-28.

[0124] Example 31 includes a method of facilitating patient therapy, the method comprising: obtaining symptom information from a patient indicating symptoms experienced by the patient, wherein the symptom information is collected through a user interface designed to prompt detailed and structured responses; obtaining motion image data indicating at least one characteristic of movement of the patient; obtaining anatomy image data indicating a condition of at least one portion of an anatomy of the patient, wherein the anatomy image data is obtained through advanced imaging techniques that provide high-resolution, multi-dimensional anatomical representations; processing the symptom information, the motion image data, and the anatomy image data using a machine learning algorithm to identify a musculoskeletal disorder affecting the patient, wherein the machine learning algorithm is trained on a curated dataset of clinically-validated cases to identify a musculoskeletal disorder affecting the patient, the processing further comprising: converting the motion image data into a set of parameters; and utilizing the machine learning algorithm to identify the musculoskeletal disorder; determining a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder, wherein the regimen is customized by integrating patient-specific information; and providing the self-administrable therapy regimen to the patient, wherein the regimen is delivered through an interactive application that guides the patient through each step of the therapy, monitors adherence, and adjusts the regimen dynamically based on patient input and progress data.

[0125] Example 32 includes the method of Example 31, wherein obtaining the symptom information and the motion image data includes receiving the symptom information and the motion image data from a device used by the patient, the device comprising the at least one camera.

[0126] Example 33 includes the method of Example 32, wherein the device used by the patient is a mobile station comprising at least one of a smartphone or a portable computer. [0127] Example 34 includes the method of any of Examples 31-33, wherein obtaining the anatomy image data includes accessing a database storing the anatomy image data.

[0128] Example 35 includes the method of any of Examples 31-34, wherein the at least one characteristic of movement indicated by the motion image data comprises at least one of a velocity, an acceleration, a range of motion, or a type of motion.

[0129] Example 36 includes the method of any of Examples 31-35, wherein the anatomy image data comprises data from a previously performed imaging study of a portion of the patient's anatomy associated with the at least one characteristic of movement.

[0130] Example 37 includes the method of any of Examples 31-36, wherein the machine learning algorithm is trained using historical data associated with the patient.

[0131] Example 38 includes the method of any of Examples 31-37, wherein the machine learning algorithm is trained using historical data associated with at least one other patient.

[0132] Example 39 includes the method of any of Examples 31-38, wherein the machine learning algorithm is configured to detect at least one motion feature from the motion image data.

[0133] Example 40 includes the method of Example 39, wherein the at least one motion feature comprises at least one of a motion speed, a motion pattern, or a motion direction.

[0134] Example 41 includes the method of Example 39, wherein the at least one motion feature comprises a motion direction relative to a coordinate system.

[0135] Example 42 includes the method of Example 39, wherein the at least one motion feature comprises a motion direction relative to a body part of the patient.

[0136] Example 43 includes the method of any of Examples 31-42, wherein the motion image data comprises video data or still image data.

[0137] Example 44 includes a non-transitory computer-readable medium storing instructions that, when executed by a computing device, cause the computing device to perform the method of any of Examples 31-43.

[0138] Example 45 includes a device for facilitating patient therapy, the device comprising: a processor; and a memory coupled with the processor and including instructions that, when executed by the processor, are configured to cause the processors to cause the device to perform the method of any of Examples 31-43.

[0139] The preceding description is exemplary rather than limiting in nature. Valuations and modifications to the disclosed examples may become apparent to those skilled in the art that do not necessarily depart from the essence of the present disclosure. The scope of legal protection given to the present disclosure can only be determined by studying the following claims.

[0140] As used herein, unless explicitly stated otherwise, any term specified in the singular may include its plural version. For example, “a computer that stores data and runs software,” may include a single computer that stores data and runs software or two computers - a first computer that stores data and a second computer that runs software. Also “a computer that stores data and runs software,” may include multiple computers that together stored data and run software. At least one of the multiple computers stores data, and at least one of the multiple computers runs software.

[0141] As used herein, the term “computer-readable medium” encompasses one or more computer readable media. A computer-readable medium may include any storage unit (or multiple storage units) that store data or instructions that are readable by processing circuitry. A computer-readable medium may include, for example, at least one of a data repository, a data storage unit, a computer memory, a hard drive, a disk, or a random access memory. A computer- readable medium may include a single computer-readable medium or multiple computer- readable media. A computer-readable medium may be a transitory computer-readable medium or a non-transitory computer-readable medium.

[0142] As used herein, the term “memory” includes one or more memories, where each memory may be a computer-readable medium. A memory may encompass memory hardware units (e.g., a hard drive or a disk) that store data or instructions in software form. Alternatively or in addition, the memory may include data or instructions that are hard-wired into processing circuitry.

[0143] The implementations of this disclosure can be described in terms of functional block components and various processing operations. Such functional block components can be realized by a number of hardware or software components that perform the specified functions. For example, the disclosed implementations can employ various integrated circuit components (e.g., memory elements, processing elements, logic elements, look-up tables, and the like), which can carry out a variety of functions under the control of one or more microprocessors or other control devices. Similarly, where the elements of the disclosed implementations are implemented using software programming or software elements, the systems and techniques can be implemented with a programming or scripting language, such as C, C++, Java, JavaScript, assembler, or the like, with the various algorithms being implemented with a combination of data structures, objects, processes, routines, or other programming elements.

[0144] Functional aspects can be implemented in algorithms that execute on one or more processors. Furthermore, the implementations of the systems and techniques disclosed herein could employ a number of conventional techniques for electronics configuration, signal processing or control, data processing, and the like. The words “mechanism” and “component” are used broadly and are not limited to mechanical or physical implementations, but can include software routines in conjunction with processors, etc. Likewise, the terms “system” or “tool” as used herein and in the figures, but in any event based on their context, may be understood as corresponding to a functional unit implemented using software, hardware (e.g., an integrated circuit, such as an ASIC), or a combination of software and hardware. In certain contexts, such systems or mechanisms may be understood to be a processor-implemented software system or processor- implemented software mechanism that is part of or callable by an executable program, which may itself be wholly or partly composed of such linked systems or mechanisms.

[0145] Implementations or portions of implementations of the above disclosure can take the form of a computer program product accessible from, for example, a computer-usable or computer-readable medium. A computer-usable or computer-readable medium can be a device that can, for example, tangibly contain, store, communicate, or transport a program or data structure for use by or in connection with a processor. The medium can be, for example, an electronic, magnetic, optical, electromagnetic, or semiconductor device.

[0146] Other suitable mediums are also available. Such computer-usable or computer- readable media can be referred to as non-transitory memory or media, and can include volatile memory or non-volatile memory that can change over time. The quality of memory or media being non-transitory refers to such memory or media storing data for some period of time or otherwise based on device power or a device power cycle. A memory of an apparatus described herein, unless otherwise specified, does not have to be physically contained by the apparatus, but is one that can be accessed remotely by the apparatus, and does not have to be contiguous with other memory that might be physically contained by the apparatus.

[0147] While the disclosure has been described in connection with certain implementations, it is to be understood that the disclosure is not to be limited to the disclosed implementations but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims, which scope is to be accorded the broadest interpretation so as to encompass all such modifications and equivalent structures as is permitted under the law.

Claims

What is claimed is:

1. A non-transitory computer-readable medium storing instructions that, when executed by a computing device, cause the computing device to perform a method of facilitating patient therapy, the method comprising: obtaining symptom information from a patient indicating symptoms experienced by the patient, wherein the symptom information is collected through a user interface designed to prompt detailed and structured responses; obtaining motion image data indicating at least one characteristic of movement of the patient; obtaining anatomy image data indicating a condition of at least one portion of an anatomy of the patient, wherein the anatomy image data is obtained through advanced imaging techniques that provide high-resolution, multi-dimensional anatomical representations; processing the symptom information, the motion image data, and the anatomy image data using a machine learning algorithm to identify a musculoskeletal disorder affecting the patient, wherein the machine learning algorithm is trained on a curated dataset of clinically-validated cases to identify a musculoskeletal disorder affecting the patient, the processing further comprising: converting the motion image data into a set of parameters; and utilizing the machine learning algorithm to identify the musculoskeletal disorder; determining a self-administrable therapy regimen for the patient based on the identified musculoskeletal disorder, wherein the regimen is customized by integrating patient-specific information; and providing the self-administrable therapy regimen to the patient, wherein the regimen is delivered through an interactive application that guides the patient through each step of the therapy, monitors adherence, and adjusts the regimen dynamically based on patient input and progress data.

2. The non-transitory computer-readable medium of claim 1, wherein obtaining the symptom information and the motion image data includes receiving the symptom information and the motion image data from a device used by the patient, the device comprising at least one camera.

3. The non-transitory computer-readable medium of claim 2, wherein the device used by the patient is a mobile station comprising at least one of a smartphone or a portable computer.

4. The non-transitory computer-readable medium of claim 1, wherein obtaining the anatomy image data includes accessing a database storing the anatomy image data.

5. The non-transitory computer-readable medium of claim 1, wherein the at least one characteristic of movement indicated by the motion image data comprises at least one of a velocity, an acceleration, a range of motion, or a type of motion.

6. A method of facilitating patient therapy, the method comprising: obtaining, via at least one camera of a patient device, motion image data indicating at least one characteristic of movement of a patient; providing at least the motion image data for input to a machine learning component configured to determine a diagnosis corresponding to a musculoskeletal disorder affecting the patient based at least in part on motion image data; and providing, for display via the patient device, an output indicative of treatment information associated with the musculoskeletal disorder.

7. The method of claim 6, wherein the output comprises an indication of a self- administrable therapy regimen for the patient based on the musculoskeletal disorder.

8. The method of claim 6, wherein the patient device comprises at least one of a smartphone or a portable computer.

9. The method of claim 6, further comprising obtaining anatomy image data by accessing a database storing the anatomy image data.

10. The method of claim 9, wherein the anatomy image data comprises data from a previously performed imaging study of a portion of an anatomy of the patient associated with the at least one characteristic of movement.

11. The method of claim 6, wherein the at least one characteristic of movement comprises at least one of a velocity, an acceleration, a range of motion, or a type of motion.

12. The method of claim 6, wherein the machine learning component is trained using historical data associated with the patient.

13. The method of claim 6, wherein the machine learning component is trained using historical data associated with at least one other patient.

14. The method of claim 6, wherein the machine learning component is configured to detect at least one motion feature from the motion image data.

15. The method of claim 14, wherein the at least one motion feature comprises at least one of a motion speed, a motion pattern, or a motion direction.

16. The method of claim 14, wherein the at least one motion feature comprises a motion direction relative to a coordinate system.

17. The method of claim 14, wherein the at least one motion feature comprises a motion direction relative to a body part of the patient.

18. The method of claim 6, wherein the motion image data comprises video data or still image data.

19. A device for facilitating patient therapy, the device comprising: obtaining, via at least one camera of a patient device, motion image data indicating at least one characteristic of movement of a patient; providing at least the motion image data for input to a machine learning component configured to determine a diagnosis corresponding to a musculoskeletal disorder affecting the patient based at least in part on motion image data; and presenting, via the patient device, an output indicative of treatment information associated with the musculoskeletal disorder.

20. The device of claim 19, wherein the motion image data comprises video data or still image data.