WO2017062484A1

WO2017062484A1 - Method for arriving at diagnoses of illnesses through comparison of retina images

Info

Publication number: WO2017062484A1
Application number: PCT/US2016/055540
Authority: WO
Inventors: Emeric SILBERMANN
Original assignee: Individual
Current assignee: Individual
Priority date: 2015-10-09
Filing date: 2016-10-05
Publication date: 2017-04-13
Anticipated expiration: 2018-04-09

Abstract

A method for arriving at diagnoses for illnesses of patients based on comparison of retina images with predetermined ranges and thresholds. A patient's retina image is processed to extract information relating to predetermined characteristics with associated parameter ranges. Sets of characteristics with predetermined parameter ranges are matched with conditions. Predetermined conditions are matched to arrive to diagnoses. The parameter ranges are adjustable to increase/decrease the certainty of diagnoses or include others that are less probable.

Description

I. TITLE: "METHOD FOR ARRIVING AT DIAGNOSES FOR

ILLNESSES THROUGH COMPARISON OF RETINA IMAGES"

II. BACKGROUND OF THE INVENTION

1. Field of the Invention.

[001] The present invention relates to a diagnosis method and, more

particularly, to such a method that utilizes images from a patient's retina.

2. Description of the Related Art.

[002] Several methods using retina images have been used in the past. None of them, however, include a database storing predetermined characteristics extracted from the retina images that are compared with previously analyzed retina images from other patients with diagnosed conditions or illnesses. The algorithm used to extract the relevant characteristics from the retina images processes this information along with other general information about the patient including age, race, etc. to arrive at weighed probable diagnoses. By using predetermined thresholds for these weighed probable diagnoses a report is generated.

[003] Applicant believes that the closest reference corresponds to U.S. patent No. 8,243,999 and its continuation application resulting in patent No. 8,503,749 issued to Tobin for a Method and System for the Diagnosis of Disease Using Retinal Image Content and an Archive of Diagnosed Human Patient Data.

However, it differs from the present invention because Tobin' s patent reference relies on a comparison of photos to arrive at results (diagnostics) based on subjective considerations. These subjective considerations involve comparison of a patient's ocular photos with pre-existing sets of photos of other person's retinas. Depending on the resemblance of the patient's photo with those of eyes with different diseases, a result is obtained based on the probabilities of a patient having the same disease. On the other hand, the present invention identifies a

predetermined number of ocular characteristics that are objectively and specifically compared with ranges that have been determined and correspond to healthy eye parameters or not. And, if not falling under the healthy parameters, compared with the ranges that are associated with predetermined diseases.

[004] The fundus of the eye allows us to see and recognize changes induced by diseases affecting the circulatory system. Since the retina has a terminal vascular system type, supplied only by one artery without arterial-arterioles

communications, and drained by only one vein, the characteristics displayed in patient's retinas contain valuable and relevant information that can be used to arrive at valuable and relevant diagnoses. The fundus of the eye is the interior surface of the eye, opposite the lens, and includes the retina, optic disc, macula and fovea, and posterior pole. The fundus can be examined by ophthalmoscopy and/or fundus photography. The term fundus may also be inclusive of Brunch's

membrane and the choroid.

[005] The color of the fundus varies both between and within species. In one study of primates the retina is blue, green, yellow, orange, and red; only the human fundus (from a lightly pigmented blond person) is red. The major differences noted among the "higher" primate species were size and regularity of the border of the macular area, size and shape of the optic disc, apparent 'texturing' of retina, and pigmentation of retina. [006] Retinal venous blood is transported by the central retinal vein, which the tributary vein of the superior ophthalmic vein that drains into the cavernous sinus. Infectious and vascular diseases at this level can express in the fundus of the eye (septic thrombosis, carotid-cavernous fistulas). Central retinal vein occlusion (CRVO) is essentially a diagnostic finding of painless unilateral loss of vision, ophthalmoscopy findings consist of variable dot and flame hemorrhages in all four quadrants, optic swelling, retinal vein engorgement and tortuosity, cotton wool spots are few.

[007] The ophthalmic artery is the first branch of the internal carotid artery and enters the orbit underneath the optic nerve, visual loss central retinal artery occlusion (CRAO) occurs from the loss of blood supply to the inner layer of the retina. CRAO may be due to embolism or thrombosis. Emboli may come from any of the following: atherosclerotic plaques, endocarditis, fat, and atrial myxoma. Patients with central retinal arterial occlusion, have a risk of stroke.

[008] Hollenhorst plaques are cholesterol crystals that get stuck in the branches of the arterioles; they are highly refractive to the light. Its presence denounces the existence of a severe systemic atherosclerosis and is associated to increase risk of stroke, myocardial infarction, aortic dissection, and congestive heart failure.

[009] The fundus of the eye is the only place in the human body where we can directly visualize a small piece of brain white matter. These are compartmentalized axons of retinal ganglion cells that form the optic nerve. The optic fibers lack the myelin sheath that surrounds nerves, and are therefore translucent resembling more a brain tissue than a nerve tissue. Papilledema is one of the main signs of intracranial hypertension of any origin, and alerts the clinician of the serious threat that could threaten the patient. The optic nerve is unsheathed in all three meningeal layers (dura, arachnoid and pia mater) the subarachnoid space of the brain is continuous with the optic nerve sheath. Hence, as the cerebrospinal fluid (CSF) pressure increases, the pressure is transmitted to the optic nerve, and the optic nerve sheath acts as a tourniquet to impede axoplasmic transport. This leads to a buildup of material at the level of the lamina cribrosa, resulting in the characteristic swelling of the nerve head.

[010] Many viral diseases can be reflected in the fundus of the eye. CMV retinitis is diagnosed through a standard ophthalmologic exam. This condition is characterized by a necrotizing retinitis associated to venous vasculitis and hemorrhages; usually there are no signs of inflammation and the vitreous is clear. The typical lesion is a triangular area with gross whitish infiltration and irregular margins that separates this area from the healthy retina.

[011] Other documents describing the closest subject matter provide for a number of more or less complicated features that fail to solve the problem in an efficient and economical way. None of these patents suggest the novel features of the present invention.

Ill SUMMARY OF THE INVENTION [012] It is one of the main objects of the present invention to provide a method for readily diagnosing conditions and/or illnesses using the patient's retina images. [013] It is another object of this invention to provide such a method that can be utilized remotely and with real time access to the results.

[014] It is still another object of the present invention to provide a method that is not intrusive to the patient to diagnose: hypertension, intracranial hypertension, diabetes, brain tumor, stroke, carotid atherosclerosis, stenosis, viral infection with cytomegalovirus, viral infection with human immunodeficiency virus, systemic lupus erythematous, infection with toxoplasmosis, takayasu disease, multiple myeloma, multiple sclerosis, optic neuritis, neuroretinitis, temporal arteritis, hypertriglyceridemia, lupus anticoagulant or antiphospholipid syndrome, infectious endocarditis, macular degeneration, eye melanoma, thrombotic thrombocytopenic purpura, septic retinitis (bacterial septicemia), glaucoma, hypercoagulable states, pseudoxanthoma elasticum, paget disease, sarcoidosis, and syphilis. [015] It is yet another object of this invention to provide such a method that is inexpensive while retaining its effectiveness with adjustable statistical confidence levels for arriving at diagnoses for predetermined diseases.

[016] Further objects of the invention will be brought out in the following part of the specification, wherein detailed description is for the purpose of fully disclosing the invention without placing limitations thereon.

IV. BRIEF DESCRIPTION OF THE DRAWINGS [017] With the above and other related objects in view, the invention consists in the details of construction and combination of parts as will be more fully understood from the following description, when read in conjunction with the accompanying drawings in which:

Figure 1 represents a flow chart representing the main steps of the method subject of the present invention.

Figure 2 is a diagram representation of the characteristics detected that are compared against adjustable reference ranges to arrive at a predetermined number of conditions that when combined result in one or more diagnoses for illnesses.

Figure 3 represents four tables in a database, namely,

CHARACTERISTICS, RANGES, CONDITIONS, and DIAGNOSES with symbolic fields in each.

Figure 4 represents the method shown in figure 1 with the conditions being taken in a loop n times.

V. DETAILED DESCRIPTION OF THE EMBODIMENTS OF THE

INVENTION

[018] The present invention can be implemented with the use of an eye adapter for digital cameras capable of obtaining images of a patient's retina with a predetermined resolution capable of identifying a predetermined number of characteristics of the retina. The digitized images are then transferred to a computerized server station with sufficient software and storage resources to extract a predetermined number of characteristics of the retina images that could be used, in conjunction with the general information of the patient to grade or provide different weights to the probable diagnoses for the patient.

[019] The characteristics (CH_1-n) extracted from the retina images used in the present invention include:

CHI . Color of the fundus

CH2. Size of the macular area

CH3. Regularity of the macular area

CH4. Size of the optic disc

CH5. Shape of the optic disc

CH6. Pigmentation of the retina

CH7. Characteristics of arterial systems

CH8. Characteristics of venous systems

CH9. Presence of hemorrhages

CH10. Presence of exudates

CH11. Presence of macular edema

CH12. Presence of optic disc edema

CHI 3. Presence of optic disc atrophy [020] The general information about a patient can also be taken into

consideration, such as, age and race.

[021] After receiving the images from a remote ophthalmoscopic device, the relevant retina characteristics (CH_1-n) are extracted and compared with a number of pre-stored characteristics derived from other patients with predetermined

conditions met to arrive to predetermined diagnoses. A matching event is saved associating the patient's characteristics with groups of the pre-stored characteristics of different persons with predetermined conditions thereby arriving at ranges for the characteristics required for a given condition for a possible diagnosis. A statistical profile is arrived at for a preliminary probable diagnosis based on the highest number of matches. The object being to eliminate those conditions that are more remotely associated with the patient so that a more focused analysis can be undertaken for the most probable diagnoses. For this more precise portion of the process, a specific set of ranges (RNG _1-m) for the retina characteristics (CH_1-n) are associated with predetermined conditions (COND_1-p) for a given illness diagnosis and compared with the relevant characteristics of a patient. For example, a given condition for a given illness diagnosis is met when retina characteristics CHI ; CH2 and CH3 fall within ranges NGi; RNG₂; and RNG₃. Different grading or weights that will affect the ultimate result for arriving at a diagnosis are given to the retina characteristics depending on their relevancy for a particular condition of a suspected diagnosis. Also, the weights applied may be affected by the general information of the patient.

[022] After all the probable conditions are evaluated, a threshold is selected to refine the confidence level required from the method to determine one or more diagnoses (DIAG_1-q). This may eliminate questionable diagnoses leaving the ones with more certainty. Each diagnosis (DIAG_X) will require that at least one condition (COND_x) be met for at least one characteristic (CH_X) detected within one predetermined range (RNG_X). The ranges can be limited or expanded

independently from each other and depending on the user (sometimes he/she could be a physician).

[023] With the present invention, a user can change the threshold, as

discussed, to make more or less strict the diagnoses determination. [024] The foregoing description conveys the best understanding of the objectives and advantages of the present invention. Different embodiments may be made of the inventive concept of this invention. It is to be understood that all matter disclosed herein is to be interpreted merely as illustrative, and not in a limiting sense.

VI. INDUSTRIAL APPLICABILITY [025] It is apparent from the previous paragraphs that an improvement of the type for such a method is quite desirable for arriving at diagnoses for illnesses through comparison of retina images.

Claims

VII. CLAIMS What is claimed is:

1. A method for arriving at diagnoses for a predetermined number of illnesses or conditions of a patient comprising the steps of:

A) acquiring images of a patient's retina;

B) transmitting said images to a computerized server for extracting at least one retina characteristic from said images;

C) comparing said at least one retina characteristic with one

predetermined range for each of said at least one characteristic;

D) generating a grading value for each of said at least one retina

characteristic that falls, within said at least one range; and

E) generating a report listing at least one illness diagnosis when

predetermined grading values are reached.

2. The method set forth in claim 1 wherein said range for each of said at least one characteristic is set for each of said at least one diagnosis.

3. The method set forth in claim 2 wherein at least one of said ranges for at least one of said characteristics is adjustable to increase or decrease the certainty probabilities for said at least one diagnosis.

4. The method set forth in claim 3 wherein said characteristics include color of the fundus, size of the macular area, regularity of the macular area, size of the optic disc, shape of the optic disc, pigmentation of the retina characteristics of arterial systems, characteristics of venous systems, presence of hemorrhages, presence of exudates, presence of macular edema, presence of optic disc edema, and presence of optic disc atrophy for diagnosing at least one of a group of predefined illnesses.