Glaucoma screening method, ophthalmic device, program, and recording medium

Abstract

A glaucoma screening method according to an embodiment includes the following steps. In the calculation model preparation step, a calculation model is prepared, the calculation model including a linear coupling of one or more teat parameters representing the thickness of the retinal tissue in the optic papilla region and a linear coupling of one or more macular parameters representing the thickness of the retinal tissue in the macular region. In the OCT data generation step, OCT data is generated by applying an OCT scan to the fundus of an examined eye. In the measurement value calculation step, a plurality of measurement values are calculated on the basis of the OCT data, the plurality of measurement values including a nipple measurement value corresponding to each nipple parameter of the calculation model and a macular measurement value corresponding to each macular parameter. In the measurement value input step, the plurality of calculated measurement values are input into the calculation model. In the calculation result providing step, calculation results output from the calculation model in accordance with the input of the plurality of measurement values are provided.

Description

Technical Field

[0001] The present disclosure relates to a glaucoma screening method, an ophthalmic device, a program, and a recording medium. Background Art

[0002] Glaucoma is one of the leading causes of blindness in early life. It is estimated that the number of glaucoma patients in the world is as high as 60 million or more, and it is expected to increase further. According to some epidemiological surveys, most glaucoma patients have not received a diagnosis and are unaware that they have glaucoma.

[0003] In order to prevent visual field loss or blindness caused by glaucoma from limiting daily life, it is important to detect the disease early and start treatment. In order to promote early detection, it is believed that widely providing screening using eye examinations is effective. However, currently, collective examinations (large-scale screening) for glaucoma are not common in many countries.

[0004] Patent Document 1: Japanese Patent Application Laid-Open No. 5-184542

[0005] Patent Document 2: U.S. Patent Application Publication No. 2018 / 0025112 Summary of the invention

[0006] In order to achieve large-scale glaucoma screening, there are some issues that need to be addressed. Two representative issues are described below.

[0007] The first problem is the reliability of glaucoma determination. For example, in the previous glaucoma determination method using image analysis, due to the influence of artifacts and vascular images, even normal eyes are often mistakenly determined to have retinal thinning. That is, the previous method has a high incidence of false positives. If it is a professional doctor, it is easy to distinguish false positives caused by artifacts. However, large-scale screening is usually implemented in an environment without professional doctors, so normal eyes may be classified as suspected glaucoma eyes. The examinee receives a precise examination in a special hospital, but the result is of course negative. The occurrence of this situation not only reduces the reliability of the large-scale screening implementation facilities and the reliability of the screening method, but also reduces the quality of the statistical data created based on the data collected through large-scale screening. In addition, for each examinee, it is a futile burden psychologically, labor-intensively, financially and in time. Therefore, one condition for achieving large-scale glaucoma screening is to improve the quality of disease determination (precision, accuracy, etc.).

[0008] The second problem is the difficulty and complexity of determining normal-tension glaucoma. Normal-tension glaucoma is one of the types of glaucoma. Unlike the type of glaucoma that develops due to increased intraocular pressure damaging the optic nerve, it is a type of glaucoma in which the intraocular pressure is normal but the optic nerve is damaged. In previous glaucoma screening methods, there were many cases in which screening using intraocular pressure as an indicator (biomarker) was initially performed, so it was impossible to determine normal-tension glaucoma that did not show increased intraocular pressure. In order to determine normal-tension glaucoma, in addition to intraocular pressure examination, the doctor can only make a judgment based on fundus examination and visual field examination. Obviously, such a method is not suitable for large-scale screening.

[0009] One object of the present disclosure is to provide a glaucoma determination technology suitable for large-scale screening.

[0010] An exemplary embodiment of the present disclosure is a method for glaucoma screening, which includes preparing a computational model, the computational model including a linear coupling of one or more papillary parameters representing the thickness of retinal tissue in an optic nerve head region and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in a macular region, applying optical coherence tomography to the fundus of an eye under examination to generate OCT data, calculating a plurality of measurement values ​​based on the OCT data, the plurality of measurement values ​​including one or more papillary measurement values ​​corresponding to the one or more papillary parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters, inputting the plurality of measurement values ​​into the computational model, and providing a calculation result output from the computational model corresponding to the input of the plurality of measurement values.

[0011] Another exemplary embodiment of the present disclosure is an ophthalmic device that can be used for glaucoma screening, comprising: a storage unit, storing a pre-created calculation model, the calculation model including a linear coupling of one or more papilla parameters representing the thickness of retinal tissue in the optic nerve papilla area and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in the macular area; an OCT data acquisition unit, acquiring optical coherence tomography (OCT) data of the fundus of the examined eye; a measurement value calculation unit, calculating a plurality of measurement values ​​based on the OCT data, the plurality of measurement values ​​including one or more papilla measurement values ​​corresponding to the one or more papilla parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters; and a risk information generation unit, generating glaucoma risk information of the examined eye based on the calculation model and the plurality of measurement values.

[0012] Another exemplary embodiment of the present disclosure is to enable a computer including a processor and a memory to execute a program for glaucoma screening, causing the processor to execute: control of storing a calculation model in the memory, the calculation model including a linear coupling of one or more papillary parameters representing the thickness of retinal tissue in the optic nerve head area and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in the macular area; control of receiving optical coherence tomography (OCT) data of the fundus of the examined eye; processing of calculating multiple measurement values ​​based on the OCT data, the multiple measurement values ​​including one or more papillary measurement values ​​corresponding to the one or more papillary parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters; processing of inputting the multiple measurement values ​​into the calculation model; and processing of providing calculation results output from the calculation model corresponding to the input of the multiple measurement values.

[0013] Another exemplary embodiment of the present disclosure is a computer-readable non-temporary recording medium having recorded thereon a program capable of causing a computer including a processor and a memory to perform glaucoma screening, the program causing the processor to execute: control of storing a calculation model in the memory, the calculation model including a linear coupling of one or more papillary parameters representing the thickness of retinal tissue in an optic nerve head area and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in a macular area; control of receiving optical coherence tomography (OCT) data of the fundus of an eye under examination; processing of calculating a plurality of measurement values ​​based on the OCT data, the plurality of measurement values ​​including one or more papillary measurement values ​​corresponding to the one or more papillary parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters; processing of inputting the plurality of measurement values ​​into the calculation model; and processing of providing calculation results output from the calculation model corresponding to the input of the plurality of measurement values.

[0014] According to an exemplary embodiment of the present disclosure, a glaucoma determination technology suitable for large-scale screening can be provided. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart illustrating steps of a glaucoma screening method according to one exemplary mode of implementation.

[0016] Figure 2 is a flow chart illustrating steps of a glaucoma screening method according to one exemplary mode of implementation.

[0017] Figure 3 is a flowchart for explaining the creation of a computational model and verification of its quality in a glaucoma screening method according to one exemplary embodiment.

[0018] Figure 4A: is a diagram for explaining OCT measurement predictive factors in a glaucoma screening method according to one exemplary embodiment.

[0019] Figure 4B : is a diagram for explaining OCT measurement predictive factors in a glaucoma screening method according to one exemplary embodiment.

[0020] Figure 4C : is a diagram for explaining OCT measurement predictive factors in a glaucoma screening method according to one exemplary embodiment.

[0021] Figure 5 This is a table showing characteristics of cases and controls of development data used for development of a calculation model in a glaucoma screening method according to an exemplary embodiment.

[0022] Figure 6 is a histogram showing case distribution and control distribution of OCT measurement predictors in a glaucoma screening method according to one exemplary embodiment.

[0023] Figure 7 is a table showing three calculation models in a glaucoma screening method according to one exemplary embodiment.

[0024] Figure 8 Receiver operating characteristic curves of three computational models in a glaucoma screening method according to an exemplary embodiment.

[0025] Fig. 9 1 is a table showing the validation results of three calculation models in a glaucoma screening method for explaining one exemplary embodiment.

[0026] Fig.10 1 is a table showing the validation results of three calculation models in a glaucoma screening method for explaining one exemplary embodiment.

[0027] Fig.11 1 is a table showing the validation results of three calculation models in a glaucoma screening method for explaining one exemplary embodiment.

[0028] Fig.12 It is a block diagram showing the configuration of an ophthalmologic apparatus according to one exemplary form of the embodiment. DETAILED DESCRIPTION

[0029] Please refer to the attached Figure 1 Some non-limiting embodiments of the present disclosure are described in detail.

[0030] Any known technology can be combined with the embodiments of the present disclosure. For example, a matter freely selected from the matters recorded in any document cited in this specification can be combined with the embodiments of the present disclosure. In addition, any technology known in any technical field associated with the present disclosure can be combined with the embodiments of the present disclosure. The matters combined with the embodiments of the present disclosure are not limited to these. For example, any technical matter disclosed by the applicant of the present application (for example, any technical matter disclosed by any means in a patent application, paper, website, etc.) can be cited to the present disclosure by reference.

[0031] Two or more illustrative aspects of the present disclosure may be combined at least partially.

[0032] Any of the various functions of the various elements described in the embodiments of the present disclosure is implemented using circuitry or processing circuitry. Circuitry or processing circuitry includes general-purpose processors, special-purpose processors, integrated circuits, CPUs (Central Processing Units), GPUs (Graphics Processing Units), ASICs (Application Specific Integrated Circuits), programmable logic devices (e.g., SPLDs (Simple Programmable Logic Devices), CPLDs (Complex Programmable Logic Devices), FPGAs (Field Programmable GateArray: Field Programmable Gate Array), previous circuit compositions and any combination of these. The processor is considered to include transistors and / or other circuit compositions, processing circuit compositions or circuit compositions. In the present disclosure, circuit compositions, units, devices or similar terms are hardware that performs certain functions or hardware that is programmed to perform certain functions. The hardware may be any of the hardware described in the embodiments of the present disclosure, or may be hardware that is programmed or configured to perform certain functions. In the case where a processor that can be regarded as a certain type of circuit composition is used as hardware, the circuit composition, unit, device or similar terms represent an element that includes a combination of hardware and software, and the software is used to constitute the hardware and / or processor.

[0033] <Overview of implementation>

[0034] One purpose of some exemplary methods of the present disclosure is to propose a glaucoma determination technology suitable for large-scale screening. To this end, the glaucoma screening technology of these exemplary methods focuses on the two aforementioned problems, namely, the reliability problem in glaucoma determination (the first problem) and the difficulty and complexity problem in normal intraocular pressure glaucoma determination (the second problem), and aims to solve or alleviate these problems. Those skilled in the art can understand from the embodiments described below that some exemplary methods of the present disclosure solve or alleviate the first problem and the second problem.

[0035] The embodiments of the present disclosure are not limited to the method of focusing on both of the two problems. For example, the method may focus on only one of the two problems, or may focus on another problem in addition to either or both of the two problems, or may focus on another problem different from the two problems. In other words, the embodiments of the present disclosure may solve or alleviate only one of the two problems, or may solve or alleviate another problem in addition to either or both of the two problems, or may solve or alleviate another problem different from the two problems. Hereinafter, unless otherwise specified, "problem solving" and "problem alleviation" have the same meaning.

[0036] Optical coherence tomography (OCT) is a non-contact, high-speed, and non-invasive imaging method. Although OCT is a relatively new technology, its popularity in the field of ophthalmology has been amazing, and it is now used in many medical institutions including small clinics. Therefore, OCT can be said to be an imaging method suitable for large-scale screening.

[0037] In the embodiment of the present disclosure, OCT is used to generate a fundus cross-sectional image. More specifically, the glaucoma determination of the embodiment of the present disclosure is essentially performed only based on the image obtained using OCT (referred to as OCT image). In contrast, in various known glaucoma determination techniques, not only the image obtained using OCT (referred to as OCT image) but also the background information of the subject, intraocular pressure data, visual field examination data, images obtained using ophthalmic methods other than OCT (slit lamp microscope, fundus camera, scanning laser ophthalmoscope, etc.) are essentially used. In addition, even if it is assumed that there is a glaucoma determination technology that uses only OCT images, it is considered to be different from the glaucoma determination of the embodiment of the present disclosure at least in that it cannot solve the first problem and / or the second problem and does not have the structure for solving these problems.

[0038] The type of OCT used in the embodiments of the present disclosure may be any, but typically may be spectral domain OCT (SD-OCT) or swept source OCT (SS-OCT). In the embodiments described below, spectral domain OCT is used, but those skilled in the art will appreciate that embodiments with the same structure and the same effect can be obtained even if other types of OCT are used.

[0039] In an embodiment of the present disclosure, an algorithm suitable for large-scale screening of glaucoma using OCT is created and verified. To this end, in an embodiment described later, an algorithm (computational model) is created that functions to generate a glaucoma risk score based on OCT images using important indicators selected from various indicators obtained from OCT data of the fundus.

[0040] Next, some methods of glaucoma screening methods that can be implemented using the algorithm are described. These methods are illustrative and are not intended to limit the embodiments of the present disclosure.

[0041] Figure 1 One embodiment of a glaucoma screening method (glaucoma identification method) is shown. In addition, details of each step will be described later by way of non-limiting example.

[0042] In the method of this embodiment, first, a calculation model for obtaining a glaucoma risk score from OCT images is created (step S11 ).

[0043] Step S11 is executed by a computer that acts according to a computational model creation program. The program may include a program that utilizes machine learning technology. For example, development data (learning data, teacher data, training data) and a mathematical model such as a neural network are prepared, and machine learning using the development data is applied to the mathematical model to construct a computational model, wherein the development data includes a plurality of OCT data (or, a plurality of measured values ​​calculated based on a plurality of OCT data) to which a label (glaucoma risk score value) created by a professional doctor or another computational model is attached. Some examples of the method for creating a computational model are described later.

[0044] The calculation model created in step S11 includes parameters related to the optic nerve head (papillary parameters) and parameters related to the macula (macula parameters). That is, the calculation model is a mathematical model configured to receive inputs of measurement values ​​corresponding to the papilla parameters and measurement values ​​corresponding to the macula parameters and output calculation results based on these measurement values.

[0045] The papilla parameter is an arbitrary parameter measured or calculated about the optic nerve head, for example, a parameter indicating the thickness of the retinal tissue in an area set for the optic nerve head (optic nerve head area). The optic nerve head area may be, for example, at least a part of the optic nerve head area, an area around the optic nerve head area, and an area combining these areas. The retinal tissue thickness indicates the size (thickness) of one or more tissues constituting the retina. The retina is histologically divided into the following 10 layers: retinal pigment epithelium, photoreceptor layer, external limiting membrane, external granular layer, external plexiform layer, internal granular layer, internal plexiform layer, ganglion cell layer, nerve fiber layer, and internal limiting membrane. The retinal tissue thickness may be the thickness of one or more of these layers. The papilla parameter is not limited to these examples, and may be, for example, an arbitrary parameter derived from data collected from the fundus area set based on the optic nerve head by OCT scanning, such as cup size (diameter, area, etc.), disc size, rim size, ratio of two sizes, papilla tilt, etc.

[0046] The macular parameter is any parameter measured or calculated with respect to the macula, for example, a parameter indicating the thickness of retinal tissue in an area set for the macula (macula area). The macular area may be, for example, at least a portion of the macula, an area around the macula, or an area combining these areas. The macular parameter is not limited to these examples, and may be, for example, any parameter derived from data collected from a fundus area set based on the macula by OCT scanning, such as macular size (diameter, area, depth, etc.). The retinal tissue corresponding to the macular parameter may be the same as or different from the retinal tissue corresponding to the papillary parameter.

[0047] The calculation model of the present method is expressed as a formula of a predetermined method including one or more papilla parameters and one or more macula parameters as variables. The form of the formula can be arbitrary.

[0048] Some exemplary computational models include linear coupling of one or more papillary parameters, and / or linear coupling of one or more macula parameters. In other words, some exemplary computational models include linear equations that have one or more papillary parameters and / or one or more macula parameters as variables.

[0049] As we all know, N parameters P1, P2, ... N The linear coupling uses coefficients c1, c2, ... c N It is expressed as the following formula: c1P1+c2P2+……+c N P N In addition, there are N variables x1, x2, ... N The linear equation y = f(x) uses coefficients a0, a1, a2, ... a N It is expressed as the following formula: y = a0 + a1x1 + a2x2 + ... + aN x N .

[0050] The formulas representing some exemplary computing models may include formal terms other than the zero-order terms and the first-order terms. For example, the formulas of some exemplary computing models may include polynomials of degree 2 or higher. In addition, the formulas of some exemplary computing models may include mathematical symbols other than the four arithmetic operations (e.g., square roots, differentials, integrals, etc.).

[0051] Three examples of calculation models defined as multivariable linear equations with both the papilla parameter and the macula parameter as variables are shown below (Y(Model 1), Y(Model 2), and Y(Model 3)). Each variable (for example, TSNITlower) will be described later.

[0052] [First calculation model]

[0053] Y(Model1)=14.4305-0.0404×TSNITlower-0.0303×cpRNFLqS-0.0304×cpRNFLqIt-0.0271×cpRNFLq In-0.1424×mGCIPL_IT+1.1427×log_mGCIPL_STvsIT+0.6971×log_mGCIPL_ITvsIN+0.0602×mGCC_IN

[0054] [Second calculation model]

[0055] Y(Model2)=12.6694-0.0329×TSNITlower-0.0344×cpRNFLqS-0.0318×cpRNFLqIt-0.0398×cpRNFLqIn+1.1646×log_mGCIPL_STxvsITx+0.712×log_mGCC_ITxvsINx

[0056] [Third calculation model]

[0057] Y(Model3)=12.5935-0.066×TSNITlower-0.0296×cpRNFLqIn-0.0289×mGCC+0.9845×log_mGCIPL_SNyvsINy+0.5671×log_mGCIPL_STyvsITy+0.6613×log_mGCC_ITyvsINy

[0058] In this method, step S11 (creation of a glaucoma risk score calculation model) can be performed at any time before the evaluation of the eye under examination is performed using the calculation model (steps S12 to S15). The created calculation model is stored in a computer that performs the evaluation of the eye under examination, and / or in a storage device that can be accessed by the computer. The storage device can be, for example, a node of a computer network such as a local area network (LAN), the Internet, or a wide area network (WAN) to which the computer belongs. Alternatively, the storage device can be a peripheral device directly or indirectly connected to the computer.

[0059] In addition, if after this method Figure 2 As described together, the calculation model can also be updated using the data used in the evaluation of the eye to be examined (corresponding to the OCT data generated in step S12 of the present embodiment).

[0060] The above-mentioned exemplary calculation model is defined as a formula that only uses parameters derived from OCT data as variables, but other calculation models may include additional parameters as variables, for example, to improve the calculation quality (accuracy, precision, reproducibility, etc.). Examples of such additional or auxiliary parameters are as follows: (1) parameters other than papillary parameters and macula parameters obtained from OCT data; (2) parameters related to predetermined subtissues of the retina; (3) parameters related to tissues other than the retina (for example, parameters related to the choroid, sclera, vitreous body, canthus, etc.); (4) parameters related to fundus blood vessels (for example, vascular density, etc.); (5) parameters related to the size and / or shape of the area where lesions or damage occur; (6) parameters related to fundus blood flow; (7) parameters related to the background information of the subject (for example, age, gender, race, medical history, treatment history, medication history, family history, etc.); (8) parameters derived from data obtained by ophthalmic methods other than OCT; (9) parameters related to data obtained by ophthalmic examination (for example, intraocular pressure, visual field examination parameters, etc.); (10) parameters related to data obtained by examinations of diagnostic and treatment departments other than ophthalmology (for example, blood pressure, blood test parameters, genetic test parameters, etc.).

[0061] In step S12, OCT scanning is applied to the fundus of the eye to be examined to generate OCT data. The OCT scanning is performed by, for example, a spectral domain type OCT device (OCT scanner).

[0062] The type (type, form, etc.) of the OCT data generated in step S12 is typically determined according to the processing content of step S13. For example, the OCT data may include any one of data collected by OCT scanning (collected data), image data obtained by applying Fourier transform or other imaging processing to the collected data (OCT image data), data obtained during the imaging processing (pre-imaging data), and data obtained by applying a predetermined processing to the OCT image data (processed OCT image data).

[0063] When the collected data is processed to generate OCT data, the processing is executed by a computer operating in accordance with an OCT data generation program. The OCT data generation program may be a part of a program that causes a computer to execute a series of processes of steps S12 to S15.

[0064] The number of OCT scans applied to the fundus of the eye under examination in step S12 may be any number of times more than one. In some examples, multiple OCT scans can be applied to the same area of ​​the fundus of the eye under examination. In this case, an average image (addition average image, etc.) with reduced random noise, an OCT angiography image (OCTA image) depicting fundus blood vessels, OCT blood flow measurement data representing fundus blood flow dynamics, etc. can be generated based on multiple collected data acquired from the same area.

[0065] In other examples, OCT scanning can be applied to two or more different regions of the fundus of the eye to be examined, respectively, and OCT data of the above two or more regions can be generated respectively. For example, OCT scanning of a region including the optic nerve head region (optic nerve head region scanning) and OCT scanning of a region including the macula region (macula region scanning) can be performed separately. In this case, OCT data of the optic nerve head region can be generated based on the optic nerve head region scanning, and OCT data of the macula region can be generated based on the macula region scanning.

[0066] On the other hand, a single OCT scan (wide-area scan) may be applied to a region including both the optic nerve head region and the macula region. In this case, OCT data of both the optic nerve head region and the macula region can be generated based on data collected by the wide-area scan.

[0067] In the next step S13, based on the OCT data generated in step S12, the data to be input to the calculation model created in step S11 is obtained. As described above, the calculation model of this embodiment includes one or more papilla parameters and one or more macula parameters as variables. Therefore, in step S13, the papilla measurement value corresponding to each papilla parameter and the macula measurement value corresponding to each macula parameter are calculated. In addition, measurement values ​​other than these values ​​can be calculated additionally or auxiliary.

[0068] Step S13 is executed by a computer operating according to a measurement value calculation program. The measurement value calculation program may be a part of a program that causes a computer to execute a series of processes of steps S12 to S15.

[0069] In a method that also uses other information (e.g., the additional or auxiliary parameters described above) different from the measured values ​​calculated based on the OCT data, the computer executes a process of acquiring the information. For example, the computer may execute a process of accessing a medical information database (e.g., an electronic medical record system, a medical image archiving system, a medical information sharing system, etc.) to acquire the information of the subject.

[0070] In the next step S14, each papillary measurement value and each macular measurement value calculated in step S13 are input into the calculation model created in step S11. Each papillary measurement value is substituted into the corresponding papillary parameter (variable), and each macular measurement value is substituted into the corresponding macular parameter (variable).

[0071] In some exemplary embodiments, in addition to the measurement value calculated in step S13, additional information is input into the calculation model. In some exemplary embodiments, the result output from the calculation model based on the input of the measurement value calculated in step S13 is adjusted (corrected) based on the additional information.

[0072] Step S14 is executed by a computer operating according to a measured value input program. The measured value input program may be a part of a program that causes a computer to execute a series of processes of steps S12 to S15.

[0073] The computer obtains a calculation result (glaucoma risk score) based on the calculation model created in step S11 and the measured values ​​(and other information) input into the calculation model in step S14. In step S15, the obtained glaucoma risk score is provided.

[0074] The destination of providing the glaucoma risk score is predetermined or determined each time a process is performed. Examples of the destination include a user, another computer, a medical system, an ophthalmic device, a storage device, etc. In some exemplary embodiments, the glaucoma risk score may be displayed on a display device, sent to a doctor's terminal, sent to a subject's terminal, sent to a server, sent to a database, sent to an examination device, sent to a treatment device, or stored in a storage device.

[0075] Step S15 is executed by a computer operating in accordance with an information providing program. The information providing program may be a part of a program that causes a computer to execute a series of processes of steps S12 to S15.

[0076] Figure 1The method shown is configured to obtain a glaucoma risk score based on the papillary measurement value and the macula measurement value, wherein the papillary measurement value and the macula measurement value are calculated based on the OCT data of the fundus of the examined eye, and the intraocular pressure of the examined eye is not required to evaluate glaucoma. Therefore, this method can determine normal intraocular pressure glaucoma. Moreover, although the details will be described later, according to this method, glaucoma can be determined with high reliability. Therefore, this method provides a glaucoma determination technology suitable for large-scale screening.

[0077] As described above, some exemplary methods may be configured to update the computational model using data used in the evaluation of the eye under examination. Figure 2 One aspect of such a glaucoma screening method (glaucoma determination method) will be described.

[0078] Steps S21 (creation of calculation model), S22 (generation of OCT data), S23 (calculation of papillary measurement value and macula measurement value) and step S24 (input of measurement value into calculation model) can be respectively performed in the same manner as Figure 1 The same method as steps S11, S12, S13 and S14 is performed. However, another method may also be applied.

[0079] In the next step S25 , the computer acquires a glaucoma risk score based on the calculation model created in step S21 and the measured values ​​input into the calculation model in step S24 , and adjusts the coefficients of the calculation model based on the glaucoma risk score.

[0080] Typically, the calculation model of the present method is defined as a multivariable linear equation with the papilla parameter and the macula parameter as variables. Each term of the multivariable linear equation is the product of the coefficient and the variable. In the present method, for example, the papilla measurement value and the macula measurement value calculated in step S23 are added to the development data used in the calculation model creation in step S21 to update the development data. The calculation model coefficients can be adjusted by applying machine learning using the updated development data to the calculation model.

[0081] Step S25 is executed by a computer that operates according to a calculation model update program. The calculation model update program may be a part of a program that causes the computer to execute a series of processes of steps S22 to S25 (or steps S21 to S25). In addition, the calculation model update program may be the same as the calculation model creation program used in step S21.

[0082] In the next step S26, each papilla measurement value and each macula measurement value calculated in step S23 are input to the calculation model updated in step S25.

[0083] The computer derives a glaucoma risk score based on the new calculation model obtained in step S25 and the measured values ​​(and other information) input into the new calculation model in step S26. In step S27, the glaucoma risk score thus obtained is provided. The provision of the glaucoma risk score is, for example, in the same manner as Figure 1 The same method as step S15 is performed. However, another method may also be applied.

[0084] Figure 2 The method shown is the same as Figure 1 In the same manner as in the method of the present invention, normal pressure glaucoma can be determined and glaucoma can be determined with high reliability, so a glaucoma determination technology suitable for large-scale screening can be provided. In addition, the present method can use the OCT data of the fundus of the eye to be examined as the object of glaucoma evaluation for updating the calculation model, so the quality of the calculation model can be improved at any time.

[0085] This method uses the subject's data for updating the calculation model, so it is considered preferable to perform processing for protecting the subject's personal information. In some exemplary methods, items equivalent to personal information (personal information items) are predefined in advance. Examples of personal information items include subject identifiers (subject IDs), names, addresses, public personal numbers, background information, etc. At the stage of processing specific subject information, the computer excludes information equivalent to each personal information item from all information about the subject, and only uses the remaining information (including OCT data and / or measurement values) for updating the calculation model. Thereby, the level of personal information protection is improved. Such processing is performed by a computer that acts in accordance with a personal information protection program. The personal information protection program can be part of the aforementioned calculation model update program.

[0086] <Exemplary method>

[0087] The above is a description of the outline of the embodiment. Some exemplary forms of such an embodiment will be described below.

[0088] First, the creation of a computational model used in a glaucoma screening method in an exemplary manner will be described together with an example performed by the inventors. Figure 3 The flowchart shows an overview of the steps for creating the computational model used in the glaucoma screening method of the present invention.

[0089] Step S31, which is the first step of the process, is the creation of development data. The development data used in this example includes data from 11,487 eyes of 7,572 participants. The average age of the participants was 51.3 years old (±10.0 years old), and the age range was 35 to 74 years old. Each participant was subjected to medical record confirmation, fundus photography using a digital fundus camera, visual field examination using an FDT (Frequency Doubling Technology) perimeter, automatic OCT measurement using spectral domain OCT, and axial length measurement using an optical axial length meter. That is, the data of each participant includes medical record data, fundus photographs, visual field data, fundus OCT data, and axial length data. In addition, considering the detection of normal-tension glaucoma, the development data in this example does not include intraocular pressure values.

[0090] The next step S32 is the selection of qualified cases and qualified controls. In the development of the glaucoma risk score in this example, a case-control design (matched case-control design) was adopted. Each case is a participant who was diagnosed with glaucoma by an ophthalmologist based on the results of visual field examination and fundus examination. In addition, each control is a participant who has no history of glaucoma and has not been diagnosed with glaucoma by fundus photography or FDT visual field examination.

[0091] The next step S33 is to match the background conditions between the cases and controls. In this example, a control is randomly selected for each case by matching the gender (male or female) and the age (5 age groups). The final development data set used for analysis includes 191 cases with 284 eyes and 277 controls with 287 eyes.

[0092] The next step S34 is to develop a calculation model using the development data obtained in step S33. The details of the method (statistical analysis method) for developing the calculation model performed in this example will be described later with some examples.

[0093] The next step S35 is to create a verification data set. Verification data is data different from development data. In this example, the development data set is obtained from participants who received a health check in 2016, and the verification data set is obtained from participants who received a health check in 2018. The verification data used in this example is OCT data of 9720 eyes. In addition, the verification data in this example does not include intraocular pressure values.

[0094] The next step S36 is to verify the calculation model of the positive hit rate (Positive Predictive Value). In addition, the positive hit rate is defined as the value of the probability that it is a true positive when it is determined to be positive. In some embodiments, it is defined as the probability that it is actually positive when the glaucoma risk score is a "high" score described later.

[0095] In this example, the calculation model obtained in step S34 was applied to each verification data (OCT data of each of 9720 eyes). Furthermore, for verification, the score range was divided into three levels of "low", "medium" and "high", and 723 eyes in each score level were randomly extracted. In addition, the inventors independently evaluated the OCT reports of each extracted eye and determined which of the four options "normal", "glaucoma", "eye disease other than glaucoma" and "cannot be determined (further precise ophthalmological examination is required)" each eye corresponds to.

[0096] The verification of this step was carried out using the following five OCT parameters in order to determine the changes in glaucomatous properties: (1) the average thickness of the retinal nerve fiber layer (RNFL) on a predetermined circle (a circular area set around the optic nerve head); (2) the presence or absence of local thinning of the RNFL and the height difference between the double humps in the TSNIT curve; (3) a quadrant RNFL thickness chart and a clock hour RNFL thickness chart; (4) a retinal thickness map, a ganglion cell layer ++ thickness map (GCL++ thickness map) and a GCL+ thickness map in the macular area, and a RNFL thickness map in the optic nerve head area; (5) a GCL++ thickness deviation map, a GCL+ thickness deviation map and a RNFL thickness deviation map. Among them, GCL+ represents a composite layer of the ganglion cell layer and the inner plexiform layer, also known as GCIPL. In addition, GCL++ represents a composite layer of the retinal nerve fiber layer, the ganglion cell layer and the inner plexiform layer, also known as the ganglion cell complex (GCC).

[0097] The TSNIT curve is data representing the RNFL thickness distribution in a predetermined circular area around the optic nerve head. More specifically, the TSNIT curve is RNFL thickness distribution data expressed using a two-dimensional coordinate system developed by a first coordinate axis representing the position (angle) in the circular area and a second coordinate axis representing the RNFL thickness. The reference position (angle zero) of the first coordinate axis of the TSNIT curve is the ear side (T) and the positive direction of the first coordinate axis is the rotation direction that passes through the ear side (T), the upper side (S), the nose side (N) and the lower side (I) in sequence and returns to the ear side (T).

[0098] Glaucomatous RNFL defects often present an arcuate defect with the borderline of the ear. For OCT reports with inconsistent evaluations between the inventors, the evaluation was repeated until the evaluation of all personnel was consistent. All personnel (ophthalmologists) participating in the evaluation of the validation data performed the evaluation in an environment where the glaucoma score was blinded and no information other than the OCT report could be accessed.

[0099] The next step S37 is to verify the calculation model of sensitivity and specificity. The "sensitivity" is defined as the value of the probability of correctly judging a case that should be judged as positive as positive, and the "specificity" is defined as the value of correctly judging a case that should be judged as negative as negative.

[0100] For the verification of this step, a follow-up survey (follow-up) was conducted for participants who were suspected of glaucoma based on fundus images (fundus photographs) among the participants created by the development data of step S31. In this follow-up survey, 129 eyes of 66 participants were additionally examined using a Humphrey perimeter (HFA (registered trademark)) using a standard procedure based on the 24-2 Swedish Interactive Threshold Algorithm. The inventors evaluated the results of the Humphrey perimetry and applied a computational model to the data set of the 129 eyes of the follow-up survey subjects to verify the consistency of the evaluation results of the Humphrey perimetry.

[0101] From the above description Figure 3 The steps of the flowchart (steps S31 to S37) are used to obtain the verified calculation model of this example.

[0102] Next, the OCT measurement predictors used in the creation of the calculation model of this example are described. The OCT measurement predictors are predictors generated from data obtained by OCT scanning.

[0103] In this example, inner retinal layer thickness is obtained using OCT scanning based on a 3D wide scan protocol. The 3D wide scan protocol is a protocol for applying OCT scanning to a wide 3D region including both the retinal region and the optic nerve head region. The inner retina refers to the range from the inner limiting membrane to the outer limiting membrane. Layer thickness distribution data in the retinal region and layer thickness distribution data in the optic nerve head region are derived from the 3D data set obtained by the OCT scanning using the 3D wide scan protocol.

[0104] In another example, an OCT scan of an area including a retinal area and an OCT scan of an area including an optic nerve head area may be performed separately, and layer thickness distribution data in the retinal area may be generated from the former, and layer thickness distribution data in the optic nerve head area may be generated from the latter. In addition, the OCT scan of the area including the retinal area may be an OCT scan of the entire area, or may be a combination of OCT scans of two or more partial areas of the area.

[0105] The inventors reconfigured the grid of the optic nerve head area (segmentation) and the macular area and created variables as promising predictors. Figure 4A , Figure 4B and Figure 4C Select from the 312 variables defined in the method shown.

[0106] In the first step, layer thickness data related to the optic nerve head region (peripapillary RNFL (cpRNFL) thickness data) and several types of layer thickness data related to the macula region are obtained.

[0107] Figure 4A FIG. 41A on the left of the two figures shown shows 12 segments set in the optic nerve head region. The optic nerve head region is a circular region (circumferential region) centered on the optic nerve head center and having a predetermined radius. The optic nerve head region is divided into 12 segments (arc-shaped regions) having an equal central angle of 30 degrees. The size (diameter) of the optic nerve head region may be, for example, a default value, or may be a value set or selected according to the size (diameter) of the optic nerve head.

[0108] In this example, the average thickness of the RNFL was calculated for each of the 12 segments of the optic nerve head region based on a three-dimensional data set collected by OCT scanning with a three-dimensional wide scan protocol.

[0109] In addition, those skilled in the art will appreciate that the names of the 12 segments of the optic nerve head region shown in FIG. 41A are defined by superior (S), inferior (I), auricular (T), nasal (N), and combinations thereof (the same applies below).

[0110] Figure 4B Figure 42A on the left of the three figures shown shows a grid that divides the macular area into 10×10=100 segments. The macular area is a square area with a predetermined size centered on the fovea, and is divided into 100 segments (square areas) of equal size. The size (diameter) of the macular area may be, for example, a default value, or a value set or selected according to the eye to be examined.

[0111] In this example, for each of the 100 segments of the macular area, the average thickness of the macular RNFL (mRNFL), the average thickness of the macular ganglion cell layer-inner plexiform layer (mGCIPL), and the average thickness of the macular ganglion cell complex (mGCC) were calculated based on the three-dimensional data set obtained by OCT scanning using a three-dimensional wide scan protocol.

[0112] In the second step, variables related to the optic nerve head area are determined. In this example, the 12 segments based on the optic nerve head area (cpRNFL) are set Figure 4A The five regions (segments) qT, qS, qN, qIn, and qIt shown in FIG41B on the right side of FIG41A are shown in FIG41B . Furthermore, five variables cpRNFLqT, cpRNFLqS, cpRNFLqN, cpRNFLqIn, and cpRNFLqIt corresponding to the five regions qT, qS, qN, qIn, and qIt are created.

[0113] Among them, cpRNFLqT is the ear-side quadrant, cpRNFLqS is the upper quadrant, and cpRNFLqN is the nasal quadrant. In contrast, the lower quadrant is divided into two areas qIn and qIt based on the analysis results described later. The division of the lower quadrant is adopted because the RNFL thickness on the lower ear side is considered to be highly useful in the detection of glaucoma.

[0114] In the third step, variables related to the macular area are determined. In this example, the segmentation (division into 10×10 segments) of the above three layer thickness distributions (mRNFL thickness distribution, mGCIPL thickness distribution, and mGCC thickness distribution) is first reconstructed as Figure 4B The variables of the macular region are defined based on the four regions ST, SN, IT, and IN shown in FIG. 42A on the left side of FIG. The variables of this example are used in the first example Y (Model 1) of the calculation model described above.

[0115] In addition, the variables are defined by excluding the peripheral part of the macular area from the 100 segments (10×10 segments) of the macular area. In this example, 64 segments belonging to the range of two segments from the outer edge (periphery) of the macular area are excluded from the 100 segments of the macular area, and the remaining 36 segments are focused on. Then, using the above 36 segments, the variables are defined as Figure 4B Variables corresponding to the four regions STx, SNx, ITx, and INx shown in FIG42B at the center. The variables of this example are used, for example, in the second example Y (Model2) of the calculation model described above.

[0116] Furthermore, the variables are defined by excluding both the peripheral part and the central part of the macular area from the 100 segments of the macular area. In this example, 64 segments belonging to the range of two segments from the outer edge of the macular area and 4 segments belonging to the range of one segment from the center of the macular area are excluded, and the remaining 32 segments are focused on. Then, using the above 32 segments, the variables are defined as Figure 4B The variables corresponding to the four regions STy, SNy, ITy, and INy shown in FIG42C on the right side of FIG42A are shown in FIG42B. The variables of this example are used in the third example Y (Model 3) of the calculation model described above.

[0117] Each variable defined for the macular region is the average value of the thickness (in micrometers) of the corresponding segment. For example, Figure 4A The RNFL thickness in the segment qS (cpRNFLqS) shown in FIG. 41B is defined as the average value of the RNFL thickness in the three segments S, ST, and SN in FIG. 41A corresponding to the segment qS: cpRNFLqS = (cpRNFL_S + cpRNFL_ST + cpRNFL_SN) / 3 (micrometers). Similarly, Figure 4B The GCC thickness (mGCC_SN) in the segment SN shown in Figure 42A is defined as the average value of the GCC thickness in 25 segments belonging to the segment SN among the 100 segments divided by the grid: mGCC_SN = (mGCC01_01+01_02+…+05_04+05_05) / 25 (micrometers).

[0118] In the fourth step, a variable is created that expresses the difference in layer thickness between two regions. In this example, a variable that expresses the difference between the upper region and the lower region and a variable that expresses the difference between the region on the ear side and the region on the nose side are created. These variables do not conform to a normal distribution. In addition, in order to avoid the calculated value being zero when the layer thickness of the two regions is equal, these variables are defined by applying a natural logarithm transformation to the real number obtained by adding 1 to the absolute value of the difference in the layer thickness values ​​of the two regions. For example, Figure 4BThe variable of the difference in mGCIPL thickness between the two regions ST and IT is defined as follows: log_mGCIPL_STvsIT=log e (|mGCIPL_ST-mGCIPL_IT|+1).

[0119] In the final fifth step, considering that the difference in cpRNFL thickness between the upper region and the lower region is large but the thickness value of the thin region is not very important, a variable TSNITlower corresponding to the layer thickness corresponding to the lower peak of the two peaks of the double hump presented by the TSNIT curve is defined. Figure 4C An example of the variable TSNITlower is shown in . Generally, a double hump is a sharp peak near the upper side (S) and a sharp peak near the lower side (I), and the variable TSNITlower is defined as the smaller value of the two layer thickness values ​​at the apex positions of the two sharp peaks.

[0120] Next, the statistical analysis used to create the calculation model of this example will be described.

[0121] In this example, all variables obtained from the OCT measurement values ​​were treated as continuous variables. In the statistical analysis of this example, the optimal prediction model in the development data set was selected by performing backward elimination using a multivariate logistic regression model using the variable reduction method with a P value of 0.1. Thus, in this example, the three calculation models Y(Model1), Y(Model2) and Y(Model3) mentioned above were obtained.

[0122] The first calculation model Y (Model 1) includes all of the values ​​associated with cpRNFL, the variables associated with mRNFL, the variables associated with mGCIPL, and the variables associated with mGCC as promising predictors. Among them, each variable associated with the macula is defined for the entirety of the 10×10 macular grid (the entirety of the macular area) (refer to Figure 4B Regarding variables associated with the macula, the second calculation model Y (Model2) uses variables defined in an area excluding the peripheral portion from the macula area (see FIG. 42B ), and the third calculation model Y (Model3) uses variables defined in an area excluding both the peripheral portion and the central portion from the macula area (see FIG. 42C ).

[0123] In addition, the three calculation models mentioned above all include the average layer thickness in the entire defined area (cpRNFL, mRNFL, mGCIPL and mGCC), a variable associated with the layer thickness difference between two areas (e.g., log_mGCIPL_STvsIT), and a variable (TSNITlower) associated with the peak of the TSNIT curve graph representing the layer thickness distribution around the optic nerve head (cpRNFL).

[0124] The effectiveness of the statistical analysis performed in this example was estimated and evaluated based on the area under the receiver operating characteristic curve (AUC-ROC).

[0125] In the validation step of this example, the risk score is calculated using the following two equations (Equation 1 and Equation 2) using the intercepts (slices) and β (slope, regression coefficient) obtained from the three regression models in the development step.

[0126] [Formula 1]

[0127]

[0128] [Formula 2]

[0129] Glaucoma screening score=round(p×100)

[0130] Here, round(X) is a function that performs mantissa processing (rounding) of X, and represents the integer closest to X.

[0131] According to the glaucoma risk score defined in this way, the scores of all participants measured by fundus OCT are in the range of 0 to 100. That is, the value range of the glaucoma risk score in this example is 0 to 100. In addition, the larger the value of the glaucoma risk score in this example, the higher the possibility of glaucoma.

[0132] Based on the score distribution, the population was classified into the following three groups: a low (low risk) group with a score range of 0 to 49; a middle (medium risk) group with a score range of 50 to 89; and a high (high risk) group with a score range of 90 to 100. Furthermore, for each group, the positive hit rate (PPV) for screening the necessity of further detailed ophthalmic examinations was calculated. In addition, the diagnostic results obtained from the subset with additional enhancement by Humphrey perimetry were used to evaluate the sensitivity and specificity.

[0133] In order to clarify the effect of axial length on glaucoma screening in this case, the improvement effect of information related to axial length on risk score was studied. In addition, in order to confirm the accuracy of distinguishing retinal thinning caused by glaucoma from retinal thinning caused by high myopia, axial length was estimated based on the angle (direction) of the double hump of the RNFL around the optic nerve head.

[0134] To this end, first, the original value (measured value) of the axial length is independently added to the third calculation model Y (Model3), and then the axial length is added to the three calculation models Y (Model1), Y (Model2) and Y (Model3), and the variable selection in the aforementioned logistic regression is implemented.

[0135] The results of this statistical analysis are explained below. Figure 5 The characteristics of the cases and controls in the development data used in this example are shown in FIG.

[0136] exist Figure 5 In the figure, cpRNFL denotes peripapillary retinal nerve fiber layer, mRNFL denotes macular retinal nerve fiber layer, mGCIPL denotes macular ganglion cell layer-inner plexiform layer, and mGCC denotes macular ganglion cell layer complex (retinal nerve fiber layer-macular ganglion cell layer-inner plexiform layer in the macular area).

[0137] In addition, cpRNFLqIn represents the cpRNFL thickness in the inferior nasal quadrant, cpRNFLqIt represents the cpRNFL thickness in the inferior auricular quadrant, and cpRNFLqS represents the cpRNFL thickness in the superior quadrant.

[0138] In addition, log_mGCIPL_ITvsIN represents the logarithmic transformation of the mGCIPL thickness difference between the lower ear quadrant and the lower nasal quadrant as a real number (the first calculation model), log_mGCIPL_SNyvsINy represents the logarithmic transformation of the mGCIPL thickness difference between the upper nasal quadrant and the lower nasal quadrant as a real number (the third calculation model), log_mGCIPL_STxvsITx represents the logarithmic transformation of the mGCIPL thickness difference between the upper ear quadrant and the lower ear quadrant as a real number (the second calculation model), and log_mGCIPL_STvsIT represents the logarithmic transformation of the mGCIPL thickness difference between the upper ear quadrant and the lower ear quadrant as a real number (the third calculation model). The difference in mGCIPL thickness between the lower ear quadrants is transformed as a logarithmic transformation of the real number (first calculation model), log_mGCIPL_STyvsITy represents the logarithmic transformation of the mGCIPL thickness difference between the upper ear quadrant and the lower ear quadrant (third calculation model), log_mGCC_ITxvsINx represents the logarithmic transformation of the mGCC thickness difference between the lower ear quadrant and the lower nasal quadrant (second calculation model), and log_mGCC_ITyvsINy represents the logarithmic transformation of the mGCC thickness difference between the lower ear quadrant and the lower nasal quadrant (third calculation model).

[0139] In addition, mGCC_IN represents the mGCC thickness in the inferior nasal quadrant (first calculation model), and mGCIPL_IT represents the mGCIPL thickness in the inferior auricular quadrant.

[0140] In addition, TSNITlower represents the cpRNFL thickness at the lower peak (peak) of the double hump of the TSNIT curve.

[0141] Figure 5 Each value shown is a numerical value (% in parentheses) or mean ± standard deviation. In addition, the cases are participants who were confirmed to have glaucoma by ophthalmologists based on a precise ophthalmological examination, and the controls are participants who were confirmed by ophthalmologists to have no glaucoma in fundus photography and FDT visual field examination. The P value is the value calculated based on the chi-square test for age and the value calculated based on the t-test for continuous variables.

[0142] In the case group, the cpRNFL was significantly thinner than that in the normal control group. In addition, the overall average values ​​of the 10×10 grids for mRNFL, mGCIPL, and mGCC were also significantly thinner than those in the normal control group. The Spearman rank correlation coefficient of cpRNFL thickness between the left and right eyes was 0.85 (P<0.01) in the control group and 0.24 (P<0.01) in the case group.

[0143] Figure 6 Show about Figure 5The case distribution and control distribution of each OCT measurement predictor. For any OCT measurement predictor, the histogram with a higher peak value in the two histograms is equivalent to the distribution of the normal control group, and the histogram with a lower peak value is equivalent to the distribution of the case group.

[0144] exist Figure 7 The diagnostic model for glaucoma screening developed by this example is shown in FIG. Figure 7 The three calculation models Y (Model1), Y (Model2) and Y (Model3) show the slice (Intercept) information and retinal thickness related prediction factor information. Figure 7 In the table, B represents slope (β, regression coefficient), SE represents standard error, OR represents odds ratio, CI represents confidence interval, and AUC-ROC represents area under the receiver operating characteristic curve.

[0145] After selecting the variables of the logistic regression model, negative regression coefficients were generally obtained for the predictors of the layer thickness (cpRNFLqS, cpRNFLqIt, cpRNFLqIn, mGCIPL_IT, mGCC, and TSNITlower) of each region. This result means that the possibility of suffering from glaucoma is low when the retinal layer of the selected region is thick.

[0146] In contrast, the predictors for the difference between the upper and lower regions (log_mGCIPL_STvsIT, log_mGCIPL_STxvsITx, log_mGCIPL_SNyvsINy, log_mGCIPL_STyvsITy) and the difference between the retinal layer thickness of the ear side region and the nasal side region (log_mGCIPL_ITvsIN, log_mGCC_ITxvsINx, log_mGCC_ITyvsINy) all obtained positive regression coefficients. This result means that the possibility of suffering from glaucoma is high when the difference in retinal layer thickness between the upper and lower regions and / or the left and right regions is large.

[0147] In addition, the predicted values ​​of the areas under the receiver operating characteristic curves of the three calculation models Y(Model1), Y(Model2) and Y(Model3) are all about 0.97 (see also Figure 8 The results show that there is no big difference between the three calculation models mentioned above.

[0148] The inventors applied the glaucoma risk score constructed in this way to the validation data set (9720 eyes of 6006 people). The above-mentioned [Formula 1] and [Formula 2] were used to calculate the glaucoma risk score. Specifically, the glaucoma risk score can be calculated using three calculation models Y (Model 1), Y (Model 2) and Y (Model 3).

[0149] Reference Fig. 9 The validation results of three computational models related to positive hit rate are described. Fig. 9 The distribution of glaucoma risk scores obtained from the three computational models and the positive hit rate (PPV) associated with the need for further careful eye examination are shown.

[0150] exist Fig. 9 The results of classifying 9720 risk scores obtained by applying each calculation model (Model1; Model2; Model3) to 9720 eyes of the validation data set into the above three levels of "Low", "Middle" and "High" are shown in FIG. Accordingly, the high-risk group (risk score of 90 or more) is equivalent to 6.1% of the total in the first calculation model Y (Model1), 6.2% in the second calculation model Y (Model2), and 6.2% in the third calculation model Y (Model3).

[0151] In order to verify the positive hit rate (and negative hit rate), two-thirds of the high-risk group, one-fifth of the medium-risk group, and 2% of the low-risk group were randomly selected from all the subjects, and four ophthalmologists made judgments based on the OCT reports of each selected subject. As a result, according to the first, second, and third calculation models, the positive hit rates of screening in the high-risk group were 80.7%, 83.3%, and 90.8%, respectively. In addition, the negative hit rates in the low-risk group were 87.9%, 88.4%, and 88.2%, respectively. These results show that among the three calculation models, the third calculation model Y (Model 3) has the best glaucoma screening accuracy.

[0152] Next, refer to Fig.10 and Fig.11 The verification results of the third calculation model Y (Model 3) related to sensitivity and specificity are described. Fig.10 The results of SD-OCT examination and Humphrey perimetry performed on participants (129 eyes) who were suspected of glaucoma by reading fundus photographs among the participants who developed data were shown, and comparison of these examination results with the glaucoma risk score of the third calculation model. Fig.11 Shows the sensitivity and specificity of the risk score in glaucoma screening using the third computational model.

[0153] like Fig.10 As shown, according to the third calculation model, the number of subjects corresponding to the high risk score (90 or more) is 67. And, 53 of these 67 eyes were diagnosed with glaucoma, 10 eyes were judged to be undiagnosable by OCT and Humphrey perimetry, and 4 eyes were judged to be undetectable (normal).

[0154] In addition, if Fig.11 As shown, if subjects with diseases other than glaucoma are excluded, the sensitivity of the risk score using the third calculation model is 85% (85.4%) and the specificity is 91% (91.3%).

[0155] In the false-positive group, two cases showed anterior visual field glaucoma (PPG), and in two cases, the optic nerve could not be tracked during OCT scanning. In the false-negative group, two cases showed mild thinning of the retinal nerve fiber layer defect (NFLD), two cases showed narrow-width NFLD, and three cases showed diffuse thinning of the retinal layer caused by myopia. In addition, two cases were diagnosed with glaucoma, but the scores of these two cases were 89 and 88, respectively, which were very close to the threshold of 90 between high-risk and medium-risk scores.

[0156] The Spearman rank correlation coefficients between the risk score of the third calculation model and the mean deviation and pattern standard deviation of Humphrey perimetry were -0.503 (P < 0.001) and 0.618 (P < 0.001). In the cases with high risk scores and confirmed glaucoma, the mean deviation and pattern standard deviation (mean ± standard deviation) were -6.80 ± 5.91 and 6.42 ± 4.64, respectively. In the 9 cases with low risk scores or medium risk scores (< 90) and confirmed glaucoma (i.e., false negative cases), the mean deviation and pattern standard deviation (mean ± standard deviation) were -1.96 ± 1.53 and 3.02 ± 0.76, respectively. The non-accidental difference between the two groups was confirmed by t-test.

[0157] The inventors studied the effect of axial length on risk score improvement. First, the original value (measured value) of axial length was added independently in the third calculation model. As a result, the odds ratio (95% confidence interval) of axial length was 1.30 (95% confidence interval: 1.02-1.65). However, no improvement in accuracy was found.

[0158] Furthermore, the value of axial length was added to the variable related to OCT, and variable selection in logistic regression was performed for the first, second, and third calculation models. As a result, the odds ratios of axial length in the first, second, and third calculation models were 1.52 (95% confidence interval: 1.21-1.91), 1.56 (95% confidence interval: 1.32-2.10), and 1.52 (95% confidence interval: 1.20-1.93), respectively. In this case, no improvement in accuracy was found.

[0159] Furthermore, the estimated axial length angle of the double hump presented by the peripapillary RNFL was not selected as a non-contingent variable in any of the variable selections in the three computational models.

[0160] The glaucoma risk score of the present embodiment described above is obtained using SD-OCT and can be used in large-scale glaucoma screening on a population basis.

[0161] The development of the glaucoma risk score of this method refers to the diagnostic logic of glaucoma specialists. This diagnostic logic includes the following seven aspects: (1) the absolute value of the thickness of each layer in the entire area, or the absolute value of the thickness of each layer in each of the multiple areas; (2) the difference in the thickness of the RNFL around the optic nerve head in the up-down direction; (3) the difference in the thickness of each layer in the macular area in the up-down direction and the difference in the thickness of each layer in the macular area in the left-right direction; (4) the presence of a wedge-shaped optic nerve fiber layer defect (NFLD) in the local area around the optic nerve head; (5) the difference in the thickness distribution (TSNIT curve) of the RNFL around the optic nerve head in the double hump pattern; (6) the layer thickness value of the side with the lower peak in the double hump pattern of the RNFL around the optic nerve head; (7) the axial length value estimated from the angle (direction) of the double hump pattern of the RNFL around the optic nerve head.

[0162] In this method, the following three modes are studied for the macular area: (A) The first mode using data of the entire macular area ( Figure 4B 42A, first calculation model); (B) a second mode using data of an area excluding the peripheral portion of the macular area in order to avoid the influence of relatively large blood vessels in the fundus (FIG. 42B, second calculation model); and (C) a third mode using data of an annular area excluding the peripheral portion of the macular area in the same way as the second mode and also excluding the central portion where abnormalities do not occur (or occur less frequently) in early glaucoma (FIG. 42C, third calculation model).

[0163] In any of the three calculation models of this method, the odds ratio of the variables related to the upper and lower differences and the left and right differences of each layer became the most influential indicator of glaucoma. In addition, among the three calculation models, the model that exerted the highest sensitivity and specificity in risk assessment using only SD-OCT data was the third calculation model.

[0164] In the verification step of the present method, two steps are applied in order to consider the following two purposes. The first purpose is to develop a score that can automatically detect glaucoma using OCT, and the second purpose is to screen for glaucoma. Therefore, in the first verification, a glaucoma specialist interprets the OCT data, and in the second verification, an evaluation of the glaucoma diagnosis using a Humphrey perimeter is performed. In addition, considering that it is important to select cases with equally separated risk scores and to correctly judge the subtle changes in early glaucoma in large-scale screening, additional cases that are difficult to diagnose as glaucoma are added. The verification work of the present method designed in this way assumes the situation in an actual group examination (based on group large-scale screening) in which an ophthalmologist judges the necessity of further detailed glaucoma examinations in the hospital.

[0165] Furthermore, for the purpose of glaucoma screening, in this method, a panel of experts was formed to blindly test the risk score. In fact, as mentioned above, the necessary decision was made by the panel of four ophthalmologists.

[0166] In order to verify the accuracy of the risk score using real data, cases with low image quality data were intentionally included in the data set used for verification of this method. The reasons for the low image quality are poor fixation, blinking, eye movement, small pupils, etc. In some of these cases, it was difficult to grasp the glaucoma observation results due to artifacts, segmentation errors, optic nerve tracking errors, etc., but 4 ophthalmologists studied their data until they reached a consensus and made the final decision. The risk score of this method obtained in this way is well correlated with the judgment required for the precise examination of glaucoma.

[0167] In addition, in this method, the validity of risk scoring was studied using Humphrey perimetry and ophthalmic diagnosis based on OCT reports, and it was confirmed that it has high performance and high accuracy with a sensitivity of 85.4% and a specificity of 91.3%.

[0168] In this method, 4 cases with no visual field abnormalities but high risk scores were considered false positive cases. Some false positive cases were judged to be likely to be anterior visual field glaucoma accompanied by retinal optic nerve fiber layer defects detected by OCT. This method is mainly aimed at detecting early and mid-stage glaucoma, but it is believed that it can also be used to identify anterior visual field glaucoma.

[0169] Nine glaucoma cases showed low risk scores and were false negative cases. Through the study of these false negative cases, we found that the risk score tends to decrease when the thinning of the RNFL is mild, the width of the retinal nerve fiber layer defect is narrow, or diffuse thinning of the retinal layer caused by myopia is found.

[0170] In recent years, research on visualization of artificial intelligence and deep learning models based on algorithms is progressing actively, but glaucoma determination technology suitable for large-scale glaucoma screening has not yet been put into practical use due to the black box nature of the algorithm and the complexity of installation in large-scale screening. As mentioned above, the algorithm of this method can not only detect glaucoma with high reliability but also detect normal pressure glaucoma. It can also provide glaucoma risk scores by installing simple calculation software on existing OCT devices or computers, so it can be said that it is suitable for the practical use of large-scale glaucoma screening.

[0171] Moreover, in this method, the cutoff value (threshold used to divide the range of risk score according to the degree of risk) can be adjusted or set according to the prevalence rate of the group of subjects in the large-scale screening. In the above example, the lower limit (cutoff value) of the risk score used to determine that further detailed examination is required is set to 90 points. The risk score in this example shows that the screening with a high positive hit rate is provided. In fact, 6% of the approximately 9,000 data sets are subjects with a prevalence rate of 5% for glaucoma.

[0172] Two glaucoma cases in the validation data set used in the validation of sensitivity and specificity were judged as false negatives, but considering that their risk scores were 88 and 89, the sensitivity could be improved by adjusting (slightly lowering) the cutoff value. The most important characteristic in large-scale screening is considered to be the height of the positive hit rate, so it can be said that the calculation model of this method has both high feasibility and high reliability.

[0173] As described above, the present method provides an OCT screening algorithm suitable for glaucoma screening. Considering the characteristics of glaucoma as an irreversible progressive disease, it is very important to observe changes over time. Therefore, it is more appropriate to express the new insights of the present method from the perspective of risk scoring rather than from the perspective of setting a clear cutoff value.

[0174] In addition, in highly myopic eyes, the axial length is extended, resulting in deformation of the eyeball. Such OCT images of the eye may reduce the reliability of the data used for glaucoma screening. Therefore, the greater the degree of myopia, the more difficult it becomes to distinguish from glaucoma. On the other hand, regardless of glaucoma, the distribution of RNFL thickness varies greatly with axial length. As mentioned above, even if the estimated axial length calculated based on the angle (direction) of the double hump in the TSNIT curve is used, the accuracy of the risk score is not found to be improved. Similarly, even considering the raw data of axial length, the quality of the risk score is not improved.

[0175] This approach has several advantages, some examples of which are described below.

[0176] The first advantage is that ophthalmologists actually read multiple images to be analyzed and use the resulting larger-scale data set to construct a computational model. That is, the construction of the glaucoma risk scoring algorithm of this method uses a data set that is suitable in both quality and quantity (a high-quality and large-scale data set).

[0177] The second advantage is that all data used in constructing the algorithm of this method are based on collective data and have no hospital bias (for example, the average severity of the disease varies due to the size, history, geographical conditions, etc. of the hospital).

[0178] The third advantage is that the algorithm of this embodiment does not provide glaucoma diagnosis results (risk scores) as output from a black box, but can explain the reasons for obtaining the results from an ophthalmological point of view. This can improve the quality of informed consent, for example.

[0179] The fourth advantage is that the algorithm of the present method is constructed using data from participants who have been appropriately followed up in the medical institution to which the inventor belongs. For example, the inventor considered the progression of glaucoma and conducted a final verification of the algorithm of the present method based on the results of an examination performed at a suitable time after the initial diagnosis (in fact, 4 years after the initial diagnosis).

[0180] The main purpose of the glaucoma risk score of this method is not to provide a final diagnosis of glaucoma, but to provide information (judgment support information) for doctors to judge whether the subject needs to undergo a detailed examination in the hospital. As described above, based on the glaucoma risk score of this method, a high-quality glaucoma determination technology suitable for large-scale screening can be provided.

[0181] The use of the glaucoma risk score of the present embodiment is not limited to the provision of the above-mentioned determination support information (typically, large-scale glaucoma screening). For example, the glaucoma risk score of the present embodiment can be used for the final diagnosis and treatment of glaucoma.

[0182] <Ophthalmic devices>

[0183] An ophthalmic device is described as a non-limiting example of the glaucoma screening method of the embodiment described above. Any matters related to the glaucoma screening method of the embodiment may be at least partially combined with the ophthalmic device of some examples. In addition, any known technology may be combined with any ophthalmic device of any example.

[0184] The ophthalmologic apparatus of this embodiment is used, for example, in collective examinations of ophthalmologic diseases including glaucoma, and generates information indicating the risk of glaucoma of an eye under examination (glaucoma risk information) by processing fundus OCT data using a computational model.

[0185] The calculation model of this embodiment may be any one of the first calculation model Y (Model1), the second calculation model Y (Model2), and the third calculation model Y (Model3) described above, but is not limited thereto.

[0186] In some examples, more than two computational models can be selectively used. The method of selecting a computational model can be arbitrary. For example, the consistency of the inspection can be ensured by constructing a system with the following functions, which include: a function of recording the type of computational model applied to each object (each inspected eye, each inspected person, each group, etc.) for each inspection; a function of notifying the facility performing the new inspection (or the ophthalmic device used for the new inspection) of the type of computational model used in the past inspection when a new inspection is performed. In addition, by constructing a system with a notification function when a new computational model is introduced and / or when an existing computational model is upgraded, the inspection quality can be improved.

[0187] The method by which the ophthalmic device of the present embodiment acquires fundus OCT data may be arbitrary. Some examples of ophthalmic devices may include an OCT scanner that applies OCT scanning to the fundus of the eye to be examined. In addition, some examples of ophthalmic devices may include an element that acquires OCT data generated by applying OCT scanning to the fundus of the eye to be examined using another ophthalmic device (ophthalmic OCT device) including an OCT scanner from another device. The other device may be, for example, an ophthalmic OCT device that generates the OCT data, a medical image archiving system (e.g., Picture Archiving and Communication System (PACS)) that stores the OCT data, or a recording medium that records the OCT data. The unit that acquires OCT data from another device may be, for example, a communication device for performing data communication between an ophthalmic OCT device or a medical image archiving system, a drive device for reading OCT data from a recording medium, or the like.

[0188] The glaucoma risk information of the present method can be any form or method to express any type of glaucoma risk information. Some examples of glaucoma risk information can include the above-mentioned glaucoma risk score, and can also include information indicating the degree of glaucoma risk (e.g., high risk, medium risk, low risk, etc.).

[0189] Fig.12 An example of the configuration of the ophthalmologic apparatus of this embodiment is shown in . The ophthalmologic apparatus 120 includes a processor 121 , a memory 125 , an OCT data acquisition unit 127 , and a user interface (UI) 128 .

[0190] The processor 121 is configured to execute various processes based on a program installed in the ophthalmic apparatus 120 and / or a program stored in a device accessible to the ophthalmic apparatus 120. The processor 121 can, for example, execute control of each element of the ophthalmic apparatus 120, control of peripheral devices of the ophthalmic apparatus 120, various data processing, and the like.

[0191] The memory 125 is a storage unit of the ophthalmic device 120. The memory 125 includes any form of storage device, for example, a main storage device (primary storage devices such as a main memory, a cache memory, and a working memory), an auxiliary storage device (secondary storage devices such as a hard disk drive, a flash memory, and a RAM (random access memory) disk), a tertiary storage device, an offline storage, a register, etc.

[0192] The memory 125 holds various data processed by the ophthalmic device 120. The memory 125 stores a calculation model 126. The calculation model 126 may be any calculation model of the above-mentioned embodiments. The calculation model 126 may be a linear equation including a linear coupling of one or more papilla parameters representing the thickness of retinal tissue in the optic nerve head region and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in the macular region.

[0193] The calculation model 126 is pre-created and stored in the memory 125. For example, the calculation model 126 can be a calculation model of any type created at least in part using the glaucoma screening method of the above-mentioned embodiment, and can include at least one of the first calculation model Y (Model1), the second calculation model Y (Model2) and the third calculation model Y (Model3).

[0194] The OCT data acquisition unit 127 acquires OCT data of the fundus of a living eye (examined eye) that is a glaucoma screening target of the ophthalmologic apparatus 120. The OCT data acquisition unit 127 may include, for example, the aforementioned OCT scanner and / or an element that acquires the aforementioned OCT data from another device.

[0195] The OCT scanner, for example, includes a light source, a splitter for splitting the light emitted from the light source into measurement light and reference light, a sampling arm for projecting the measurement light to the fundus of the eye to be examined and guiding the return light (reflected light, backscattered light, etc.) of the measurement light from the eye to be examined, a reference arm for guiding the reference light, a combiner for generating interference light by superimposing the return light of the measurement light with the reference light, and a photodetector for generating an electrical signal (detection signal) by detecting the interference light. An optical deflector (optical scanner) such as a galvanometer scanner is provided in the sampling arm to project the measurement light to a plurality of different positions of the fundus of the eye to be examined.

[0196] Furthermore, the OCT scanner includes an image construction unit (image construction circuit, composed of image construction circuits) for applying signal processing to the detection signal generated by the light detector to construct OCT data (typically, OCT image data). The image construction unit, for example, generates a reflection intensity curve (A-line curve) for each A-line based on the spectral distribution of data at each projection position (scanning point, A-line) of the measurement light, images each A-line curve to generate a plurality of A-scan image data, and arranges these A-scan image data according to a scanning pattern (configuration of a plurality of scanning points). The above-mentioned signal processing for generating the A-line curve includes noise reduction (denoising), filtering, fast Fourier transform (FFT), etc.

[0197] The OCT modality used by the OCT scanner can be arbitrary, for example, it can be spectral domain OCT or swept source OCT. Spectral domain OCT is an OCT modality that splits the light emitted from a low-coherence light source (broadband light source) into measurement light and reference light, projects the measurement light onto the object to be tested, makes the return light of the measurement light from the object to be tested overlap with the reference light to generate interference light, detects the spectral distribution of the generated interference light using a spectrometer, and performs data processing such as Fourier transform on the detected spectral distribution to construct image data.

[0198] On the other hand, swept source OCT is a method of splitting light from a wavelength-variable light source into measurement light and reference light, projecting the measurement light onto the object to be tested, causing the return light of the measurement light from the object to overlap with the reference light to generate interference light, detecting the generated interference light with a photodetector, and performing data processing such as Fourier transform on the detection data collected by scanning the wavelength and scanning the measurement light to construct image data.

[0199] In short, spectral domain OCT is an OCT modality that obtains the spectral distribution of interference light by spatial division, and swept source OCT is an OCT method that obtains the spectral distribution of interference light by temporal division. For a non-limiting specific example of spectral domain OCT, refer to U.S. Patent No. 9,295,386. In addition, for a non-limiting specific example of swept source OCT, refer to U.S. Patent No. 11,013,400.

[0200] The user interface 128 is an interface for exchanging information between the ophthalmic apparatus 120 and the user, and includes, for example, an input interface and an output interface. The input interface is an interface for the user to input information to the ophthalmic apparatus 120, and examples thereof include a pointer device, a keyboard, a switch, a sound input device, an image (video) input device, and the like. The output interface is an interface for outputting information from the ophthalmic apparatus 120, and examples thereof include a display, a sound output device, a printer, and the like.

[0201] The ophthalmic device 120 of this example may include various elements not shown in the figure, for example, any element described in the above-mentioned documents (US Pat. No. 9,295,386 and US Pat. No. 11,013,400) or any element described in other documents related to OCT.

[0202] The processor 121 includes a control unit (control circuit, control circuit component) 122 , a measured value calculation unit (measured value calculation circuit, measured value calculation circuit component) 123 , and a risk information generation unit (risk information generation circuit, risk information generation circuit component) 124 .

[0203] The control unit 122 is configured to execute control related to the measurement value calculation unit 123 , control related to the risk information generation unit 124 , control related to the memory 125 , control related to the OCT data acquisition unit 127 , and control related to the user interface 128 .

[0204] The control related to the measured value calculation unit 123 may include, for example, the start / stop of the operation of the measured value calculation unit 123, the control related to the parameters of the measured value calculation unit 123, the input of data to the measured value calculation unit 123, the reception of output data from the measured value calculation unit 123, etc. The control related to the risk information generation unit 124 may be similar.

[0205] The control related to the memory 125 may include, for example, a process of storing (saving) data in the memory 125 , a process of retrieving data from the memory 125 , a process of processing or editing data stored in the memory 125 , and the like.

[0206] The controls related to the OCT data acquisition unit 127 may include, for example, control of the light source, control of the optical scanner, control of the optical path length of the sampling arm and / or the reference arm, control of the focusing of the measurement light, control of the polarization state of the measurement light and / or the reference light, control of the light amount (intensity) of the measurement light and / or the reference light, control of the light detector, control of the data collection system (DAS, DAQ) not shown in the figure, etc.

[0207] Controls related to the user interface 128 may include, for example, control of a display, control of a graphical user interface (GUI), and the like.

[0208] The control unit 122 can perform control on other elements of the ophthalmologic apparatus 120. This control can be performed in the same manner as in the related art (for example, alignment control, focus control, tracking control, etc.).

[0209] The control unit 122 inputs the OCT data of the fundus of the subject's eye generated by the OCT data acquisition unit 127 to the measurement value calculation unit 123. The measurement value calculation unit 123 is configured to calculate a plurality of measurement values ​​input to the calculation model 126 for glaucoma screening of the subject's eye based on the OCT data.

[0210] When the calculation model 126 includes two or more calculation models, for example, the control unit 122 transmits information indicating the type of measurement values ​​corresponding to the calculation model used for glaucoma screening of the subject eye (measurement value type information) to the measurement value calculation unit 123. The measurement value calculation unit 123 identifies the types of a plurality of measurement values ​​to be calculated based on the measurement value type information, and calculates each of these measurement values ​​based on the OCT data of the fundus of the subject eye.

[0211] As described above, the calculation model 126 of some examples may be a linear equation including a linear coupling of one or more papillary parameters representing the thickness of the retinal tissue in the optic nerve papilla region and a linear coupling of one or more macular parameters representing the thickness of the retinal tissue in the macular region. When such a calculation model 126 is used in glaucoma screening of the subject's eye, the plurality of measurement values ​​calculated by the measurement value calculation unit 123 include one or more papillary measurement values ​​corresponding to the one or more papillary parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters.

[0212] When the calculation model 126 used in glaucoma screening of the examined eye is the first calculation model Y (Model 1), the measurement value calculation unit 123 calculates the measurement value corresponding to TSNITlower, the measurement value corresponding to cpRNFLqS, the measurement value corresponding to cpRNFLqIt, the measurement value corresponding to cpRNFLqIn, the measurement value corresponding to mGCIPL_IT, the measurement value corresponding to log_mGCIPL_STvsIT, the measurement value corresponding to log_mGCIPL_ITvsIN, and the measurement value corresponding to mGCC_IN based on the OCT data of the fundus of the examined eye.

[0213] When the calculation model 126 used in glaucoma screening of the examined eye is the second calculation model Y (Model 2), the measurement value calculation unit 123 calculates the measurement value corresponding to TSNITlower, the measurement value corresponding to cpRNFLqS, the measurement value corresponding to cpRNFLqIt, the measurement value corresponding to cpRNFLqIn, the measurement value corresponding to log_mGCIPL_STxvsITx, and the measurement value corresponding to log_mGCC_ITxvsINx based on the OCT data of the fundus of the examined eye.

[0214] When the calculation model 126 used in glaucoma screening of the examined eye is the third calculation model Y (Model 3), the measurement value calculation unit 123 calculates the measurement value corresponding to TSNITlower, the measurement value corresponding to cpRNFLqIn, the measurement value corresponding to mGCC, the measurement value corresponding to log_mGCIPL_SNyvsINy, the measurement value corresponding to log_mGCIPL_STyvsITy, and the measurement value corresponding to log_mGCC_ITyvsINy based on the OCT data of the fundus of the examined eye.

[0215] Similarly, when another type of calculation model 126 is used, the measurement value calculation unit 123 obtains measurement values ​​corresponding to each of a plurality of parameters (a plurality of variables) included in the calculation model 126 based on the OCT data of the fundus of the eye to be examined.

[0216] The risk information generating unit 124 generates glaucoma risk information of the subject eye based on the plurality of measurement values ​​calculated by the measurement value calculating unit 123 from the OCT data of the fundus of the subject eye and the calculation model 126 used in glaucoma screening of the subject eye.

[0217] In some examples, the risk information generating unit 124 substitutes each measurement value calculated by the measurement value calculating unit 123 into the corresponding parameter (variable) of the calculation model 126, and calculates a value (e.g., a glaucoma risk score) based on the calculation model 126. The glaucoma risk information generated and output by the risk information generating unit 124 of this example includes, for example, one or both of the calculation result (e.g., a glaucoma risk score) using the calculation model 126 and the information obtained by applying a predetermined algorithm (sequence, calculation method) to the calculation result (e.g., information indicating the degree of glaucoma risk).

[0218] When the calculation model 126 used in the glaucoma screening of the examined eye is the first calculation model Y (Model 1), the risk information generating unit 124 calculates the glaucoma risk score by substituting a plurality of measurement values ​​into a plurality of parameters (a plurality of variables) of the first calculation model Y (Model 1) as shown below: (1) substituting the measurement value corresponding to TSNITlower into the variable TSNITlower; (2) substituting the measurement value corresponding to cpRNFLqS into the variable cpRNFLqS; (3) substituting the measurement value corresponding to cpRNFLqIt into the variable cpRNFL qIt; (4) Substitute the measured value corresponding to cpRNFLqIn into the variable cpRNFLqIn; (5) Substitute the measured value corresponding to mGCIPL_IT into the variable mGCIPL_IT; (6) Substitute the measured value corresponding to log_mGCIPL_STvsIT into the variable log_mGCIPL_STvsIT; (7) Substitute the measured value corresponding to log_mGCIPL_ITvsIN into the variable log_mGCIPL_ITvsIN; (8) Substitute the measured value corresponding to mGCC_IN into the variable mGCC_IN. In addition, the risk information generation unit 124 of this example can generate glaucoma risk information including the glaucoma risk score calculated in this way and / or information indicating the degree of glaucoma risk (e.g., high risk, medium risk, or low risk) obtained based on the glaucoma risk score.

[0219] When the calculation model 126 used in the glaucoma screening of the examined eye is the second calculation model Y (Model2), the risk information generating unit 124 calculates the glaucoma risk score by substituting multiple measurement values ​​into multiple parameters (multiple variables) of the second calculation model Y (Model2) as shown below: (1) Substitute the measurement value corresponding to TSNITlower into the variable TSNITlower; (2) Substitute the measurement value corresponding to cpRNFLqS into the variable cpRNFLqS; (3) Substitute the measurement value corresponding to cpRNFLqIt into the variable cpRNFLqIt; (4) Substitute the measurement value corresponding to cpRNFLqIn into the variable cpRNFLqIn; (5) Substitute the measurement value corresponding to log_mGCIPL_STxvsITx into the variable log_mGCIPL_STxvsITx; (6) Substitute the measurement value corresponding to log_mGCC_ITxvsINx into the variable log_mGCC_ITxvsINx. Then, the risk information generating unit 124 of this example can generate glaucoma risk information including the glaucoma risk score calculated in this way and / or information indicating the degree of glaucoma risk (eg, high risk, medium risk, or low risk) obtained based on the glaucoma risk score.

[0220] When the calculation model 126 used in the glaucoma screening of the examined eye is the third calculation model Y (Model 3), the risk information generating unit 124 calculates the glaucoma risk score by substituting multiple measurement values ​​into multiple parameters (multiple variables) of the third calculation model Y (Model 3) as shown below: (1) Substitute the measurement value corresponding to TSNITlower into the variable TSNITlower; (2) Substitute the measurement value corresponding to cpRNFLqIn into the variable cpRNFLqIn; (3) Substitute the measurement value corresponding to mGCC into the variable mGCC; (4) Substitute the measurement value corresponding to log_mGCIPL_SNyvsINy into the variable log_mGCIPL_SNyvsINy; (5) Substitute the measurement value corresponding to log_mGCIPL_STyvsITy into the variable log_mGCIPL_STyvsITy; (6) Substitute the measurement value corresponding to log_mGCC_ITyvsINy into the variable log_mGCC_ITyvsINy. Then, the risk information generating unit 124 of this example can generate glaucoma risk information including the glaucoma risk score calculated in this way and / or information indicating the degree of glaucoma risk (eg, high risk, medium risk, or low risk) obtained based on the glaucoma risk score.

[0221] When another type of calculation model 126 is used, the risk information generating unit 124 can substitute the plurality of measurement values ​​calculated by the measurement value calculating unit 123 into the plurality of variables of the calculation model 126 used for glaucoma screening of the examined eye in order to generate glaucoma risk information.

[0222] The control unit 122 performs control for providing the glaucoma risk information generated by the risk information generating unit 124. In some examples, the control unit 122 can cause the display of the user interface 128 to display the glaucoma risk information. In some examples, the control unit 122 can control a communication device (not shown) to transmit the glaucoma risk information to another device (e.g., an apparatus, a system, a recording medium, etc.).

[0223] When the calculation model 126 includes a plurality of calculation models and two or more of the calculation models are used to perform glaucoma screening of the subject eye, the measurement value calculation unit 123 calculates two or more measurement value groups corresponding to the two or more calculation models based on the OCT data of the fundus of the subject eye, and the risk information generation unit 124 generates glaucoma risk information based on each of the two or more measurement value groups. Thus, two or more glaucoma risk information corresponding to the two or more calculation models used in glaucoma screening of the subject eye are obtained.

[0224] In some examples, the risk information generating unit 124 can provide the two or more glaucoma risk information.

[0225] In some examples, the risk information generating unit 124 can select one or more glaucoma risk information from the two or more glaucoma risk information and provide the selected one or more glaucoma risk information.

[0226] In some examples, the risk information generating unit 124 can generate final glaucoma risk information based on the two or more glaucoma risk information, and provide the final glaucoma risk information (and the two or more glaucoma risk information).

[0227] In some examples, the risk information generation unit 124 can generate a final glaucoma risk score and / or final information indicating the degree of glaucoma risk based on more than two glaucoma risk scores. The process of generating the final glaucoma risk information based on more than two glaucoma risk information may include, for example, a process of calculating statistical values ​​(e.g., average, weighted average, median, standard deviation, variance, etc.) of more than two glaucoma risk scores, a process of deleting outliers in more than two glaucoma risk scores, etc. The final information indicating the degree of glaucoma risk may be generated, for example, based on the final glaucoma risk score.

[0228] In some examples, the risk information generating unit 124 can generate final information indicating the degree of risk of glaucoma based on a plurality of information indicating the degree of risk of glaucoma. Such final information generating processing may include, for example, a process of selecting one information from the plurality of information by majority voting, a process of selecting one information from the plurality of information by comparing the information obtained from the eye to be examined (the examinee) through this glaucoma screening with the plurality of information, a process of selecting one information from the plurality of information by comparing the information obtained through the past examination of the eye to be examined (the examinee) with the plurality of information, and the like.

[0229] As described above, the ophthalmologic apparatus 120 of this embodiment performs glaucoma risk assessment of the examined eye using the glaucoma screening method of the above-described embodiment, and thus can provide a high-quality glaucoma identification technique suitable for large-scale screening.

[0230] <Program, Recording Medium>

[0231] A program and a recording medium are described as non-limiting example embodiments of the glaucoma screening method according to the above-mentioned embodiment.

[0232] The program of this method is a new program for executing on a computer. The computer includes a processor and a memory. The processor may have Fig.12The processor 121 of the exemplary ophthalmic apparatus 120 shown in the figure may be configured in the same manner as the processor 121 of the exemplary ophthalmic apparatus 120. That is, the processor of this embodiment may include a processor that functions as a control unit 122, a processor that functions as a measurement value calculation unit 123, and a processor that functions as a risk information generation unit 124. Moreover, the memory of this embodiment may store the calculation model (126) in the same manner as the memory 125 of the ophthalmic apparatus 120.

[0233] The program of this embodiment is configured to cause a processor to execute a sequence of steps including the following.

[0234] In the first step, the processor executes control for storing a calculation model in a memory according to the program of the present embodiment. The calculation model may be any of the calculation models of the above-mentioned embodiments, for example, a linear equation including a linear coupling of one or more retinal parameters representing the thickness of retinal tissue in the optic nerve head region and a linear coupling of one or more macula parameters representing the thickness of retinal tissue in the macula region.

[0235] In the second step, the processor executes control for receiving OCT data of the fundus of the eye to be examined (OCT data receiving control) according to the program of this embodiment.

[0236] In some examples, OCT data acceptance control may include processing of acquiring OCT data acquired from the fundus of the subject's eye by the ophthalmic OCT apparatus from another ophthalmic apparatus (ophthalmic OCT apparatus) including an OCT scanner, a medical image archiving system, or a recording medium.

[0237] In some examples, the computer is a component of an ophthalmic apparatus (e.g., the ophthalmic apparatus 120). In addition, the ophthalmic apparatus includes an OCT data acquisition unit (e.g., the OCT data acquisition unit 127). In this case, the OCT data receiving control may include a control for causing the OCT data acquisition unit to perform an OCT scan of the fundus of the eye to be examined and a control for causing the OCT data acquisition unit to perform a generation of OCT data based on the data acquired by the OCT scan.

[0238] That is, the OCT data acceptance control executed in the second step may include any one of control in which the processor accepts OCT data generated by another ophthalmic device from outside the computer and control in which the processor accepts OCT data generated by the ophthalmic device itself including the computer.

[0239] In the third step, the processor executes a process (measurement value calculation process) of calculating a plurality of measurement values ​​from the OCT data acquired in the second step according to the program of this embodiment.

[0240] In some examples, the measurement value calculation process calculates a plurality of measurement values ​​corresponding to the calculation model stored in the memory in the first process based on the OCT data acquired in the second process. When the calculation model is a linear equation including a linear coupling of one or more papillary parameters representing the thickness of retinal tissue in the optic nerve head region and a linear coupling of one or more macular parameters representing the thickness of retinal tissue in the macular region, the measurement value calculation process includes a process of calculating a plurality of measurement values ​​based on the OCT data acquired in the second process, the plurality of measurement values ​​including one or more papillary measurement values ​​corresponding to the one or more papillary parameters, respectively, and one or more macular measurement values ​​corresponding to the one or more macular parameters, respectively.

[0241] In the fourth step, the processor executes a process of inputting the plurality of measurement values ​​calculated in the third step into the calculation model stored in the memory in the first step according to the program of this aspect.

[0242] In the first step, the calculation model stored in the memory is a linear equation including a linear coupling of one or more papillary parameters representing the thickness of retinal tissue in the optic nerve papilla region and one or more macular parameters representing the thickness of retinal tissue in the macular region, and in the third step, a plurality of measurement values ​​including one or more papillary measurement values ​​corresponding to the one or more papillary parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters are calculated, in the fourth step, the corresponding papillary measurement values ​​and the corresponding macular measurement values ​​are respectively input to each papillary parameter (variable) and each macular parameter (variable) of the calculation model. Thus, a glaucoma risk score of the examined eye is derived. The processor can generate glaucoma risk information based on the glaucoma risk score.

[0243] In the fifth step, the processor executes processing for providing the glaucoma risk information generated in the fourth step according to the program of this embodiment. That is, in the fifth step, processing for providing the calculation results output from the calculation model in response to the input of the plurality of measurement values ​​calculated in the third step is executed.

[0244] As described above, the program of this embodiment uses the glaucoma screening method of the above-mentioned embodiment to perform glaucoma risk assessment of the examined eye, and thus can provide a high-quality glaucoma identification technology suitable for large-scale screening.

[0245] The program of the present method can be recorded on a recording medium. That is, the recording medium of the present method is a recording medium on which the program of the present method is recorded. The recording medium of the present method is a computer-readable non-temporary recording medium. The computer-readable non-temporary recording medium of the present method can be a recording medium in any form, for example, it can be any one of a magnetic disk, an optical disk, an optical magnetic disk, and a semiconductor memory.

[0246] Any item related to the glaucoma screening method of the embodiment and / or any item related to the ophthalmic device of the embodiment may be combined at least partially with the program of any embodiment. In addition, any known technology may be combined with the program of any embodiment.

[0247] Likewise, any matter related to the glaucoma screening method of the embodiment and / or any matter related to the ophthalmic device of the embodiment may be combined at least partially with the recording medium of any embodiment. In addition, any known technology may be combined with the recording medium of any embodiment.

[0248] <Some non-limiting characteristics>

[0249] Some non-limiting features of the various embodiments described in detail in this disclosure are described. In addition, the features of the embodiments are not limited to these.

[0250] A first aspect of the glaucoma screening method according to the embodiment includes a calculation model preparation step, an OCT data generation step, a measurement value calculation step, a measurement value input step, and a calculation result providing step.

[0251] In the calculation model preparation step, a calculation model including a linear coupling of one or more papilla parameters representing the retinal tissue thickness in the optic nerve head region and a linear coupling of one or more macula parameters representing the retinal tissue thickness in the macula region is prepared.

[0252] In the OCT data generating step, optical coherence tomography (OCT) scanning is applied to the fundus of the eye to be examined to generate OCT data.

[0253] In the measurement value calculation step, multiple measurement values ​​are calculated based on the OCT data generated in the OCT data generation step, and the multiple measurement values ​​include one or more papilla measurement values ​​corresponding to one or more papilla parameters in the calculation model prepared in the calculation model preparation step and one or more macular measurement values ​​corresponding to one or more macular parameters.

[0254] In the measurement value input step, the plurality of measurement values ​​calculated in the measurement value calculation step are input into the calculation model prepared in the calculation model preparation step.

[0255] In the calculation result providing step, the calculation result output from the calculation model corresponding to the input of the plurality of measurement values ​​to the calculation model executed in the measurement value input step is provided.

[0256] The second embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the first embodiment: the one or more papillary parameters in the calculation model prepared in the calculation model preparation step include a papillary parameter indicating the thickness of the retinal nerve fiber layer in the optic nerve head region.

[0257] The third embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the second embodiment. The calculation model prepared in the calculation model preparation step includes a linear coupling of two or more papillary parameters. The two or more papillary parameters include a first papillary parameter (TSNITlower) that represents a layer thickness value corresponding to the lower peak of two peaks of a double hump presented by a TSNIT curve representing the distribution of retinal nerve fiber layer thickness in a predetermined circular area around the optic nerve head. Moreover, the two or more papillary parameters include a second papillary parameter (cpRNFLqIn) that represents an average value of the retinal nerve fiber layer thickness in an arc-shaped area corresponding to an angle of 225 degrees to 285 degrees defined by taking the temporal direction as zero and taking the direction rotated from the temporal direction to the superior direction as the positive direction in the circular area around the optic nerve head.

[0258] The fourth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the third embodiment. The calculation model prepared in the calculation model preparation step includes a linear coupling of four papilla parameters. The four papilla parameters include a third papilla parameter (cpRNFLqS) and a fourth papilla parameter (cpRNFLqIt) in addition to the first papilla parameter (TSNITlower) and the second papilla parameter (cpRNFLqIn). The third papilla parameter (cpRNFLqS) is a papilla parameter representing the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle range of 45 degrees to 135 degrees in the circular area around the optic nerve disc. The fourth papilla parameter (cpRNFLqIt) is a papilla parameter representing the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle range of 285 degrees to 315 degrees in the circular area around the optic nerve disc.

[0259] The fifth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of any one of the first to fourth embodiments. The calculation model prepared in the calculation model preparation step includes a linear coupling of two or more macular parameters. The two or more macular parameters include a first macular parameter based on the thickness of a composite tissue of ganglion cells and inner plexiform layer (mGCIPL) in the macular area and a second macular parameter based on the thickness of a retinal ganglion cell complex (mGCC) in the macular area.

[0260] The sixth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the fifth embodiment: The first macular parameter (parameter related to mGCIPL) includes a layer thickness value parameter (mGCIPL_IT) indicating the value of the thickness of the complex tissue (mGCIPL).

[0261] In addition to the non-limiting features of the fifth embodiment, the seventh embodiment of the glaucoma screening method of the embodiment has the following non-limiting features: The first macular parameter (parameter related to mGCIPL) includes a comparison value parameter (log_mGCIPL_STvsIT, log_mGCIPL_ITvsIN; log_mGCIPL_STxvsITx; log_mGCIPL_SNyvsINy, log_mGCIPL_STyvsITy), which represents a comparison value between the first composite tissue thickness in the first sub-region of the macular region and the second composite tissue thickness in the second sub-region.

[0262] The eighth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the seventh embodiment: the comparison value represented by the comparison value parameter included in the first macular parameter (parameter related to mGCIPL) is calculated by applying a logarithmic operation to the true number of the difference between the first composite tissue thickness in the first sub-region including the macular region and the second composite tissue thickness in the second sub-region.

[0263] The ninth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments: The second macular parameter (parameter related to mGCC) includes a layer thickness value parameter (mGCC_IN; mGCC) indicating the value of the retinal ganglion cell complex (mGCC) thickness.

[0264] The tenth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments: The second macular parameter (parameter related to mGCC) includes a comparison value parameter (log_mGCC_ITxvsINx; log_mGCC_ITyvsINy) indicating a comparison value between the first retinal ganglion cell complex thickness in the third sub-region and the second retinal ganglion cell complex thickness in the fourth sub-region of the macular region.

[0265] The eleventh embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the tenth embodiment: the comparison value represented by the comparison value parameter included in the second macular parameter (parameter related to mGCC) is calculated by applying a logarithmic operation to the true number of the difference between the first retinal ganglion cell complex thickness in the third sub-region including the macular region and the second retinal ganglion cell complex thickness in the fourth sub-region.

[0266] The twelfth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of any one of the first to eleventh embodiments. In the calculation model preparation step, a calculation model is created based on a development data set collected from a first group, a first verification data set collected from a second group different from the first group, and a second verification data set collected based on a subgroup determined as suspected glaucoma in the first group.

[0267] The thirteenth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of the twelfth embodiment: the development data set includes fundus photographs acquired by digital fundus photography, fundus OCT data acquired by optical coherence tomography, and visual field data acquired by visual field examination.

[0268] In addition to the non-limiting features of the twelfth or thirteenth embodiments, the fourteenth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features: In the calculation model preparation step, the first verification data set includes fundus OCT data acquired by optical coherence tomography for verification of the positive hit rate.

[0269] The fifteenth embodiment of the glaucoma screening method of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the twelfth to fourteenth embodiments: In the calculation model preparation step, the second validation data set includes visual field data obtained by visual field examination, which is used to verify the sensitivity and specificity.

[0270] In any of the first to fifteenth aspects of the glaucoma screening method of the embodiment, any matters described or suggested in the present disclosure may be at least partially combined.

[0271] A first aspect of an ophthalmologic apparatus according to an embodiment includes a storage unit, an OCT data acquisition unit, a measurement value calculation unit, and a risk information generation unit.

[0272] The storage unit stores a pre-created calculation model. The calculation model includes a linear coupling of one or more papilla parameters representing the thickness of retinal tissue in the optic nerve head region and a linear coupling of one or more macula parameters representing the thickness of retinal tissue in the macula region. Fig.12 The memory 125 of the ophthalmologic apparatus 120 shown is an example of a storage unit in this embodiment.

[0273] The OCT data acquisition unit is configured to acquire optical coherence tomography data (OCT data) of the fundus of the eye to be examined. Fig.12 The OCT data acquisition unit 127 of the ophthalmologic apparatus 120 shown is an example of the OCT data acquisition unit of the present embodiment.

[0274] The measurement value calculation unit is configured to calculate multiple measurement values ​​based on the OCT data acquired by the OCT data acquisition unit, and the multiple measurement values ​​include one or more papilla measurement values ​​corresponding to one or more papilla parameters in the calculation model stored in the storage unit and one or more macular measurement values ​​corresponding to one or more macular parameters. Fig.12 The measured value calculation unit 123 of the ophthalmologic apparatus 120 shown is an example of the measured value calculation unit of the present embodiment.

[0275] The risk information generating unit is configured to generate glaucoma risk information of the eye to be examined based on the calculation model stored in the storage unit and the plurality of measurement values ​​calculated by the measurement value calculating unit. Fig.12 The risk information generating unit 124 of the illustrated ophthalmologic apparatus 120 is an example of a risk information generating unit according to the present embodiment.

[0276] The second embodiment of the ophthalmic apparatus of the embodiment has the following non-limiting features in addition to the non-limiting features of the first embodiment: The one or more papilla parameters in the calculation model stored in the storage unit include a papilla parameter indicating the thickness of the retinal nerve fiber layer in the optic nerve papilla region.

[0277] In addition to the non-limiting features of the second embodiment, the third embodiment of the ophthalmic device of the embodiment has the following non-limiting features. The calculation model stored in the storage unit includes a linear coupling of two or more papillary parameters. The two or more papillary parameters include a first papillary parameter (TSNITlower) representing a layer thickness value corresponding to a lower peak value of two peak values ​​of a double hump presented by a TSNIT curve, wherein the TSNIT curve represents the distribution of the retinal nerve fiber layer thickness in a predetermined circular area around the optic nerve papilla. Moreover, the two or more papillary parameters include a second papillary parameter (cpRNFLqIn) representing an average value of the retinal nerve fiber layer thickness in an arc-shaped area corresponding to an angle in a range of 225 degrees to 285 degrees defined by taking the temporal direction as zero and the direction rotated from the temporal direction to the superior direction as the positive direction in the circular area around the optic nerve papilla.

[0278] In addition to the non-limiting features of the third embodiment, the fourth embodiment of the ophthalmic device of the embodiment has the following non-limiting features. The calculation model stored in the storage unit includes a linear coupling of four papilla parameters. The four papilla parameters include a third papilla parameter (cpRNFLqS) and a fourth papilla parameter (cpRNFLqIt) in addition to the first papilla parameter (TSNITlower) and the second papilla parameter (cpRNFLqIn). The third papilla parameter (cpRNFLqS) is a papilla parameter that represents the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle in the range of 45 degrees to 135 degrees in the circular area around the optic nerve disc. The fourth papilla parameter (cpRNFLqIt) is a papilla parameter that represents the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle in the range of 285 degrees to 315 degrees in the circular area around the optic nerve disc.

[0279] In addition to the non-limiting features of any of the first to fourth embodiments, the fifth embodiment of the ophthalmic device of the embodiment further has the following non-limiting features. The calculation model stored in the storage unit includes a linear coupling of two or more macular parameters. The two or more macular parameters include a first macular parameter based on the thickness of a composite tissue of ganglion cells and inner plexiform layer (mGCIPL) in the macular area and a second macular parameter based on the thickness of a retinal ganglion cell complex (mGCC) in the macular area.

[0280] The sixth embodiment of the ophthalmic apparatus of the embodiment has the following non-limiting features in addition to the non-limiting features of the fifth embodiment: The first macular parameter (parameter related to mGCIPL) includes a layer thickness value parameter (mGCIPL_IT) indicating the value of the thickness of the composite tissue (mGCIPL).

[0281] The seventh embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of the fifth embodiment: The first macular parameter (parameter related to mGCIPL) includes a comparison value parameter (log_mGCIPL_STvsIT, log_mGCIPL_ITvsIN; log_mGCIPL_STxvsITx; log_mGCIPL_SNyvsINy, log_mGCIPL_STyvsITy) indicating a comparison value between the first composite tissue thickness in the first sub-region and the second composite tissue thickness in the second sub-region of the macular region.

[0282] The eighth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of the seventh embodiment: the comparison value represented by the comparison value parameter included in the first macular parameter (parameter related to mGCIPL) is calculated by applying a logarithmic operation to the real number of the difference between the first composite tissue thickness in the first sub-region including the macular region and the second composite tissue thickness in the second sub-region.

[0283] The ninth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments. The second macular parameter (parameter related to mGCC) includes a layer thickness value parameter (mGCC_IN; mGCC) indicating the value of the retinal ganglion cell complex (mGCC) thickness.

[0284] The tenth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments: The second macular parameter (parameter related to mGCC) includes a comparison value parameter (log_mGCC_ITxvsINx; log_mGCC_ITyvsINy) indicating a comparison value between the first retinal ganglion cell complex thickness in the third sub-region and the second retinal ganglion cell complex thickness in the fourth sub-region of the macular region.

[0285] The eleventh embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of the tenth embodiment: the comparison value represented by the comparison value parameter included in the second macular parameter (parameter related to mGCC) is calculated by applying a logarithmic operation to the true number of the difference between the first retinal ganglion cell complex thickness in the third sub-region including the macular region and the second retinal ganglion cell complex thickness in the fourth sub-region.

[0286] The twelfth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the first to eleventh embodiments. The calculation model of this embodiment is created by a calculation model creation device. The calculation model creation device can be a component of the ophthalmic device of this embodiment, or it can be another device different from the ophthalmic device of this embodiment.

[0287] In order to create the computational model of this embodiment, a development data set, a first validation data set, and a second validation data set are prepared.

[0288] The development data set is a data set collected from the first group. That is, the development data set is a data set collected by applying one or more predetermined medical examinations (including ophthalmological examinations) to each subject included in the first group.

[0289] The first validation data set is a data set collected from a second group different from the first group corresponding to the development data set. That is, the first validation data set is a data set collected by applying one or more predetermined medical examinations (including ophthalmological examinations) to each subject included in the second group. The one or more medical examinations performed for the first validation data set and the one or more medical examinations performed for the development data set may be at least one common examination, or may not include a common examination.

[0290] The second validation data set is a data set collected from a subgroup determined to be suspected of glaucoma in the first group corresponding to the development data set. The subgroup is a group consisting of subjects who are determined (diagnosed) to be suspected of glaucoma among a plurality of subjects belonging to the first group that is the object of the development data set. It is clear from its definition that the group is a part (partial set) of the first group.

[0291] Such a development data set, a first validation data set, and a second validation data set are input to a computing model creation device, which creates a computing model based on these data sets and a predetermined algorithm.

[0292] The algorithms used in the creation of the computational model may include machine learning algorithms. The method of machine learning is typically supervised learning (supervised learning), but is not limited thereto. In some examples, in the machine learning used to construct the computational model, in addition to or instead of supervised learning, any known method can be used alone or in combination, such as unsupervised learning (unsupervised learning), reinforcement learning, semi-supervised learning (semi-supervised learning), transduction, multi-task learning, etc.

[0293] The methods for constructing the computational model are not limited to the examples shown here and may include any well-known methods, such as support vector machines, Bayesian classifiers, boosting, k-means, kernel density inference, principal component analysis, independent component analysis, self-organizing maps, random forests, generative adversarial networks (GANs), etc.

[0294] The thirteenth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of the twelfth embodiment: the development data set input into the calculation model creation device includes fundus photographs acquired by digital fundus photography, fundus OCT data acquired by optical coherence tomography, and visual field data acquired by visual field examination.

[0295] In addition to the non-limiting features of the twelfth or thirteenth embodiments, the fourteenth embodiment of the ophthalmic device of the embodiment has the following non-limiting features. The first verification data set input to the calculation model creation device includes fundus OCT data acquired by optical coherence tomography. The calculation model creation device performs verification of the positive hit rate based on the input first verification data set.

[0296] The fifteenth embodiment of the ophthalmic device of the embodiment has the following non-limiting features in addition to the non-limiting features of any one of the twelfth to fourteenth embodiments. The second verification data set input to the calculation model creation device includes visual field data acquired by visual field examination. The calculation model creation device performs sensitivity and specificity verification based on the input second verification data set.

[0297] Any of the first to fifteenth aspects of the ophthalmologic apparatus of the embodiment may be at least partially combined with any of the matters described or suggested in the present disclosure.

[0298] A first aspect of the program of the embodiment is a program for causing a computer including a processor and a memory to execute, and is configured to cause the processor to execute calculation model storage control, OCT data reception control, measurement value calculation processing, measurement value input processing, and calculation result providing processing.

[0299] In the calculation model storage control, the program causes the processor to execute processing for storing in the memory a calculation model including linear coupling of one or more papilla parameters representing retinal tissue thickness in the optic nerve head area and linear coupling of one or more macula parameters representing retinal tissue thickness in the macula area.

[0300] In the OCT data acceptance control, the program causes the processor to execute a process of accepting optical coherence tomography data (OCT data) of the fundus of the eye to be examined.

[0301] In the measurement value calculation processing, the program causes the processor to execute a process of calculating multiple measurement values ​​based on the OCT data received through the OCT data receiving processing, and the multiple measurement values ​​include one or more papilla measurement values ​​corresponding to one or more papilla parameters in the calculation model stored in the memory through the calculation model storage control and one or more macular measurement values ​​corresponding to one or more macular parameters.

[0302] In the measured value input process, the program causes the processor to execute a process of inputting a plurality of measured values ​​calculated in the measured value calculation process into a calculation model stored in the memory by the calculation model storage control.

[0303] In the calculation result providing process, the program causes the processor to execute a process of providing the calculation result output from the calculation model in response to the input of a plurality of measurement values ​​to the calculation model executed in the measurement value input process.

[0304] A first aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the first aspect is recorded.

[0305] The second embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the first embodiment: One or more papilla parameters in the calculation model stored in the memory by the calculation model storage control include a papilla parameter indicating the thickness of the retinal nerve fiber layer in the optic nerve papilla region.

[0306] A second aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the second aspect is recorded.

[0307] The third embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the second embodiment. The calculation model stored in the memory by the calculation model storage control includes a linear coupling of two or more papillary parameters. The two or more papillary parameters include a first papillary parameter (TSNITlower) corresponding to a layer thickness value corresponding to the lower peak of two peaks of a double hump presented by a TSNIT curve representing the distribution of retinal nerve fiber layer thickness in a predetermined circular area around the optic nerve disc. Moreover, the two or more papillary parameters include a second papillary parameter (cpRNFLqIn) representing an average value of the retinal nerve fiber layer thickness in an arc-shaped area corresponding to an angle range of 225 degrees to 285 degrees defined by taking the temporal direction as zero and the direction rotated from the temporal direction to the superior direction as the positive direction in the circular area around the optic nerve disc.

[0308] A third aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the third aspect is recorded.

[0309] The fourth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the third embodiment. The calculation model stored in the memory by the calculation model storage control includes a linear coupling of four papilla parameters. The four papilla parameters include a third papilla parameter (cpRNFLqS) and a fourth papilla parameter (cpRNFLqIt) in addition to the first papilla parameter (TSNITlower) and the second papilla parameter (cpRNFLqIn). The third papilla parameter (cpRNFLqS) is a papilla parameter indicating the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle in the range of 45 degrees to 135 degrees in the circular area around the optic nerve papilla. The fourth papilla parameter (cpRNFLqIt) is a papilla parameter indicating the average value of the thickness of the retinal nerve fiber layer in the arc-shaped area corresponding to the defined angle in the range of 285 degrees to 315 degrees in the circular area around the optic nerve papilla.

[0310] A fourth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the fourth aspect is recorded.

[0311] The fifth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of any one of the first to fourth embodiments. The calculation model stored in the memory by the calculation model storage control includes linear coupling of two or more macular parameters. The two or more macular parameters include a first macular parameter based on the thickness of the composite tissue of ganglion cells and inner plexiform layer (mGCIPL) in the macular area and a second macular parameter based on the thickness of the retinal ganglion cell complex (mGCC) in the macular area.

[0312] A fifth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the fifth aspect is recorded.

[0313] The sixth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the fifth embodiment: The first macular parameter (parameter related to mGCIPL) includes a layer thickness value parameter (mGCIPL_IT) indicating the value of the thickness of the composite tissue (mGCIPL).

[0314] A sixth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the sixth aspect is recorded.

[0315] The seventh embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the fifth embodiment: The first macular parameter (parameter related to mGCIPL) includes a comparison value parameter (log_mGCIPL_STvsIT, log_mGCIPL_ITvsIN; log_mGCIPL_STxvsITx; log_mGCIPL_SNyvsINy, log_mGCIPL_STyvsITy) indicating a comparison value between the first composite tissue thickness in the first sub-region and the second composite tissue thickness in the second sub-region of the macular region.

[0316] A seventh aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the seventh aspect is recorded.

[0317] The eighth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the seventh embodiment: the comparison value represented by the comparison value parameter included in the first macular parameter (parameter related to mGCIPL) is calculated by applying a logarithmic operation to the real number of the difference between the first composite tissue thickness in the first sub-region including the macular region and the second composite tissue thickness in the second sub-region.

[0318] An eighth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the eighth aspect is recorded.

[0319] The ninth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments: The second macular parameter (parameter related to mGCC) includes a layer thickness value parameter (mGCC_IN; mGCC) indicating the value of the retinal ganglion cell complex (mGCC) thickness.

[0320] A ninth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the ninth aspect is recorded.

[0321] The tenth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the fifth to eighth embodiments: The second macular parameter (parameter related to mGCC) includes a comparison value parameter (log_mGCC_ITxvsINx; log_mGCC_ITyvsINy) indicating a comparison value between the first retinal ganglion cell complex thickness in the third sub-region and the second retinal ganglion cell complex thickness in the fourth sub-region of the macular region.

[0322] A tenth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium having the program according to the tenth aspect recorded thereon.

[0323] The eleventh embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the tenth embodiment: the comparison value represented by the comparison value parameter included in the second macular parameter (parameter related to mGCC) is calculated by applying a logarithmic operation to the real number of the difference between the first retinal ganglion cell complex thickness in the third sub-region including the macular region and the second retinal ganglion cell complex thickness in the fourth sub-region.

[0324] An eleventh aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the eleventh aspect is recorded.

[0325] The twelfth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of any of the first to eleventh embodiments. The program of this embodiment is configured to cause the processor to execute a calculation model creation process for creating a calculation model stored in the memory by the calculation model storage control.

[0326] In the computational model creation process, the program causes the processor to execute the process of creating a computational model based on a development data set collected from a first group, a first verification data set collected from a second group different from the first group, and a second verification data set collected from a subgroup determined to be suspected of glaucoma in the first group.

[0327] A twelfth embodiment of the recording medium of the embodiment is a computer-readable non-transitory recording medium on which the program of the twelfth embodiment is recorded.

[0328] The thirteenth embodiment of the program of the embodiment has the following non-limiting features in addition to the non-limiting features of the twelfth embodiment: the development data set input to the processor includes fundus photographs acquired by digital fundus photography, fundus OCT data acquired by optical coherence tomography, and visual field data acquired by visual field examination.

[0329] A thirteenth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the thirteenth aspect is recorded.

[0330] In addition to the non-limiting features of the twelfth or thirteenth embodiments, the fourteenth embodiment of the program of the implementation method also has the following non-limiting features. The first verification data set input to the processor includes fundus OCT data acquired by optical coherence tomography. The processor performs verification of the positive hit rate based on the input first verification data set according to the program.

[0331] A fourteenth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the fourteenth aspect is recorded.

[0332] In addition to the non-limiting features of any of the twelfth to fourteenth embodiments, the fifteenth embodiment of the program of the embodiment further has the following non-limiting features. The second verification data set input to the processor includes visual field data obtained by visual field examination. The processor performs sensitivity and specificity verification based on the input second verification data set according to the program.

[0333] A fifteenth aspect of the recording medium according to the embodiment is a computer-readable non-transitory recording medium on which the program according to the fifteenth aspect is recorded.

[0334] Any of the first to fifteenth aspects of the program of the embodiment may be at least partially combined with any matter described or suggested in the present disclosure.

[0335] Any of the first to fifteenth aspects of the recording medium of the embodiment may be at least partially combined with any matter described or suggested in the present disclosure.

[0336] The present disclosure proposes some embodiments and some methods thereof. These embodiments and methods are merely non-limiting illustrations of the present invention. Therefore, any modification (omission, substitution, addition, etc.) within the scope of the gist of the present invention can be implemented in the embodiments or methods proposed in the present disclosure.

[0337] Description of Reference Numerals

[0338] 120 Ophthalmic devices

[0339] 121 Processor

[0340] 122 Control Department

[0341] 123 Measurement value calculation unit

[0342] 124 Risk Information Generation Department

[0343] 125 Memory

[0344] 126 Computational Model

[0345] 127 OCT data acquisition unit

[0346] 128 User Interface

Claims

1. A method for glaucoma screening, wherein: preparing a computational model including a linear coupling of one or more papillary parameters representing retinal tissue thickness in the optic nerve head region and a linear coupling of one or more macular parameters representing retinal tissue thickness in the macular region, Apply optical coherence tomography to the fundus of the eye under examination to generate OCT data, calculating a plurality of measurement values ​​based on the OCT data, the plurality of measurement values ​​including one or more papilla measurement values ​​corresponding to the one or more papilla parameters respectively and one or more macular measurement values ​​corresponding to the one or more macular parameters respectively, inputting the plurality of measured values ​​into the computational model, A calculation result output from the calculation model corresponding to the input of the plurality of measurement values ​​is provided.

2. The method according to claim 1, wherein: The one or more papillary parameters include a papillary parameter representing a thickness of the retinal nerve fiber layer in the optic nerve head region.

3. The method according to claim 2, wherein: The computational model includes linear coupling of more than two nipple parameters, The two or more nipple parameters include: A first papilla parameter representing a layer thickness value corresponding to a lower peak of two peaks of a double hump presented by a TSNIT curve representing the distribution of retinal nerve fiber layer thickness in a predetermined circular region around the optic nerve papilla; and The second nipple parameter represents an average value of the retinal nerve fiber layer thickness in an arc-shaped area corresponding to an angle ranging from 225 degrees to 285 degrees defined by taking the ear-side direction as zero and the direction rotated upward from the ear-side direction as the positive direction in the circular area.

4. The method according to claim 3, wherein: The computational model includes the linear coupling of four nipple parameters, The four nipple parameters include: the first nipple parameter; the second nipple parameter; A third nipple parameter representing an average value of the thickness of the retinal nerve fiber layer in an arc-shaped area corresponding to the angle ranging from 45 degrees to 135 degrees in the circular area; and The fourth nipple parameter represents an average value of the thickness of the retinal nerve fiber layer in an arc-shaped area corresponding to the angle ranging from 285 degrees to 315 degrees in the circular area.

5. The method according to any one of claims 1 to 4, wherein: The computational model includes linear coupling of two or more macular parameters, The two or more macular parameters include: A first macular parameter is a parameter based on the composite tissue thickness of ganglion cells and inner plexiform layer in the macular region; and The second macular parameter is a parameter based on the thickness of the retinal ganglion cell complex in the macular area.

6. The method according to claim 5, wherein: The first macular parameter comprises a layer thickness value parameter representing a value of the composite tissue thickness.

7. The method according to claim 5, wherein: The first macular parameter includes a comparison value parameter representing a comparison value between a first composite tissue thickness in a first sub-region and a second composite tissue thickness in a second sub-region of the macular region.

8. The method according to claim 7, wherein: The comparison value is calculated by applying a logarithmic operation to a real number including a difference between the first composite tissue thickness and the second composite tissue thickness.

9. The method according to claim 5, wherein: The second macular parameter includes a layer thickness value parameter representing a value of the retinal ganglion cell complex thickness.

10. The method according to claim 5, wherein: The second macular parameter includes a comparison value parameter representing a comparison value between a first retinal ganglion cell complex thickness in a third sub-region and a second retinal ganglion cell complex thickness in a fourth sub-region of the macular region.

11. The method according to claim 10, wherein: The comparison value is calculated by applying a logarithmic operation to a real number including a difference between the first retinal ganglion cell complex thickness and the second retinal ganglion cell complex thickness.

12. The method according to any one of claims 1 to 4, wherein: The computational model is created based on a development data set collected from a first population, a first validation data set collected from a second population different from the first population, and a second validation data set collected from a subpopulation of the first population determined to be suspected of glaucoma.

13. The method according to claim 12, wherein: The development data set includes fundus photographs obtained by digital fundus photography, fundus OCT data obtained by optical coherence tomography, and visual field data obtained by visual field examination.

14. The method according to claim 12, wherein: The first verification data set includes fundus OCT data acquired by optical coherence tomography, which is used for verification of the positive hit rate.

15. The method according to claim 12, wherein: The second validation data set includes visual field data obtained through visual field examination, and is used for validation of sensitivity and specificity.

16. An ophthalmic device comprising: A storage unit storing a pre-created calculation model, the calculation model including a linear coupling of one or more papilla parameters representing the thickness of retinal tissue in the optic nerve papilla region and a linear coupling of one or more macula parameters representing the thickness of retinal tissue in the macula region; an OCT data acquisition unit for acquiring optical coherence tomography (OCT) data of the fundus of the eye under examination; a measurement value calculation unit that calculates a plurality of measurement values ​​based on the OCT data, the plurality of measurement values ​​including one or more papilla measurement values ​​corresponding to the one or more papilla parameters and one or more macular measurement values ​​corresponding to the one or more macular parameters; as well as The risk information generating unit generates glaucoma risk information of the inspected eye based on the calculation model and the plurality of measurement values.

17. A program for causing a computer comprising a processor and a memory to execute, wherein: The program causes the processor to execute: storing a computational model in the memory, the computational model comprising a linear coupling of one or more papillary parameters representing retinal tissue thickness in the optic nerve head region and a linear coupling of one or more macular parameters representing retinal tissue thickness in the macular region; A control receiving optical coherence tomography (OCT) data of the fundus of an eye under examination; a process of calculating a plurality of measurements based on the OCT data, the plurality of measurements including one or more papillary measurements corresponding to the one or more papillary parameters, respectively, and one or more macular measurements corresponding to the one or more macular parameters, respectively; inputting the plurality of measured values ​​into the computational model; as well as A process of providing a calculation result output from the calculation model in response to input of the plurality of measurement values.

18. A computer-readable non-transitory recording medium having the program according to claim 17 recorded thereon.

Images