WO2023136460A1 - System and method for glaucoma analysis using archetypal analysis and fcm analysis - Google Patents

Abstract

The present invention relates to a system and method for glaucoma analysis using archetype analysis and FCM analysis. More specifically, the system comprises: a data acquisition unit for acquiring central visual field data of a subject; an archetype determination unit for determining a plurality of archetypes by using archetypal analysis (AA); a disassembly unit for performing Fuzzy C-Means (FCM) clustering on the central visual field data and then disassembling the FCM-clustered data into at least one archetype among the plurality of archetypes determined by the archetype determination unit; and a glaucoma determination unit for determining glaucoma according to the result of disassembling the central visual field data.

Description

Glaucoma analysis system and method using prototype analysis and FCM analysis techniques

The present invention relates to a glaucoma analysis system and method using circular analysis and FCM analysis techniques, and more specifically, to analyze central visual field data of a subject using circular analysis and FCM analysis techniques and diagnose glaucoma. It relates to a glaucoma analysis system and method using prototype analysis and FCM analysis techniques.

Glaucoma is a disease that can lead to blindness due to damage to retinal ganglion cells (RGCs) and their axons. Because glaucoma is chronic and progresses irreversibly, early detection of glaucoma can be delayed through treatment or surgery. However, due to the nature of the disease, there are many cases where there are no subjective symptoms such as visual field defect or visual impairment that the patient can clearly feel until the terminal stage. In addition, since the treatment prognosis is poor in the case of severely advanced glaucoma, the importance of early detection through examination is very great.

In addition, usually, medical staff reading medical images and diagnosing glaucoma is a very difficult task that only highly skilled ophthalmologists can do, and since each ophthalmologist has different diagnostic criteria, objective and consistent diagnosis results are derived. difficulties exist.

To solve this problem, studies on fundus image analysis methods and glaucoma diagnosis methods using artificial intelligence are being actively conducted. Related document 1 (Korean Patent Registration No. 10-2295426) relates to an eye disease detection device and method based on artificial intelligence, and related document 2 (Korean Patent Publication No. 10-2021-0124024) uses artificial intelligence A quality score calculation method and computer program for fundus image data. Through related documents 1 and 2, it is possible to detect eye diseases based on artificial intelligence using fundus image data of better quality, but the central field of view (Visual Field; VF) is easy to confirm the location and change of ganglion cells (RGC). ) There are technical limitations in providing more objective and consistent glaucoma diagnosis results based on data. Therefore, a related technology is required.

The present invention is to solve the above problems, and based on the central visual field data that can easily confirm the location and change of ganglion cells (RGC), central visual field data can be determined more objectively and consistently with Fuzzy C-Means. After clustering, decomposes into at least one of the plurality of archetypes determined through the archetype analysis (AA) technique and uses the archetype analysis and FCM analysis technique to determine whether or not glaucoma is present according to the decomposition result. To obtain a system and method for analyzing glaucoma aims to

The technical problems to be achieved by the present invention are not limited to the above-mentioned technical problems, and other technical problems not mentioned can be clearly understood by those skilled in the art from the description of the present invention.

In order to achieve the above object, a glaucoma analysis system using circular analysis and FCM analysis techniques of the present invention includes a data acquisition unit for obtaining central field data of a subject; An archetype determining unit for determining a plurality of archetypes using an archetypal analysis (AA) technique; a decomposition unit for decomposing the central field data into at least one archetype among the plurality of archetypes determined by the archetype determination unit after Fuzzy C-Means clustering (FCM clustering); and a glaucoma determining unit determining whether or not glaucoma is present according to a classification result of the central visual field data.

In order to achieve the above object, a glaucoma analysis method using circular analysis and FCM analysis techniques of the present invention includes a data acquisition step of obtaining central field data of a subject by a data acquisition unit; An archetype determination step in which a plurality of archetypes are determined by using an archetypal analysis (AA) technique by the archetype determiner; a decomposition step of decomposing, by a decomposition unit, the central field data into at least one archetype among the plurality of archetypes determined by the archetype determination unit after Fuzzy C-Means clustering (FCM clustering); and a glaucoma determination step of determining, by a glaucoma determination unit, whether or not glaucoma is present according to a decomposition result of the central visual field data.

As described above, according to the present invention, after Fuzzy C-Means clustering of central visual field data, decomposition into at least one archetype among the plurality of archetypes determined through the archetype analysis (AA) technique, and glaucoma according to the decomposition result By determining whether or not, there is an effect of determining glaucoma more objectively and consistently based on central field data that facilitates checking the location and change of ganglion cells (RGC).

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the detailed description and claims.

1 is a block diagram of a system for analyzing glaucoma using circular analysis and FCM analysis techniques according to the present invention.

2 is a diagram explaining the difference between Archetypal Analysis (AA) and Fuzzy C-Means clustering (FCM Clustering) according to an embodiment of the present invention.

3 is a diagram showing 10 archetypes according to an embodiment of the present invention.

4 is a view showing that central field (VF) data of a subject is decomposed into three archetypes of AT2, AT9, and AT6 according to an embodiment of the present invention.

5 is a flow chart of a glaucoma analysis method using circular analysis and FCM analysis techniques of the present invention.

The terms used in this specification have been selected from general terms that are currently widely used as much as possible while considering the functions in the present invention, but these may vary depending on the intention of a person skilled in the art, precedent, or the emergence of new technologies. In addition, in a specific case, there is also a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the invention. Therefore, the term used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, not simply the name of the term.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in this application, it should not be interpreted in an ideal or excessively formal meaning. don't

Glaucoma analysis system using prototype analysis and FCM analysis techniques

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings. 1 is a block diagram of a system for analyzing glaucoma using circular analysis and FCM analysis techniques according to the present invention. 2 is a diagram explaining the difference between Archetypal Analysis (AA) and Fuzzy C-Means clustering (FCM Clustering) according to an embodiment of the present invention. 3 is a diagram showing 10 archetypes according to an embodiment of the present invention. 4 is a view showing that central field (VF) data of a subject is decomposed into three archetypes of AT2, AT9, and AT6 according to an embodiment of the present invention.

Referring to FIG. 1 , the glaucoma analysis system using the prototype analysis and FCM analysis techniques of the present invention includes a data acquisition unit 100, a prototype determination unit 200, a decomposition unit 300, and a glaucoma determination unit 400.

Glaucoma is a disease that causes visual field defect and visual impairment due to disorders in the optic nerve due to an increase in intraocular pressure or angiopathy. Intraocular pressure is generally determined by aqueous humor produced in the eye, and when a large amount of aqueous humor is produced or the discharge is reduced due to flow obstruction, the pressure inside the eye rises, and through this process, the intraocular pressure rises, causing glaucoma. Acute glaucoma, in which symptoms such as visual acuity decrease, headache, vomiting, and congestion appear as the intraocular pressure rapidly increases, and the optic nerve is slowly destroyed, so you do not feel any special symptoms, but feel stuffy at the end when your vision improves, leading to blindness if further progresses. chronic glaucoma, etc.

In treating such glaucoma, in the case of acute glaucoma, it is important to rapidly drop intraocular pressure to preserve the optic nerve. To this end, the intraocular pressure is lowered by instillation and administration of drugs, and after the intraocular pressure is lowered, a small hole is pierced in the iris using a laser to help the circulation and discharge of aqueous humor. In the case of chronic glaucoma, intraocular pressure-lowering drugs may be helpful to prevent further damage to the optic nerve. When intraocular pressure cannot be controlled by drug or laser treatment, glaucoma surgery can be performed, but the purpose is to control intraocular pressure, not to restore the already damaged optic nerve. Therefore, early detection of glaucoma is very important rather than special prevention, so conventionally, in addition to intraocular pressure measurement, fundus photography is used to confirm the presence or absence of defects in the optic nerve fiber layer.

However, in the present invention, considering the fact that about 50% of ganglion cells (RGC) are located within 4.5 mm of the macula, glaucoma is diagnosed and progressed by confirming changes in central visual field data that are clearly visible in the viewing direction. to be used for observation. The central field of view data is obtained by digitally formatting the central field of view of the subject through a field measurement device.

More specifically, the present invention is described as follows. First, the data acquisition unit 100 of the present invention acquires central field data of the subject. The central visual field data is not classified by diagnosis, and may be one of central visual field data of a normal person, central visual field data of a glaucoma patient, and other central visual field data of an optic neuropathy patient. However, central field data of patients with retinal diseases or intraocular opacities such as cataracts and corneal opacities may be excluded.

In addition, when the data acquisition unit 100 acquires the central visual field data for both eyes, it is characterized in that the central visual field data is converted and matched based on an arbitrary eyeball. For example, the data acquisition unit 100 may randomly acquire central-field data for the left and right eyes, or obtain central-field data for both eyes. Further, the data acquisition unit 100 may convert the central field of view data of the right eye to match the central field of view data format of the left eye when the left eye is used as a reference, and convert the central field of view data of the left eye to match the data format of the left eye when the right eye is used as a reference. can be matched with the central field data format of

Next, the archetype determining unit 200 determines a plurality of archetypes using an archetypal analysis (AA) technique. In addition, the archetype determiner 200 calculates an initial value setter 210 for setting initial values (β ₀ ) for the plurality of archetypes and a reconstruction error, and then calculates the residual sum of squares (Residual An update unit 220 for updating the initial value (β ₀ ) until Sum of Squares (RSS) is equal to or less than a preset value or reaches a maximum number of repetitions. When the update from the updater 220 is completed, the archetype determiner 200 may determine an optimal value β ₁ for the plurality of archetypes.

In other words, the central visual field data of the test subject has 68 segmented positions in the shape of the eyeball, and the visual field measurement results for each segmented position may be displayed. That is, the central field of view data may be 68-dimensional data from the point of view of perimeter measurement data analysis. Since it is quite difficult to cluster such high-dimensional data through unsupervised learning, the circularity determination unit 200 attempts to analyze the central field data, which is high-dimensional data, using an circular analysis (AA) technique. The circular shape determining unit 200 may extract a plurality of data points from the central field of view data and derive a feature pattern formed by the plurality of data points. In general, the circular analysis (AA) technique has an effect of easily finding a characteristic pattern in a high-dimensional space by representing members of a multivariate data set with a convex combination of data poles.

However, the archetype analysis (AA) technique projects a plurality of data points into an arbitrary archetype, resulting in projection loss, which can lead to inaccurate results. exist.

In this regard, in (a) of FIG. 2 , a plurality of archetypes (red _circles ) and a plurality of original data points ( gray circles), and a number of Projection Data Points (green circles) projected according to the circular analysis (AA) technique are shown. And referring to (b) of FIG. 2, the circular analysis (AA) technique is used to express the archetype as a convex combination when the optimal value (β ₁ ) for the archetype is determined. A plurality of original data points may be projected onto the convex hull composed of the archetype, and the plurality of projection data points may be generated. At this time, the convex combination coefficient of the archetype calculated from the projection data point is a decomposition coefficient. Here, the convex combination coefficients may be 3.2 and 1.8 as shown in (c) of FIG.

Next, the decomposition unit 300 performs fuzzy C-means clustering (FCM clustering) on the central field data, and then selects at least one archetype among the plurality of archetypes determined by the archetype determination unit 200. decompose into In addition, the decomposition unit 300 may include a calculation unit 310 that calculates a membership grade between the optimal value β ₁ and a data point. The decomposition unit 300 may decompose the central field of view data into at least one archetype using the membership grade.

In general, there are two main types of clustering: hard clustering and soft or fuzzy clustering. Hard clustering, such as K-means clustering, allows an object to belong to only one cluster. That is, when hard clustering is used when it is difficult to express data as one cluster, many pieces of information in the data may be lost. Soft or fuzzy clustering is designed to compensate for the disadvantages of hard clustering, and the decomposition unit 300 of the present invention uses fuzzy C-Means clustering (FCM clustering) included in soft clustering to solve the central field of view (VF). ) Data may be clustered with at least one archetype among the plurality of archetypes, rather than clustered with only a specific archetype.

Referring to (c) of FIG. 2 , the calculation unit 310 determines the membership grade of the archetype in proportion to the distance from a plurality of original data points to an arbitrary archetype. can be computed. For example, Membership Grade 4 may be calculated in proportion to the distance from any original data point to Archetype 1, and from the same original data to Archetype 2. Membership Grade 3 may be calculated in proportion to the distance.

At this time, as mentioned above, using the AA technique, a random original data point is projected onto archetype 1 and archetype 2, and projection data points ), the convex combination coefficients computed can be 3.2 and 1.8. It can be confirmed that the convex combination coefficient does not match the values 4 and 3, which are distances from an arbitrary original data point to an archetype, using Fuzzy C-Means clustering (FCM Clustering).

Therefore, the AA technique does not accurately reflect the distance between the original data point and the archetype as the original data point is projected onto the convex hull, resulting in loss. It can be defined as the projection loss.

That is, the present invention intends to apply Fuzzy C-Means clustering (FCM Clustering) together in order to minimize the projection loss generated in the archetypal analysis (AA) technique and output more accurate decomposition results.

Also, the decomposition unit 300 may output a combination of a plurality of archetypes and a ratio for each archetype as the decomposition result for the central field of view data of the subject. As shown in FIG. 4 , the decomposition result may be output as AT2 20.3%, AT9 18.7%, and AT6 16.5% after the central field data of the test subject is decomposed by the decomposition unit 300 .

Next, the glaucoma determining unit 400 determines whether or not glaucoma is present according to a decomposition result of the central visual field data. The plurality of archetypes are hyper parameters and may be preset to 10 as shown in FIG. 3 . Each archetype may represent a different ocular loss. For example, Temporal loss (AT 1), Nasal loss (AT 2), Central loss (AT 3), Temporal islands (AT 4), Differential quadrant Superotemporal quadrant sparing (AT 5), Superior loss (AT 6), Superonasal loss (AT 7), Inferior loss (AT 8), High quadrant sparing ( Inferotemporal quadrant sparing (AT 9) and intact field (AT 10) may be included.

At this time, the glaucoma determination unit 400 determines that the disassembly result is output as shown in FIG. 4 from the decomposition unit 300, and the nasal loss (AT2) is 20.3%, the upper quadrant preservation (AT9) is 18.7%, and the upper quadrant loss (AT6) is determined. In 16.5%, it can be judged as glaucoma damage.

Glaucoma analysis method using archetype analysis and FCM analysis technique

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings. 5 is a flow chart of a glaucoma analysis method using circular analysis and FCM analysis techniques of the present invention.

Referring to FIG. 5 , the glaucoma analysis method using the prototype analysis and FCM analysis techniques of the present invention includes a data acquisition step (S100), a prototype determination step (S200), a disassembly step (S300), and a glaucoma determination step (S400).

More specifically, in the data acquisition step (S100), central visual field data of the subject is acquired by the data acquisition unit 100. The central visual field data is not classified by diagnosis, and may be one of central visual field data of a normal person, central visual field data of a glaucoma patient, and other central visual field data of an optic neuropathy patient. However, central field data of patients with retinal diseases or intraocular opacities such as cataracts and corneal opacities may be excluded.

Meanwhile, in the data acquisition step (S100), when the central visual field data for both eyes are acquired, the central visual field data are converted and matched based on an arbitrary eyeball. For example, in the data acquisition step ( S100 ), central field data for the left and right eyes may be obtained randomly or central field data for both eyes may be acquired. In the data acquisition step (S100), when the left eye is used as a reference, the right eye's central field of view (VF) data is converted to match the left eye's central field of view (VF) data format, and when the right eye is used as a reference, the left eye's center field of view (VF) data is converted. The visual field data may be converted to match the format of the central visual field data of the right eye.

Next, in the archetype determination step (S200), a plurality of archetypes are determined by the archetype determiner 200 using an archetypal analysis (AA) technique. Most preferably, an optimal value (β ₁ ) for a plurality of archetypes is determined.

To this end, on the other hand, the archetype determining step (S200) is an initial value setting step (S210) in which the initial values (β ₀ ) for the plurality of archetypes are set by the initial value setting unit 210 and the update unit In step 220, after the Reconstruct Error is calculated, the initial value (β ₀ ) is updated until the residual sum of squares (RSS) is less than or equal to a preset value or the maximum number of iterations is reached. It may include an update step (S220) to be. Therefore, in the archetype determination step (S200), when the update from the update step (S220) is completed, an optimal value (β ₁ ) for the plurality of archetypes may be determined.

In other words, the central field (VF) data of the test subject has 68 segment positions in the shape of the eyeball, and the visual field measurement results for each segment position may be displayed. That is, the central field of view data may be 68-dimensional data from the point of view of perimeter measurement data analysis. Since it is quite difficult to cluster such high-dimensional data through unsupervised learning, the archetype determination step (S200) attempts to analyze the central field data, which is high-dimensional data, using an AA technique. In the circular shape determination step (S200), a plurality of data points are extracted from the central field of view data, and a feature pattern formed by the plurality of data points may be derived. In general, circular analysis (AA) techniques are very easy to find characteristic patterns in high-dimensional space by representing members of a multivariate data set as a convex combination of data poles.

However, the circular analysis (AA) technique has a problem in that a plurality of data points are projected onto an arbitrary archetype, resulting in projection loss, which can lead to inaccurate results. exist.

In this regard, referring to (b) of FIG. 2, the circular analysis (AA) technique is to express as a convex combination of the archetype when the optimal value (β ₁ ) for the archetype is determined. To this end, the plurality of original data points may be projected onto the convex hull composed of the archetype, and the plurality of projection data points may be generated. At this time, the convex combination coefficient of the archetype calculated from the projection data point is a decomposition coefficient. Here, the convex combination coefficients may be 3.2 and 1.8 as shown in (c) of FIG.

Next, in the decomposition step (S300), the central field of view (VF) data is subjected to Fuzzy C-Means clustering (FCM Clustering) by the decomposition unit 300, and then the plurality of prototypes determined from the prototype determination unit 200. (Archetype) is decomposed into at least one archetype.

Meanwhile, the decomposition step (S300) may include a calculation step (S310) in which a membership grade between the optimal value (β ₁ ) and a data point is calculated by the calculation unit 310. . In the decomposing step (S300), the central field of view data may be decomposed into at least one archetype by using the membership grade.

In general, there are two main types of clustering: hard clustering and soft or fuzzy clustering. Hard clustering, such as K-means clustering, allows an object to belong to only one cluster. That is, when hard clustering is used when it is difficult to express data as one cluster, many pieces of information in the data may be lost. Soft or fuzzy clustering is designed to compensate for the disadvantages of hard clustering, and the decomposition step (S300) of the present invention uses Fuzzy C-Means clustering (FCM Clustering) included in soft clustering to focus on the central field of view ( VF) data may be clustered into at least one archetype among the plurality of archetypes, rather than clustered only by a specific archetype.

Referring to (c) of FIG. 2, in the calculation step (S310), the membership grade of the archetype is proportional to the distance from a plurality of original data points to an arbitrary archetype. ) can be computed. For example, Membership Grade 4 may be calculated in proportion to the distance from any original data point to Archetype 1, and from the same original data to Archetype 2. Membership Grade 3 may be calculated in proportion to the distance.

At this time, as mentioned above, using the AA technique, a random original data point is projected onto archetype 1 and archetype 2, and projection data points ), the convex combination coefficients computed can be 3.2 and 1.8. It can be confirmed that the convex combination coefficient does not match the values 4 and 3, which are distances from an arbitrary original data point to an archetype, using Fuzzy C-Means clustering (FCM Clustering).

Therefore, in the AA technique, as the original data point is projected onto the convex hull, the distance between the original data point and the archetype is not accurately reflected and loss occurs. It can be defined as projection loss.

That is, the present invention intends to apply Fuzzy C-Means clustering (FCM Clustering) together in order to minimize the projection loss generated in the archetypal analysis (AA) technique and output more accurate decomposition results.

In addition, in the decomposition step (S300), a combination of a plurality of archetypes and a ratio of each archetype may be output as the decomposition result for the central field of view data of the subject. As shown in FIG. 4 , the decomposition result may be output as AT2 20.3%, AT9 18.7%, AT6 16.5% by disassembling the central field of view (VF) data of the subject from the decomposition step (S300).

In the glaucoma determination step (S400), the glaucoma determination unit 400 determines whether or not glaucoma is present according to the decomposition result of the central field (VF) data. The plurality of archetypes are hyper parameters and may be preset to 10 as shown in FIG. 3 . Each archetype may represent a different ocular loss. For example, Temporal loss (AT 1), Nasal loss (AT 2), Central loss (AT 3), Temporal islands (AT 4), Differential quadrant Superotemporal quadrant sparing (AT 5), Superior loss (AT 6), Superonasal loss (AT 7), Inferior loss (AT 8), Upper quadrant sparing (AT 8) Inferotemporal quadrant sparing (AT 9) and intact field (AT 10) may be included.

In the glaucoma determination step (S400), when the disassembly result is output as shown in FIG. 4 from the decomposition unit 300, the nasal loss (AT2) is 20.3%, the upper quadrant preservation (AT9) is 18.7%, and the upper quadrant loss (AT6) is 16.5%. % glaucomatous damage can be judged to be present.

As described above, according to the present invention, after Fuzzy C-Means clustering of central visual field data, decomposition into at least one archetype among the plurality of archetypes determined through the archetype analysis (AA) technique, and glaucoma according to the decomposition result By determining whether or not, there is an effect of determining glaucoma more objectively and consistently based on central field data that facilitates checking the location and change of ganglion cells (RGC).

Embodiments may be implemented by hardware, software, firmware, middleware, microcode, hardware description language, or any combination thereof. When implemented in software, firmware, middleware or microcode, the program code or code segments that perform necessary tasks may be stored on a computer readable storage medium and executed by one or more processors.

And aspects of the subject matter described herein may be described in the general context of computer-executable instructions, such as a program module or component executed by a computer. Generally, program modules or components include routines, programs, objects, and data structures that perform particular tasks or implement particular data types. Aspects of the subject matter described herein may be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

As described above, although the embodiments have been described with limited examples and drawings, those skilled in the art can make various modifications and variations from the above description. For example, the described techniques are performed in a different order than the described method, and/or the components of the system, structure, device, circuit, etc. described in are combined or combined in a different form than the described method, or in a different configuration. Appropriate results can be achieved even when substituted or substituted by elements or equivalents.

Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

Claims (5)

A data acquisition unit for obtaining central visual field data of the test subject; An archetype determining unit for determining a plurality of archetypes using an archetypal analysis (AA) technique; a decomposition unit for decomposing the central field data into at least one archetype among the plurality of archetypes determined by the archetype determination unit after Fuzzy C-Means clustering (FCM clustering); and Glaucoma analysis system using circular analysis and FCM analysis techniques including; According to claim 1, The circular crystal part, an initial value setting unit that sets initial values (β ₀ ) for the plurality of archetypes; and An update unit for updating the initial value (β ₀ ) until a residual sum of squares (RSS) is less than or equal to a preset number of repetitions after calculating a reconstruction error (Reconstruct Error). do, Glaucoma analysis system using archetype analysis and FCM analysis, characterized in that for determining the optimum value (β ₁ ) for the plurality of archetypes when the update is completed from the update unit. According to claim 2, the disassembly, A calculation unit for calculating a membership grade between the optimal value (β ₁ ) and a data point; includes, A glaucoma analysis system using an archetype analysis and an FCM analysis technique, characterized in that for decomposing the central field data into at least one archetype using the membership grade. According to claim 1, The data acquisition unit, Glaucoma analysis system using circular analysis and FCM analysis techniques, characterized in that when the central visual field data for both eyes are acquired, the central visual field data are converted and matched based on an arbitrary eyeball. a data acquisition step in which central visual field data of the subject is acquired by the data acquisition unit; An archetype determination step in which a plurality of archetypes are determined by using an archetypal analysis (AA) technique by the archetype determiner; a decomposition step of decomposing, by a decomposition unit, the central field data into at least one archetype among the plurality of archetypes determined by the archetype determination unit after Fuzzy C-Means clustering (FCM clustering); and A glaucoma analysis method using circular analysis and FCM analysis techniques, comprising: a glaucoma determination step of determining, by a glaucoma determination unit, whether glaucoma is present according to a decomposition result of the central visual field data.

Images