WO2022265024A1

WO2022265024A1 - Visual field testing method, visual field testing device, and visual field testing program

Info

Publication number: WO2022265024A1
Application number: PCT/JP2022/023857
Authority: WO
Inventors: ちから中村; 大志青木; 照生堀川; 日佐雄大澤
Original assignee: 株式会社ニコン
Priority date: 2021-06-15
Filing date: 2022-06-14
Publication date: 2022-12-22

Abstract

This visual field testing method comprises: a step for measuring the sensitivity of a plurality of first test points in a visual field range to obtain a first test result; a step for inferring a medical condition on the basis of the first test result; and a step for setting a plurality of second test points that are different from the plurality of first test points in the visual field range on the basis of the inferred medical condition.

Description

Visual field inspection method, visual field inspection device, and visual field inspection program

The technology of the present disclosure relates to a visual field inspection method, a visual field inspection device, and a visual field inspection program.

Japanese Patent No. 5048284 discloses a visual field test device that tests the sensitivity of an eye to be examined to a light stimulus. A visual field inspection apparatus that does not place a burden on the subject is desired.

A visual field inspection method according to a first aspect of the technology of the present disclosure includes steps of measuring sensitivities of a plurality of first inspection points in a visual field range to obtain a first inspection result; and setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case.

A visual field inspection apparatus according to a second aspect of the technology of the present disclosure includes a processor, the processor measuring sensitivities of a plurality of first inspection points in a visual field range to obtain a first inspection result; estimating a case based on one test result; and setting a plurality of second test points different from the plurality of first test points in the visual field range based on the estimated case. conduct.

A program according to a third aspect of the technology of the present disclosure is provided in a computer, measuring the sensitivity of a plurality of first inspection points in a visual field range to obtain a first inspection result, and based on the first inspection result, estimating a case; and setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case.

1 is a block diagram of an ophthalmic system 100; FIG. 2 is a block diagram showing the configuration of a perimeter 110. FIG. 3 is a functional block diagram of a CPU 22 of the perimeter 110; FIG. 2 is an explanatory diagram showing the structure of an eye 12 to be examined; FIG. 19 is an image of a region of interest 190 in a normal fundus. FIG. 3 is a schematic diagram showing an inspection point set 200, which is a set of inspection points to which index light is presented in an inspection target area 190; It is a visual field sensitivity map showing the results of a visual field test of a normal fundus. FIG. 10 is an explanatory diagram showing that initial inspection points are set from an inspection point set 200; It is an explanatory view showing the result of inspecting the initial inspection points. FIG. 9 is an explanatory diagram showing a case where additional inspection points are set based on inspection results; FIG. 11 is an explanatory diagram showing a case where additional inspection points are set; FIG. 4 is a schematic diagram showing an example of a visual field defect pan in estimation of a case; FIG. 2 is a schematic diagram showing a case where the visual field examination area is subdivided into

areas

220, 222, 224, 226, 228 and 230; 4 is a flowchart of visual field inspection processing executed by a CPU 22 of the perimeter 110. FIG. FIG. 9 is a flow chart of processing for estimating and interpolating visual field sensitivities of all inspection points in step 136 of FIG. 8. FIG. It is a figure which shows the update process of a cumulative function. FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 11 is another diagram showing update processing of the cumulative function; FIG. 10 is an explanatory diagram showing a luminance value-percentage of correct answer curve showing the relationship between the luminance value and the probability f _a,b (θ) that an inspection point of the optic nerve of the subject's eye 12 recognizes the index light of the luminance value. . FIG. 3 is a schematic diagram showing the relationship between inspection points (including uninspected points and inspected points) and the estimated brightness value of each inspection point; FIG. 10 is a schematic diagram showing an example of inspection results at each inspection point when there is no abnormality; It is the schematic which showed the example of the visual field sensitivity map in the case of no abnormality. FIG. 4 is a schematic diagram showing setting of additional inspection points when there is no abnormality; FIG. 11 is a schematic diagram showing an example of test results for each test point when the upper region 202 is a nasal perforation; FIG. 11 is a schematic diagram showing an example of a visual field sensitivity map when the upper region 202 is nasal perforation; FIG. 11 is a schematic diagram showing the setting of additional test points when the upper region 202 is a nasal perforation; FIG. 11 is a schematic diagram showing an example of test results at each test point when the upper region 202 has a temporal wedge-shaped defect. FIG. 4 is a schematic diagram showing an example of a visual field sensitivity map when the upper region 202 has a temporal wedge-shaped defect; FIG. 11 is a schematic diagram showing an example of setting additional inspection points when the upper region 202 has a temporal wedge-shaped defect. FIG. 10 is a schematic diagram showing an example of inspection results for each inspection point in the case of nose-side stairs; It is a schematic diagram showing an example of a visual field sensitivity map when the lower region 204 is the nose-side stairs. FIG. 11 is a schematic diagram showing an example of setting additional inspection points in the case of nose-side stairs; 4 is a graph showing the estimated luminance value of each inspection point in the total inspection point set; FIG. 17B is a graph obtained by adding reliability to the graph of FIG. 17A; It is a figure which shows the visual field sensitivity map 510M. FIG. 10 is a diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the prior art. FIG. 10 is another diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the prior art. FIG. 10 is another diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the prior art. FIG. 4 is a diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the present embodiment. FIG. 9 is another diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the present embodiment. FIG. 9 is another diagram showing the relationship between the number of inspections and the difference between the correct value of sensitivity and the estimated sensitivity in the present embodiment. FIG. 10 is a diagram illustrating a method of interpolating estimated luminance values of uninspected points; 4 is a graph showing estimated luminance values and reliability obtained in the present embodiment.

Hereinafter, embodiments of the technology of the present disclosure will be described in detail with reference to the drawings.

The configuration of the ophthalmic system 100 will be described with reference to FIG. As shown in FIG. 1, the ophthalmic system 100 includes a static visual field test device (hereinafter referred to as a "perimeter") 110, a management server device (hereinafter referred to as a "server") 140, and an image display device (hereinafter referred to as a " viewer") 150.

The perimeter 110 is an example of the "visual field inspection device" of the technology of the present disclosure.

The perimeter 110 is a device for examining the visual field sensitivity (brightness value) of the subject's eye to be examined, which will be described later in detail, and is used for diagnosing glaucoma, retinitis pigmentosa, and the like.

Here, the visual field sensitivity is the intensity (luminance value: luminance (dB)) of the index light that reaches the inspection point to be inspected in the optic nerve that exists in the retina of the subject's eye and is recognized by the subject. Note that the larger the luminance value expressed in dB, the smaller the intensity of the index light reaching the inspection point. In other words, the smaller the brightness value expressed in dB, the greater the intensity of the index light reaching the inspection point. That is, the larger the luminance value expressed in dB, the darker the index light, and the smaller the luminance value expressed in dB, the brighter the index light.

The server 140 stores the visual field sensitivity test results (estimated sensitivity, etc.) of the subject's eye to be examined by the perimeter 110 in association with the patient ID. The viewer 150 displays medical information such as the visual field sensitivity test results of the subject's eye acquired from the server 140 .

The perimeter 110, the server 140, and the viewer 150 are interconnected via the network 130.

The configuration of the perimeter 110 is shown in FIG.

When the perimeter 110 is installed on a horizontal plane, the horizontal direction is the "X direction", the vertical direction to the horizontal plane is the "Y direction", and the direction connecting the center of the pupil of the anterior segment of the subject's eye 12 and the center of the eyeball. Let it be "Z direction". Therefore, the X, Y and Z directions are perpendicular to each other.

As shown in FIG. 2, the perimeter 110 includes a control device 10, an index presentation section 30, an external storage device 40, an input/display section 50, and a response section 60.

The control device 10 has a CPU (Central Processing Unit) 12, a ROM (Read-Only Memory) 14, a RAM (Random Access Memory) 16, and an input/output (I/O) port 18. are interconnected by a bus 20; The ROM 14 stores a visual field inspection program, which will be described later.

The CPU 12 is an example of the "processor" of the technology of the present disclosure. A processor executes a visual field test program.

The I/O port 18 is connected to an index presentation unit 30, an external storage device 40, a communication interface (I/F) 45, an input/display unit 50, and a response unit 60.

The input/display unit 50 has a graphic operator interface that displays images and receives various instructions from the operator. Graphic operator interfaces include touch panel displays.

The response unit 60 includes a switch (not shown) operated by the subject (patient) and a transmission unit. When the subject recognizes the index light during the visual field test, which will be described later, the subject turns on the switch. When the switch is turned on, the transmitter transmits a recognition signal indicating that the subject has recognized the index light to the control device 10 .

A communication interface (I/F) 45 is connected to the server 140 and the viewer 150 via the network 130 .

The index presenting unit 30 presents indices (specifically, projects light) to a dome 30D whose hemispherical inner surface is a reflecting surface, and to inspection points at a plurality of positions on the inner surface of the dome 30D. device. Under the control of the control device 10 according to a visual field test program for visual field test, which will be described later, the projection device presents indices at a plurality of points (index presentation points) at different positions on the inner surface of the dome 30D at different times. do. The index presenting point corresponds to the retina of the subject's eye. The index light from the index presentation point reaches the examination point on the retina of the eye 12 to be examined. As described above, the subject who recognizes the index light turns on the switch, and the transmitter transmits the recognition signal to the control device 10 .

Note that, in the technology of the present disclosure, the configuration of the index presentation unit 30 is not limited to the configuration including the dome 30D and the projection device. In the technique of the present disclosure, for example, a configuration in which a point on the inner surface of the dome 30D emits light by itself, or a configuration in which an inspection point on the retina of the subject's eye 12 is directly irradiated with index light can be employed as the configuration of the index presentation unit 30. .

The server 140 and viewer 150 are equipped with a computer equipped with a CPU, RAM, ROM, etc., an input device, a display, an external storage device, and the like.

A functional block diagram of the CPU 22 of the perimeter 110 is shown in FIG. Various functions realized by executing the visual field inspection program by the CPU 22 of the perimeter 110 will be described. The visual field inspection program has an inspection point setting function, an image processing function and a processing function. When the CPU 22 executes the visual field inspection program having these functions, the CPU 22 functions as an inspection point setting section 72, an image processing section 74 and a processing section 76, as shown in FIG.

FIG. 4A is an explanatory diagram showing the structure of the eye 12 to be examined. The eyeball constituting the eye 12 to be examined has a substantially spherical shape surrounded by the sclera 170 . Inside the sclera 170 is a choroid 172 , further inside the choroid 172 is a retina 174 , and the eyeball covered with the retina 174 is filled with a gel-like vitreous body 176 .

Visual cells are arranged in a plane in the retina 174, and the visual cells convert visual images (optical information) into neural signals (electrical signals). Neural signals acquired by visual cells are transmitted from the optic disc 184 through the optic nerve 182 to the brain.

The macula 178 is an area where visual cells are densely arranged in the retina 174, and the central fovea 180 corresponding to the center of the macula 178 has the most densely arranged visual cells, so the resolution is the highest in the visual field. expensive. Also, since the optic papilla 184 is a site where the optic nerve 182 converges, there are no visual cells. As a result, the area on retina 174 where optic disc 184 resides becomes blind spot 186 .

In this embodiment, the visual field sensitivity of the subject's eye 12 is measured, and the area where visual cells are significantly arranged on the retina 174 of the fundus is defined as the inspection target area 190 .

FIG. 4B is an image of the inspection target area 190 in the normal fundus of the right eye. The horizontal line across the image is the horizontal meridian 206, which is the boundary between the upper region 202 and the lower region 204 of the visual field test area when setting additional test points based on test data.

FIG. 4C is a schematic diagram showing an inspection point set 200, which is a set of inspection points to which index light is presented, in the inspection target area 190. As shown in FIG. Since there are a large number of inspection points included in the inspection point set 200, in the present embodiment, a thinning inspection is performed in which index light is presented to some selected inspection points to obtain the reaction of the subject, and the index light is used. The luminance values of inspection points not presented are interpolated by a technique such as Gaussian process regression. In addition, there are cases in which the continuity between the upper region 202 and the lower region 204 is weak in diseases of the fundus. Therefore, when setting the additional examination points, the additional examination points are separately set in the upper region 202 and the lower region 204 according to the cases inferred from the examination results.
Here, a case means a case of visual field defect in glaucoma. Specifically, there are cases of nasal perforation, nasal steps, temporal wedge defects, and arcuate scotoma.

FIG. 4D is a visual field sensitivity map showing the results of a visual field test of the normal fundus of the right eye. In the visual field sensitivity map, areas with low luminance values indicating visual field sensitivity are shown as dark areas. A dark area 188 is also present in FIG. 4D and corresponds to the optic disc 184 which is the blind spot 186 .

5A to 5D are explanatory diagrams showing the outline of the visual field test. In this embodiment, when the visual field inspection is started, initial inspection points shown in gray are set from the inspection point set 200 as shown in FIG. 5A. Basically, the same initial test points are set regardless of the presence or absence of past diagnostic data for the subject, and whether or not there is past diagnostic data. Initial inspection points may be set according to the data. For example, if past diagnostic data describes the presence of an area with low visual field sensitivity, the initial inspection points may be set intensively in an area including the area. Also, the initial inspection points do not need to be set vertically symmetrically as shown in FIG. 5A, but may be set asymmetrically.

FIG. 5B is an explanatory diagram showing the result of inspecting the initial inspection points. In FIG. 5B, inspection points indicated by black triangles are visual field sensitivity poor points where the visual field sensitivity is equal to or less than a predetermined threshold value, and other inspection points are visual field sensitivity appropriate points where the visual field sensitivity exceeds the predetermined threshold value. be.

FIG. 5C is an explanatory diagram showing an example of setting additional inspection points based on inspection results. In FIG. 5C , the deteriorated region 210 where the visual field sensitivity defect points are concentrated in the upper region 202 is recognized, so the additional inspection points indicated by the pentagons are preferentially set in the upper region 202 rather than the lower region 204 .

FIG. 5D is an explanatory diagram showing a case where additional inspection points are set. As shown in FIG. 5D, new additional inspection points indicated by hexagons are preferentially set in areas where many visual field sensitivity defects exist.

As will be described later, in this embodiment, cases of disease in the subject's eye to be examined are estimated based on the distribution of points of poor visual field sensitivity. FIG. 6 is a schematic diagram showing an example of a visual field defect pan in estimating a case. For example, if the visual field sensitivity defect point 300 exists in the area 212 close to the nose side of the right eye, it is estimated to be nasal perforation. Also, if the visual field sensitivity defect point 302 exists in the region 214 close to the ear side of the right eye, it is estimated to be an ear-side wedge-shaped defect.

Also, the visual field test area may be subdivided into

areas

220, 222, 224, 226, 228, and 230, for example, as shown in FIG.

FIG. 8 shows a flowchart of visual field inspection processing executed by the CPU 12 of the perimeter 110 . The visual field inspection process shown in the flowchart of FIG. 8 is realized by the CPU 12 executing the visual field inspection program. The visual field inspection process is started when the operator operates a start button (not shown) displayed on the input/display unit 50 .

At step 100 , the image processing unit 74 displays an input screen for the patient ID on the input/display unit 50 . A patient ID is input to the input/display unit 50 by the operator. At step 102, the processing unit 76 acquires a patient ID.

At step 104, the inspection point setting unit 72 inquires of the server 140 whether the visual field sensitivity inspection result corresponding to the acquired patient ID is stored. That is, an inquiry is made as to whether the visual field sensitivity test result is stored in correspondence with the acquired patient ID. The inspection point setting unit 72 acquires the inquiry result from the server 140, and determines whether or not there is past data of the visual field sensitivity inspection result corresponding to the patient ID based on the acquired inquiry result. The past data is data corresponding to each subject and inspection points of each subject. The past data are, for example, the patient's visual field sensitivity, the estimated sensitivity of each inspection point, the number of inspections, and a cumulative function, which will be described later. The past data may be acquired data of all inspections performed in the past, or may be data updated after the latest inspection.

If it is determined in step 104 that there is past data of visual field sensitivity test results corresponding to the patient ID, then in step 106 the processing unit 76 extracts the test results from the past data corresponding to the input patient ID. The latest estimated sensitivity (visual field sensitivity) and the number of inspections of each inspection point in the point set 200 are read. On the other hand, if it is determined in step 104 that there is no past data of visual field sensitivity test results corresponding to the patient ID, then in step 108 the processing unit 76 predefines each test point in the test point set 200. read the specified luminance value. The predetermined brightness value is, for example, a reference value for each inspection point in the inspection point set 200 for normal eyes.

At step 110, the inspection point setting unit 72 sets a set of initial inspection points as shown in FIG. 5A. Then, at step 112 , the inspection point setting unit 72 selects the luminance value of the index light to be presented at the initial inspection point set at step 110 . In step 112, the brightness value may be selected randomly, selected by an operator, or automatically selected based on historical data, for example.

At step 114, the inspection point setting unit 72 initializes the cumulative function. The cumulative function is a function that indicates the relationship between the luminance value of the index light and the number of inspections. More specifically, it is a function that associates the luminance value of each index light with the cumulative number of times used in inspection. . The initialization of the cumulative function is a process of zeroing the number of inspections corresponding to each luminance value as shown in FIG. 10A. The cumulative function will be described later with reference to FIGS. 10A-10I.

At step 116 , the inspection point setting unit 72 selects one inspection point from the set of initial inspection points set at step 110 . This inspection point may be randomly selected from a set of initial inspection points, may be selected by an operator, or may be automatically selected based on historical data.

At step 118 , the inspection point setting unit 72 acquires the cumulative number of inspections of the inspection point selected at step 116 . The cumulative number of inspections can be extracted from a cumulative function to be described later, but the cumulative number of inspections may be held as data independent of the cumulative function.

At step 120, it is determined whether or not the cumulative number of inspections is 1 or more. At step 120, if the cumulative number of inspections is one or more, the procedure proceeds to step 122, and if the cumulative number of inspections is not one or more, the procedure proceeds to step .

At step 122, the inspection point setting unit 72 presents index light to the inspection point selected at step 116 with a luminance value based on the cumulative function. The inspection point setting unit 72 sets the brightness value of the index light to be presented from the range of brightness values extracted from the cumulative function. In the technique of the present disclosure, the luminance value of the index light to be presented may be set by randomly extracting from the extracted luminance value range, or may be set by extracting an arbitrarily determined value. . For example, the inspection point setting unit 72 may extract the brightness value of the index light to be presented from the range, such as the median value or the 3/4 value in the range. Next, the inspection point setting unit 72 controls the projection device so that the index light is incident on the inspection point selected in step 116 with the extracted luminance value.

In step 124, the subject is presented with the index light having the initial brightness value set in step 112.

At step 126, the inspection point setting unit 72 acquires the subject's reaction. When the subject recognizes the index light presented in step 126, the subject switches on the responder 60. FIG. A recognition signal is thereby sent to the control device 10 . If the subject does not recognize the indicator light even if it is presented, the subject does not turn on the switch of the response unit 60 . The inspection point setting unit 72 determines whether or not the subject recognizes the index light based on whether or not the recognition signal has been transmitted before a predetermined time has elapsed since the index light was presented. For example, if the recognition signal is transmitted before the elapse of a predetermined time from when the index light is presented, the inspection point setting unit 72 may indicate that the subject recognizes the index light. to get If the recognition signal is not transmitted even after the predetermined time has passed, the subject's reaction that the subject did not recognize the index is acquired. The inspection point setting unit 72 saves the subject's reaction acquired in step 126 in the external storage device 40 .

At step 128, the inspection point setting unit 72 updates the cumulative function. In this embodiment, the processing from step 116 to step 128 is repeated, and the update processing of the cumulative function in step 128 is also repeated. Also, the cumulative number of inspection points is updated. By repeating the processing from step 116 to step 128, index lights with different brightness values are presented a plurality of times to each inspection point belonging to the set of inspection points set in step 110, and the subject's response to each indicator light is presented. , update the cumulative function corresponding to each test point based on the subject's response at each presentation. A specific description will be given below.

As for the cumulative function, if there is no past data, the cumulative number of inspections before inspection is 0 for each luminance value as shown in FIG. 10A, and the cumulative function does not exist. By the processing of step 124, for example, the subject is presented with index light with an initial luminance value of 28 dB. When the subject does not recognize the index light, the inspection point setting unit 72 sets the number of inspections for each luminance value in the range defined by the boundary of 28 dB, that is, the range of 28 dB or more, as shown in FIG. 10B. , is increased by a predetermined amount. The predetermined amount to be increased is 1, for example. Therefore, as shown in FIG. 10B, the number of inspections for each luminance value in the range of 28 dB or more is one.

Here, in FIG. 10B, although only the luminance value of 28 dB is presented, the reason why the number of inspections for each luminance value in the range of 28 dB or more is 1 is as follows. If the subject does not recognize the index light of 28 dB, it is presumed that the subject cannot recognize the index light with a luminance value higher than 28 dB, that is, the light darker than the presented index light. Therefore, it is presumed that the subject's reaction would be that he could not recognize the index with a luminance value greater than 28 dB for this inspection point where the index light of 28 dB was presented. Therefore, for a luminance value greater than 28 dB, the subject's reaction is assumed without actually being tested, so the number of times of testing is increased by 1 assuming that it has been tested. The predetermined amount to be increased may be a different value depending on each luminance value. For example, when the subject does not recognize the index light of 28 dB, the number of inspections is increased by 1 for each luminance value in the range of 28 dB to less than 32 dB, and the number of inspections is increased by 2 for each luminance value in the range of 32 dB to less than 36 dB. The number of tests may be increased by 3 for each luminance value in the range of 36 dB or more. This is based on the assumption that the examinee will not be able to recognize all of the 28 dB index light, as well as the 32 dB index light and the 36 dB index light that are presented in a pseudo manner. This corresponds to performing an operation of increasing the number of inspections for each luminance value by 1 with respect to the inspection of .

At step 128, if the number of inspections is the first, the cumulative function is updated as shown in FIG. 10B.

In step 130, it is determined whether or not the visual field sensitivity can be estimated with sufficient accuracy for the set of inspection points set in step 116. In this embodiment, as will be described later, the cumulative function draws a downwardly convex linear shape as shown in FIG. . When the number of inspections is 1, the cumulative function does not have a downwardly convex linear shape as shown in FIG. 10B, so the visual field sensitivity cannot be estimated with sufficient accuracy.

In order to determine in step 130 whether the visual field sensitivity can be estimated with sufficient accuracy, index lights with different luminance values are presented to the inspection points selected in step 116 to acquire the subject's reaction. The cumulative function is updated according to the brightness value of the index light and the subject's reaction.

For example, when the result shown in FIG. 10B is obtained, as shown in FIG. It is necessary to search (test) whether the range of values is recognized.

Therefore, by repeating the above, in step 122, from the cumulative function (see FIG. 10B), for example, 16 dB is extracted from the range of luminance values in which the number of inspections is less than 1, as shown in FIG. is presented. Suppose that the subject did not recognize the index light with a luminance value of 16 dB. In this case, for the reason described above, the inspection point setting unit 72 increases the number of inspections for each luminance value within the range defined by 16 dB (range of 16 dB or more) by a predetermined amount (for example, 1). Therefore, as shown in FIG. 10D, the cumulative function is updated so that the number of inspections is 1 for luminance values from 16 dB to less than 28 dB, and the number of inspections is 2 for luminance values in the range of 28 dB or more. In this case, as shown in FIG. 10E, it is necessary to search (check) whether or not a range of luminance values smaller than 16 dB can be recognized.

By repeating the processing from step 120 to step 128 in this manner and updating the cumulative function as shown in FIGS. . As a result, the upper and lower limits of the range of luminance values that need to be searched (inspected) are determined, and a downwardly convex linear cumulative function can be created.

When the cumulative function of the inspection point selected in step 116 is in the state shown in FIG. 10I, next, the index light with a brightness value of 12 dB is presented to the inspection point selected in step 116 to obtain the subject's reaction. do. However, if the cumulative function of the inspection point selected in step 116 is as shown in FIG. 10I, it is expected that the subject's visual field sensitivity value at this inspection point is around 12 dB. In this way, it is determined whether the visual field sensitivity of the inspection point selected in step 116 can be estimated.

The examples described above using FIGS. 10A to 10I are examples in which there is no past data as described above. On the other hand, if it is determined in step 104 of FIG. 8 that there is past data, there is a cumulative function and an estimated value of visual field sensitivity for each inspection point in the inspection point set 200 corresponding to the patient ID. In this case, when the process of step 104 in FIG. 8 is executed, step 104 is determined affirmatively, and step 122 determines the cumulative function and visual field sensitivity estimated value for each inspection point in the inspection point set 200 in the past data. Either or both are used. As for the method of selecting the brightness value of the index term to be presented, for example, when the brightness value with the smallest value of the cumulative function is 32 dB and the estimated value of the visual field sensitivity is 28 dB, the average value of 30 dB is presented.

At step 132, the inspection point setting unit 72 determines whether or not the determination conditions are satisfied. The determination condition in step 132 is whether or not it is determined in step 128 that the visual field sensitivity can be estimated for all inspection points belonging to the set of initial inspection points set in step 110 . If it is determined in step 132 that visual field sensitivities can be estimated for all inspection points, the process proceeds to step 134;

At step 134, the inspection point setting unit 72 reads the inspection data obtained in the procedure up to step 132. Then, in step 136 , the inspection point setting unit 72 determines the overall visual field sensitivity, that is, the visual field sensitivity for each inspection point in the inspection point set 200 based on the cumulative function obtained for each inspection point in the inspection point set 200 . Perform data interpolation to estimate sensitivity (estimated luminance value).

In this embodiment, the inspection point setting unit 72 estimates the visual field sensitivity (estimated brightness value) for each inspection point in the inspection point set 200 including the initial inspection points. The inspection point setting unit 72 uses a stochastic process to estimate an estimated brightness value of an uninspected point from the estimated brightness value of each inspected point, and sets a confidence value representing the likelihood of the estimated brightness value of the estimated uninspected point. Estimate degrees.

In this embodiment, the inspection point setting unit 72 numerically obtains the estimated luminance value of the uninspected points. The stochastic process in this case is called a "stochastic field". In this embodiment, the inspection point setting unit 72 uses a stochastic process or a random field to estimate the estimated brightness value of an uninspected point from the estimated brightness value of each inspected point, and uses the stochastic process or random field to , to estimate the reliability of the estimated brightness values of the estimated uninspected points.

In this embodiment, Gaussian Process Regression (GPR) is used as the stochastic process. Note that the technique of the present disclosure is not limited to Gaussian process regression as the stochastic process. Another example of a stochastic process is t-process regression.

As described above, the reliability is numerical data representing the likelihood of the estimated luminance value of the estimated uninspected points. Specifically, the reliability is numerical data indicating the possible range of the visual field sensitivity of the uninspected point centered on the estimated luminance value of the uninspected point.

FIG. 9 shows a flowchart of the process of estimating and interpolating the visual field sensitivity of the entire inspection points in step 136 of FIG. As shown in FIG. 9, at step 301, the inspection point setting unit 72 calculates the estimated brightness value of each inspected point. Specifically, first, in order to calculate the estimated luminance value (visual field sensitivity) of each inspected point, the inspection point setting unit 72 sets the index light of the luminance value with respect to the luminance value shown in FIG. A luminance value-percentage of correct answer curve showing the relationship between the probability f _a,b (θ) of recognition by the inspection point of the optic nerve of the subject's eye 12 is used. The luminance value-correct answer rate curve is defined by the following formula. In the following formula, θ is the brightness value used in each inspection at the inspection point.

As shown below, a is a constant greater than 0, and b is a value included in R (R is the set of all real numbers).

The inspection point setting unit 72 obtains the likelihood L(a, b) of each inspected point from the following equation using the equation representing the curve. More specifically, for example, in the first inspection at a certain inspected point, the luminance value is 20 dB, and if the reaction of the subject is recognized (Yes), the inspection point setting unit 72 Use the probability f _a,b (20). In the second inspection, when the luminance value is 24 dB and the subject's reaction is not recognized (No), the inspection point setting unit 72 sets (1-f _a,b (24)). use. In the third inspection, the luminance value is 16 dB, and if the subject's reaction is recognized (Yes), the inspection point setting unit 72 uses the probability f _a,b (16). For each inspected point, the inspection point setting unit 72 uses the product of values corresponding to the results of all inspections as described above, and obtains b that maximizes the likelihood (a, b).

Let b obtained for each inspected point be the estimated luminance value of each inspected point. For example, at the inspected point (x ₂ ), 21 dB is calculated as the estimated luminance value, at the inspected point (x ₃ ), 19.6 dB is calculated as the estimated luminance value, and at the inspected point (x ₅ ) , 31.5 dB is calculated as an estimated luminance value.

At step 303, the inspection point setting unit 72 uses Gaussian process regression to calculate the estimated luminance value of each uninspected point that has not been inspected. Specifically, for each uninspected point, the inspection point setting unit 72 uses the estimated brightness value of each inspected point to calculate an estimated brightness value E[X(x*)|D] from the following equation. do.

K(x, x') of k* _D in _Equation 4 and K(x, x') in KD in Equation 5 is the following Gaussian RBF kernel (radial basis function kernel). In the formula below, θ ₁ and θ ₂ are real numbers.

Each of x ₁ , x ₂ , . . . Denote X _N and x ₁ , x ₂ , . . . Each of _XN is the XY coordinates of the location of the inspected point.

The x' of K(x, x') is the XY coordinates of each uninspected point x*.

_y1 , y2, . . . _YN is the estimated luminance value of each inspected point.

The estimated luminance value E[X(x*)|D] of each uninspected point obtained as described above is the average luminance value of each uninspected point estimated from the estimated luminance value of each inspected point. value.

In step 305, using Gaussian process regression, the inspection point setting unit 72 determines the reliability of the estimated brightness value of each uninspected point that has not been inspected, from the following equation: variance V[X(x*)|D] Calculate

k _** is as follows.

When the processing of step 305 ends, the processing of step 136 in FIG. 8 ends.

FIG. 12 is a diagram showing the relationship between inspection points (including uninspected points and inspected points) and the estimated luminance value of each inspection point. When the process of step 136 in FIG. 8 is completed, for example, as shown in FIG. 12, the estimated brightness values of the inspected points (x ₂ , x ₃ , x ₅ , . . . ) are obtained, and the uninspected points ( _x1 , _x4 , _x6 , _x7 , _x8 , ...) and their reliability are obtained. In FIG. 12, the colored area around the curve indicates the range of error. If the width of the colored area is small, the reliability of the estimated luminance value is high, and if the width of the colored area is large, the reliability of the estimated luminance value is high. low.

As shown in FIG. 12, for example, uninspected point (x ₄ ) is adjacent to inspected point (x ₃ , x ₅ ) and relatively close to inspected point (x ₃ , x ₅ ). However, the unchecked point (x ₈ ) is relatively far from the checked point (x ₅ ). Therefore, the error range of the estimated luminance value of the uninspected point (x ₄ ) is relatively small and the reliability value is high, and the error range of the estimated luminance value of the uninspected point (x ₈ ) is relatively large and the reliability is high. low.

At step 138, the inspection point setting unit 72 determines whether additional inspection is necessary. The necessity of additional inspection is determined based on the error range of the estimated luminance value. In step 138, if there is an inspection point whose error range exceeds a predetermined range among the inspection points for which the estimated brightness value has been calculated, it is determined that additional inspection is necessary. The predetermined range in step 138 is specifically determined through a test of estimated luminance value calculation.

If it is determined in step 138 that additional inspection is not required, the process is terminated, and if additional inspection is required, the procedure proceeds to step 140.

At step 140, the inspection point setting unit 72 sets a set of additional inspection points. In this embodiment, the glaucoma case of the subject is estimated based on the test results of the initial test points, and additional test points are set according to the estimated case. Case estimation and setting of additional test points may be performed using, for example, the control device 10 that has been machine-learned by an RNN (Regressive Neural Network) or the like. In machine learning using RNN or the like, the control device 10 is trained by using test results for each case as teacher data.

FIG. 13A is a schematic diagram showing an example of inspection results for each inspection point when there is no abnormality. If all of the initial test points indicated by squares in FIG. 13A are normal, it is assumed that there is only a dark area 188 corresponding to the optic papilla 184, which is the blind spot 186, in the visual field sensitivity map shown in FIG. 13B. Then, as shown in FIG. 13C, the same number of additional inspection points indicated by pentagons are set in the upper area 202 where the initial inspection points are sparse and the lower area 204 where the initial inspection points are sparse.

FIG. 14A is a schematic diagram showing an example of test results for each test point when the upper region 202 is nasal perforation. As shown in FIG. 14A, if the visual field sensitivity defect indicated by the triangle is prominent in the nasal region 202N of the upper region 202, there is a possibility that the nasal perforation of the upper region 202 has occurred.

FIG. 14B is a schematic diagram showing an example of a visual field sensitivity map when the upper region 202 is a nasal perforation. If there is a possibility of nasal perforation of the upper region 202 in step 140, it is assumed that the visual field sensitivity points are distributed as shown in FIG. 14B. Then, as shown in FIG. 14C, the same number of additional inspection points indicated by pentagons are set in the upper area 202 where the initial inspection points are sparse and the lower area 204 where the initial inspection points are sparse.

FIG. 15A is a schematic diagram showing an example of inspection results at each inspection point when the upper region 202 has a temporal wedge-shaped defect. As shown in FIG. 15A , when the visual field sensitivity defect points indicated by triangles are conspicuous in the ear-side region 202E of the upper region 202, there is a possibility that the ear-side wedge-shaped defect of the upper region 202 is present.

FIG. 15B is a schematic diagram showing an example of a visual field sensitivity map when the upper region 202 has a temporal wedge-shaped defect. If in step 140 there is a possibility of an ear-side wedge-shaped defect in the upper region 202, it is assumed that the visual field sensitivity defect points are distributed as shown in FIG. 15B. Then, as shown in FIG. 15C, additional test points are set in the upper region 202 where the initial test points are sparse and the lower region 204 where the initial test points are sparse. Since the sensitivity of the visual field region on the side is low, the inspection point setting unit 72 preferentially sets additional inspection points indicated by pentagons in the ear-side region 204E of the upper region 202 .

FIG. 16A is a schematic diagram showing an example of inspection results for each inspection point in the case of nose-side stairs. As shown in FIG. 16A , if the visual field sensitivity defect points indicated by triangles are conspicuous in the nose-side region 204N of the lower region 204, there is a possibility that the nose-side steps of the lower region 204 are present.

FIG. 16B is a schematic diagram showing an example of a visual field sensitivity map when the lower region 204 is the nose-side stairs. In step 140, if there is a possibility of nasal steps in the lower region 204, it is assumed that the visual field sensitivity points are distributed as shown in FIG. 16B. Then, as shown in FIG. 16C, additional inspection points indicated by pentagons are set in the upper area 202 where the initial inspection points are sparse and the lower area 204 where the initial inspection points are sparse. In addition to the above-described nasal perforation, temporal wedge-shaped defect, and nasal staircase, we also determined diseases such as arcuate scotoma, paracentral scotoma, horizontal hemianopia-like visual field, and central residual visual field. However, additional test points may be set according to the determined disease.

At step 142 , the inspection point setting unit 72 selects the brightness value of the index light to be presented at the additional inspection points set at step 140 . In step 142, the brightness value may be selected randomly, selected by an operator, or automatically selected based on historical data, for example.

In step 144, one inspection point is selected from the set of additional inspection points set in step 140. The inspection points to be selected may be randomly selected from a set of additional inspection points, may be selected by the operator, or may be automatically selected based on historical data.

At step 146, the cumulative number of inspections for one inspection point selected at step 144 is obtained. The cumulative number of inspections can be extracted from the aforementioned cumulative function, but the cumulative number of inspections may be held as data independent of the cumulative function.

At step 148, it is determined whether or not the cumulative number of inspections is 1 or more. At step 148, if the cumulative number of inspections is 1 or more, the procedure proceeds to step 150; otherwise, the procedure proceeds to step 152.

In step 150, index light is presented to the inspection point selected in step 144 with a luminance value based on the cumulative function. The inspection point setting unit 72 sets the brightness value of the index light to be presented from the range of brightness values extracted from the cumulative function. In the technique of the present disclosure, the luminance value of the index light to be presented may be set by randomly extracting from the extracted luminance value range, or may be set by extracting an arbitrarily determined value. . For example, the inspection point setting unit 72 may extract the brightness value of the index light to be presented from the range, such as the median value or the 3/4 value in the range. Next, the inspection point setting unit 72 controls the projection device so that the index light is incident on the inspection point selected in step 144 with the extracted luminance value.

At step 152, the subject is presented with the index light having the initial brightness value set at step 142.

At step 154, the inspection point setting unit 72 acquires the subject's reaction. When the subject recognizes the index light presented in step 154, the subject switches on the response unit 60. FIG. A recognition signal is thereby sent to the control device 10 . If the subject does not recognize the indicator light even if it is presented, the subject does not turn on the switch of the response unit 60 . The inspection point setting unit 72 determines whether or not the subject recognizes the index light based on whether or not the recognition signal has been transmitted before a predetermined time has elapsed since the index light was presented. For example, if the recognition signal is transmitted before the elapse of a predetermined time from when the index light is presented, the inspection point setting unit 72 may indicate that the subject recognizes the index light. to get If the recognition signal is not transmitted even after the predetermined time has passed, the subject's reaction that the subject did not recognize the index is acquired. The inspection point setting unit 72 saves the subject's reaction acquired in step 154 in the external storage device 40 .

At step 156, the inspection point setting unit 72 updates the cumulative function. In this embodiment, the processing from step 144 to step 156 is repeated, and the update processing of the cumulative function in step 156 is also repeated. Also, the cumulative number of inspection points is updated. By repeating the processing from step 144 to step 156, index lights with different brightness values are presented a plurality of times to each inspection point belonging to the set of additional inspection points set in step 140, and the subject's response to each indicator light is presented. , the cumulative function corresponding to each test point is updated based on the subject's response at each presentation, as in step 128 above.

At step 158, the inspection point setting unit 72 determines whether the visual field sensitivity of the inspection point selected at step 144 can be estimated with sufficient accuracy. In this embodiment, as in step 130, the cumulative function draws a downwardly convex linear shape as shown in FIG. 10I, and it is determined whether or not the brightness of the index light recognized by the subject can be estimated.

At step 160, the inspection point setting unit 72 determines whether or not the determination conditions are satisfied. The judgment condition in step 160 is whether or not the visual field sensitivity performed in step 158 can be estimated for the additional inspection points set in step 140 . At step 140, if the visual field sensitivity can be estimated for the additional inspection points, then go to step 162;

In step 162, the inspection data obtained in the procedure up to step 160 are read. Then, in step 164 , the inspection point setting unit 72 determines the overall visual field sensitivity, that is, the visual field sensitivity for each inspection point in the inspection point set 200 based on the cumulative function obtained for each inspection point in the inspection point set 200 . Data interpolation for estimating sensitivity (estimated luminance value) is performed in the same manner as in step 136 described above.

At step 166, the image processor 74 computes the cumulative function of each inspected point and each additional inspection point, if performed, the estimated luminance value of each inspection point in the total inspection point set, and an estimate of the uninspected points. Create screen data for visualizing the reliability of luminance values.

Specifically, as the screen data, first, as shown in FIG. 17A, there is a graph showing the estimated brightness value of each inspection point in the set of all inspection points.

Secondly, as shown in FIG. 17B, a graph obtained by adding the reliability of the estimated brightness value at the uninspected point to the graph showing the estimated brightness value of each inspection point in the set of all inspection points, centering on the visual field sensitivity. is.

Thirdly, as shown in FIG. 18, it is a visual field sensitivity map. A visual field sensitivity map is an example of a visual field sensitivity distribution display method. The visual field sensitivity map displays the distribution of visual field sensitivity data of a plurality of inspection points included in the inspection point set. A visual field sensitivity map may be generated for the entire inspection point set, or may be generated for a part of the inspection point set. Note that the visual field sensitivity map also includes reliability data. Specifically, as shown in FIG. 18, the visual field sensitivity map 510M is a map in which * is added to an image simulating the fundus of the subject's eye 12, and inspection points whose visual field sensitivity is less than a predetermined value (dB). This is screen data. Also, the visual field sensitivity map 510M is screen data that can display a range 510 of inspection points whose reliability is equal to or higher than a predetermined value with a dotted line. Further, for example, among the inspection points in the set of all inspection points, uninspected points are displayed in a color different from that of inspected points and additional inspection points. , screen data that can display the reliability of the estimated brightness value of the uninspected point.

As described above, in this embodiment, by updating the cumulative function for each inspection point in the inspection point set 200, the range of luminance values that need to be searched (inspected) is gradually narrowed down. , it is possible to complete the visual field test in a shorter time than in the case of randomly setting the luminance value of the index light. The effectiveness of this embodiment using the cumulative function will be described below with reference to FIGS. 19A to 20C. In each of FIGS. 19A to 20C, the horizontal axis indicates the number of inspections, and the vertical axis indicates the difference between the correct value of sensitivity and the estimated sensitivity. 0 on the vertical axis means that there is no difference between the correct value and the estimated value, and the estimated value is highly accurate.

FIGS. 19A to 19C show the required number of inspections to ensure sufficient accuracy in the visual field test results when randomly selecting light intensities at three different points on the optic nerve of the fundus of the subject's eye. there is FIGS. 20A to 20C show the number of inspections required to ensure sufficient accuracy when the method of this embodiment is applied to three different points on the optic nerve of the fundus of the subject's eye.

70 to 80 inspections are required when randomly selecting luminance values (FIGS. 19A to 19C). 15 tests are sufficient. Therefore, it can be understood that the method of the present embodiment contributes to reduction in the number of inspections.

In the case of the method of linearly interpolating the estimated brightness values of uninspected points as shown in FIG. 21, the reliability of the interpolated values is not considered. For example, in the case of a method of interpolating by connecting the estimated brightness values of two uninspected points with a straight line or using a spline method (interpolation using a polynomial), the reliability of the interpolated value is not considered.

On the other hand, in the present embodiment, the reliability of the estimated brightness value is estimated along with the estimated brightness value of the uninspected points. Specifically, in FIG. 22, the horizontal axis indicates each inspection point, and the vertical axis indicates the corresponding estimated brightness value of each inspection point. The dotted line is the correct value line, the solid line is the estimated luminance value line, and the reliability is shown as a width around the estimated luminance value. As shown in FIG. 22, the degree of reliability is shown, so according to this embodiment, the operator can recognize the likelihood of the estimated luminance value.

In the embodiment described above, Gaussian process regression is used as the stochastic process, and the Gaussian RBF kernel is used. In the technology of the present disclosure, the stochastic process is not limited to Gaussian process regression, and for example, the following polynomial kernel may be used. In the formula below, c is a real number and p is a positive integer.

Also, the following Matern kernel may be used. In the equations below, Kv is a modified Bessel function of the second kind, v is a real number, and Γ(v) is a gamma function.

At step 136 in FIG. 8 of the present embodiment, the inspection point setting unit 72 estimates the estimated luminance value of the uninspected points, and estimates the reliability of the estimated luminance values of the estimated uninspected points. The technology of the present disclosure is not limited to this. The inspection point setting unit 72 calculates the range of possible values of the estimated brightness value of each uninspected point from the estimated brightness value of each inspected point, and determines the value (for example, the central value) within the calculated range. , may be calculated as the estimated luminance value of each uninspected point.

Although each stochastic process described above is the same for each inspection point in the entire inspection point set, the technology of the present disclosure is not limited to this. For example, different stochastic processes may be used for the central region of a predetermined range including the center of the fundus and for the peripheral region around the central region.

In addition, the untested point for which the estimated brightness value is interpolated is located in the range where the index light reaches through the pupil of the eye 12 to be inspected, but the range adjacent to the reach range, that is, the index light does not reach In other words, the estimated luminance value may be estimated even for points at positions where visual field inspection is not possible.

In each of the examples described above, the case where visual field test processing is realized by a software configuration using a computer has been exemplified, but the technology of the present disclosure is not limited to this. For example, instead of a software configuration using a computer, image processing may be performed only by a hardware configuration such as FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit). A part of the image processing may be performed by a software configuration, and the rest of the image processing may be performed by a hardware configuration.

In this way, the technology of the present disclosure includes the following technology, as it includes cases where visual field test processing is realized by software configuration using a computer and cases where it is not.

(First technology)
a processing unit that measures the sensitivity of a plurality of first inspection points in a visual field range and obtains a first inspection result;
setting an inspection point for estimating a case based on the first inspection result and setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case; Department and
A visual field test device comprising:

(Second technology)
a processing step in which the processing unit measures the sensitivity of a plurality of first inspection points in the visual field range to obtain a first inspection result;
An inspection point setting unit estimates a case based on the first inspection result, and determines a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case. an inspection point setting step for setting
A visual field test method comprising:

The following technique is proposed from the above disclosure.
(Third technology)
A computer program product for testing a visual field, comprising:
The computer program product comprises a computer readable storage medium that is not itself a transitory signal;
The computer-readable storage medium stores a program,
Said program
to the computer,
measuring the sensitivity of a plurality of first inspection points in the field of view to obtain a first inspection result;
estimating a case based on the first test result;
setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case;
A computer program product that causes the execution of

Note that the control device 10 is an example of the "computer program product" of the technology of the present disclosure.

The visual field inspection processing described above is just an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps added, and the order of processing may be changed without departing from the scope of the invention. Further, the technology disclosed in the present specification is a method for inspecting an eye to be inspected, in which light is presented at a plurality of light intensities to an inspection point set on the retina of the eye to be inspected, and the retina detecting the sensitivity at the inspection point of; estimating the sensitivity at a site other than the inspection point based on the detected sensitivity at the inspection point; and evaluating the reliability of the estimated sensitivity and a method comprising:

All publications, patent applications and technical standards mentioned herein, as if each individual publication, patent application and technical standard were specifically and individually indicated to be incorporated by reference. incorporated herein by reference.

The disclosure of Japanese Patent Application No. 2021-099540 is incorporated herein by reference in its entirety. In addition, all publications, patent applications, and technical standards mentioned herein are to the same extent as if each individual publication, patent application, or technical standard were specifically and individually noted to be incorporated by reference. , incorporated herein by reference.

Claims

measuring the sensitivity of a plurality of first inspection points in the field of view to obtain a first inspection result;
estimating a case based on the first test result;
setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case;
A visual field test method comprising:
measuring the sensitivity of the plurality of second test points to obtain a second test result;
2. The visual field inspection method according to claim 1, wherein a visual field sensitivity map is generated based on said first inspection result and said second inspection result.
The visual field inspection method according to claim 2, wherein the case is estimated based on a distribution pattern of inspection points whose sensitivity is equal to or less than a predetermined threshold in the first inspection result.
4. The case according to claim 3, wherein the case to be estimated is any one of nasal perforation, nasal step, temporal wedge-shaped defect, arcuate scotoma, paracentral scotoma, horizontal hemianopia-like visual field, and central residual visual field. visual field test method.
The visual field range includes a plurality of third inspection points other than the first inspection point and the second inspection point,
The field of view according to any one of claims 2 to 4, wherein the sensitivity of the plurality of third inspection points is estimated using a stochastic process based on the first inspection result or/and the second inspection result. Inspection methods.
The visual field testing method according to claim 5, wherein the stochastic process is a Gaussian process.
The visual field inspection method according to any one of claims 1 to 6, wherein the plurality of first inspection points are a plurality of preset initial inspection points.
with a processor
The processor
determining the sensitivity of a plurality of first test points in the field of view to obtain a first test result;
estimating a case based on the first test result;
setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case;
A visual field tester.
to the computer,
measuring the sensitivity of a plurality of first inspection points in the field of view to obtain a first inspection result;
estimating a case based on the first test result;
setting a plurality of second inspection points different from the plurality of first inspection points in the visual field range based on the estimated case;
A visual field test program that runs a