Eye movement testing device

The eye movement testing device uses a learning model to analyze numerical data from Hess chart images, enabling accurate disease identification even for inexperienced doctors, thus overcoming the requirement for expert diagnosis.

JP7594714B2Active Publication Date: 2024-12-05THINK OUT INC
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2022500222
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-02-10
Filing Date
2020-09-11
Publication Date
2024-12-05
Estimated Expiration
2040-09-11

AI Technical Summary

Technical Problem

Diagnosis of eye movement disorders using the Hess chart requires expertise, making it difficult for inexperienced doctors to accurately determine the underlying disease.

Method used

An eye movement testing device that includes an image acquisition section, data calculation section, and disease name acquisition section, utilizing a learning model to analyze numerical data on deviations between basic coordinates and light spots in Hess chart images, allowing for accurate disease identification.

Benefits of technology

Enables inexperienced doctors to accurately diagnose eye movement disorders by estimating disease names with high precision through numerical data analysis, reducing the need for direct input of Hess chart images into the learning model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007594714000001
    Figure 0007594714000001
  • Figure 0007594714000002
    Figure 0007594714000002
  • Figure 0007594714000003
    Figure 0007594714000003
Patent Text Reader

Abstract

This ocular movement inspection device comprises an image acquisition unit, a data calculation unit, a disease name acquisition unit, and a display unit. The image acquisition unit acquires one or more Hess chart images. For each of basic coordinates in said one or more Hess chart images acquired by the image acquisition unit, the data calculation unit calculates numerical data related to a deviation between each basic coordinate and a light spot. The disease name acquisition unit inputs, to a learning model, deviation data obtained by arranging, according to a predetermined rule, the pieces of numerical data for each of the basic coordinates acquired by the data calculation unit, and acquires, from the learning model, a disease name corresponding to the Hess chart. The display unit displays the disease name acquired by the disease name acquisition unit.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an eye movement testing device. [Background technology]

[0002] Ocular movement disorders appear as symptoms of various diseases, including not only ophthalmological diseases but also brain surgical diseases and internal diseases (see Non-Patent Document 1). The Hess chart is very important for understanding the content and causes of ocular movement disorders. For example, the Hess chart illustrates and records the eye position and eye muscle function at each measurement point of the eye position in nine directions as a unified eye position diagram for each of the left and right eyes. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] https: / / www.tsukazaki-hp.jp / care / ophthalmology / eyelid-disorders Summary of the Invention

[0004] However, in order to identify possible diseases from the Hess chart, a diagnosis by an experienced doctor is essential, so it is desirable to provide an eye movement testing device that allows even inexperienced doctors to use the Hess chart.

[0005] An eye movement testing device according to one embodiment of the present invention includes an image acquisition unit, a data calculation unit, a disease name acquisition unit, and a display unit. The image acquisition unit acquires one or more Hess chart images. The data calculation unit calculates numerical data relating to the deviation between a basic coordinate and a light spot for each basic coordinate in one or more Hess chart images acquired by the image acquisition unit. The disease name acquisition unit inputs deviation data, obtained by arranging the numerical data for each basic coordinate in a predetermined rule, acquired by the data calculation unit into a learning model, and acquires a disease name corresponding to the Hess chart from the learning model. The display unit displays the disease name acquired by the disease name acquisition unit.

[0006] In the eye movement testing device according to one embodiment of the present invention, numerical data on the deviation between the basic coordinates and the light spot for each basic coordinate in one or more Hess chart images obtained by the image acquisition unit is calculated. After that, the deviation data obtained by the data calculation unit, in which the numerical data for each basic coordinate is arranged according to a predetermined rule, is input to the learning model, and the disease name corresponding to the Hess chart is obtained from the learning model. In this way, in the present invention, the Hess chart image is not directly input to the learning model, but the deviation data expressed in numerical values ​​obtained from the Hess chart image is input to the learning model. This makes it possible to estimate the disease name with high accuracy in the learning model. As a result, even inexperienced doctors can handle the Hess chart. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram for explaining an outline of the Hess Screen Test. [Diagram 2] 1 is a diagram illustrating an example of a schematic configuration of an eye movement testing device according to a first embodiment of the present invention. [Diagram 3] 3 is a diagram illustrating an example of input and output values ​​to the learning model of FIG. 2. [Figure 4] 3 is a diagram illustrating an example of a test procedure in the eye movement testing device of FIG. 2. [Diagram 5] FIG. 3 is a diagram illustrating a modified example of input and output values ​​to the learning model in FIG. [Figure 6]FIG. 3 is a diagram illustrating a modified example of the test procedure in the eye movement testing device of FIG. 2. [Figure 7] FIG. 13 is a diagram illustrating an example of a recording sheet for the Hess Screen Test. [Figure 8] FIG. 3 is a diagram illustrating a modified example of the eye movement testing device of FIG. 2. [Figure 9] FIG. 13 is a diagram illustrating an example of a schematic configuration of an eye movement testing system according to a second embodiment of the present invention. [Figure 10] FIG. 10 is a diagram illustrating a modified example of the schematic configuration of the eye movement testing system of FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. The following description is a specific example of the present invention, and the present invention is not limited to the following embodiment.

[0009] <First embodiment> [composition] 1 is a diagram for explaining an outline of the Hess screen test performed in this embodiment. In the Hess screen test performed in this embodiment, a Hess chart board 100, glasses 200, and a spotlight 300 are used.

[0010] The Hess chart board 100 is used to quantitatively measure eye positions in nine directions under fixation of each of the left and right eyes, and a Hess chart grid and nine basic coordinate points A to I are drawn on the surface of the Hess chart board 100. The surface of the Hess chart board 100 is black, and the Hess chart grid and nine basic coordinate points A to I are drawn in red.

[0011] The basic coordinate point A is disposed in front of the subject, and the basic coordinate points B to I are disposed in order at a predetermined interval in a clockwise direction around the basic coordinate point A. The basic coordinate point B is disposed above the subject (in the +90° direction with respect to the X-axis), the basic coordinate point C is disposed diagonally upward to the right of the subject (in the +45° direction with respect to the X-axis), the basic coordinate point D is disposed to the right of the subject (in the +0° direction with respect to the X-axis), the basic coordinate point E is disposed diagonally downward to the right of the subject (in the -45° direction with respect to the X-axis), the basic coordinate point F is disposed below the subject (in the -90° direction with respect to the X-axis), the basic coordinate point G is disposed diagonally downward to the left of the subject (in the -135° direction with respect to the X-axis), the basic coordinate point H is disposed to the left of the subject (in the +180° direction with respect to the X-axis), and the basic coordinate point I is disposed diagonally upward to the left of the subject (in the +135° direction with respect to the X-axis).

[0012] In the Hess screen test carried out in this embodiment, the subject wears glasses 200 in which red glass 210 and green glass 220 are removably fixed to a frame, and views a Hess chart board 100. While wearing the glasses 200 and viewing the Hess chart board 100, the subject irradiates green light Lg emitted from a spotlight 300 toward each of the basic coordinates A to I to form a light spot P on the Hess chart board 100. This makes it possible to separate the images (light source) seen by the left and right eyes based on the difference in color, making it difficult for motor fusion to occur.

[0013] 2 shows an example of a schematic configuration of an eye movement testing device 10 according to an embodiment of the present invention. The eye movement testing device 10 is a device that records a Hess screen test using a Hess chart board 100 and outputs a disease name according to the results of the recorded Hess screen test. The eye movement testing device 10 includes, for example, a control unit 11, an imaging unit 12, a communication unit 13, a memory 14, a learning model 15, an input unit 16, and a display unit 17.

[0014] The control unit 11 includes a central processing unit (CPU) and a graphics processing unit (GPU), and executes, for example, an operating system stored in the memory 14, a control program (not shown) that controls various devices included in the eye movement testing device 10, and the like. The control unit 11 further executes, for example, a program 14A stored in the memory 14. The program 14A causes the control unit 11 to execute a series of procedures for recording a Hess screen test using the Hess chart board 100 and outputting a disease name according to the recorded Hess screen test result.

[0015] The control unit 11 (hereinafter, simply referred to as the "control unit 11") to which the program 14A is loaded controls the imaging by the imaging unit 12 based on a control signal input from the shutter button 500 via the communication unit 13. This allows a medical professional to image the Hess chart board 100 by pressing the shutter button 500 at a predetermined timing. In the Hess screen test performed in this embodiment, a subject 400 wears glasses 200 and visually recognizes the Hess chart board 100. The medical professional presses the shutter button 500 at a timing when green light Lg emitted from a spotlight 300 is irradiated in the order of basic coordinates A, B, C, D, E, F, G, H, and I, aiming at the basic coordinates A to I, to image the Hess chart board 100 by the imaging unit 12. The control unit 11 stores the nine Hess chart images I1 to I9 obtained by the imaging unit 12 in the memory 14. The control unit 11 stores nine Hess chart images I1 to I9 obtained when the red glass 210 is worn on the left eye and the green glass 220 is worn on the right eye (hereinafter referred to as "nine Hess chart images I1 to I9 at left eye fixation") and nine Hess chart images I1 to I9 obtained when the red glass 210 is worn on the right eye and the green glass 220 is worn on the left eye (hereinafter referred to as "nine Hess chart images I1 to I9 at right eye fixation") in the memory 14. Note that hereinafter, "the nine Hess chart images I1 to I9 at left eye fixation and the nine Hess chart images I1 to I9 at right eye fixation" will be referred to as "18 Hess chart images I" as appropriate.

[0016] The control unit 11 reads out from the memory 14 nine Hess chart images I1 to I9 at the time of left eye fixation and nine Hess chart images I1 to I9 at the time of right eye fixation. The control unit 11 calculates numerical data 110 relating to the deviation between the basic coordinates A, B, C, D, E, F, G, H or I and the light spot for each of the basic coordinates A to I in the nine Hess chart images I1 to I9 at the time of left eye fixation read out from the memory 14. The control unit 11 further calculates numerical data 110 relating to the deviation between the basic coordinates A, B, C, D, E, F, G, H or I and the light spot for each of the basic coordinates A to I in the nine Hess chart images I1 to I9 at the time of right eye fixation read out from the memory 14. The light spot is an irradiation spot formed on the Hess chart board 100 by the green light Lg emitted from the spotlight 300. The control unit 11 calculates numerical data 110 in the nine Hess chart images I1 to I9 during left eye fixation for each of the basic coordinates A to I. The control unit 11 further calculates numerical data 110 in the nine Hess chart images I1 to I9 during right eye fixation for each of the basic coordinates A to I.

[0017] The control unit 11 calculates, from the Hess chart image I1, numerical data 110 relating to the deviation between the basic coordinate A and the light spot generated by irradiating light toward the basic coordinate A. The control unit 11 calculates, from the Hess chart image I2, numerical data 110 relating to the deviation between the basic coordinate B and the light spot generated by irradiating light toward the basic coordinate B. The control unit 11 calculates, from the Hess chart image I3, numerical data 110 relating to the deviation between the basic coordinate C and the light spot generated by irradiating light toward the basic coordinate C. The control unit 11 calculates, from the Hess chart image I4, numerical data 110 relating to the deviation between the basic coordinate D and the light spot generated by irradiating light toward the basic coordinate D. The control unit 11 calculates, from the Hess chart image I5, numerical data 110 relating to the deviation between the basic coordinate E and the light spot generated by irradiating light toward the basic coordinate E. The control unit 11 calculates, from the Hess chart image I6, numerical data 110 relating to the deviation between the basic coordinate F and the light spot generated by irradiating light toward the basic coordinate F. The control unit 11 calculates, from the Hess chart image I7, numerical data 110 relating to the deviation between the basic coordinate G and the light spot generated by irradiating light toward the basic coordinate G. The control unit 11 calculates, from the Hess chart image I8, numerical data 110 relating to the deviation between the basic coordinate H and the light spot generated by irradiating light toward the basic coordinate H. The control unit 11 calculates, from the Hess chart image I9, numerical data 110 relating to the deviation between the basic coordinate I and the light spot generated by irradiating light toward the basic coordinate I.

[0018] The numerical data 110 is, for example, a deviation amount 111 in basic coordinates 112 (basic coordinates A to I) as shown in FIG. 3. The deviation amount 111 is expressed as a difference between the basic coordinates 112 (basic coordinates A to I) and the light spot coordinates. The basic coordinates 112 are the coordinates of the basic coordinates A, B, C, D, E, F, G, H, or I on the Hess chart board 100. The light spot coordinates are the coordinates of the irradiation spot formed by irradiation with the spotlight 300 on the Hess chart board 100. The deviation amount 111 is expressed by a deviation amount ΔX between the basic coordinates 112 and the light spot coordinates in the left-right direction (X-axis direction) and a deviation amount ΔY between the basic coordinates 112 and the light spot coordinates in the up-down direction (Y-axis direction).

[0019] The control unit 11 generates a sequence (displacement data) in which the displacement amount 111 for each basic coordinate 112 is arranged according to a predetermined rule. For example, the control unit 11 generates the displacement amount 111 (ΔXa, ΔYa) in the basic coordinate A (Xa1, Ya1), the displacement amount 111 (ΔXb, ΔYb) in the basic coordinate B (Xb1, Yb1), the displacement amount 111 (ΔXc, ΔYc) in the basic coordinate C (Xc1, Yc1), the displacement amount 111 (ΔXd, ΔYd) in the basic coordinate D (Xd1, Yd1), the displacement amount 111 (ΔXd, ΔYd) in the basic coordinate E (Xe1, Ye1), The control unit 11 generates a sequence in which the amount of deviation 111 (ΔXe, ΔYe) at the basic coordinates F (Xf1, Yf1), the amount of deviation 111 (ΔXf, ΔYf) at the basic coordinates G (Xg1, Yg1), the amount of deviation 111 (ΔXg, ΔYg) at the basic coordinates H (Xh1, Yh1), and the amount of deviation 111 (ΔXi, ΔYi) at the basic coordinates I (Xi1, Yi1) are arranged in this order. The control unit 11 further generates a sequence for left eye fixation and a sequence for right eye fixation, and generates a sequence by combining these sequences as the input value Din of the learning model 15.

[0020] The deviation amount 111 (ΔXa, ΔYa) is, for example, a difference between a basic coordinate A (Xa1, Ya1) and a coordinate (Xa2, Ya2) of a light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate A. The deviation amount 111 (ΔXb, ΔYb) is, for example, a difference between a basic coordinate B (Xb1, Yb1) and a coordinate (Xb2, Yb2) of a light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate B. The deviation amount 111 (ΔXc, ΔYc) is, for example, a difference between a basic coordinate C (Xc1, Yc1) and a coordinate (Xc2, Yc2) of a light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate C. The deviation amount 111 (ΔXd, ΔYd) is, for example, a difference between the basic coordinate D (Xd1, Yd1) and the coordinate (Xd2, Yd2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate D. The deviation amount 111 (ΔXe, ΔYe) is, for example, a difference between the basic coordinate E (Xe1, Ye1) and the coordinate (Xe2, Ye2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate E. The deviation amount 111 (ΔXf, ΔYf) is, for example, a difference between the basic coordinate F (Xf1, Yf1) and the coordinate (Xf2, Yf2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate F. The deviation amount 111 (ΔXg, ΔYg) is, for example, a difference between the basic coordinate G (Xg1, Yg1) and the coordinate (Xg2, Yg2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate G. The deviation amount 111 (ΔXh, ΔYh) is, for example, a difference between the basic coordinate H (Xh1, Yh1) and the coordinate (Xh2, Yh2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate H. The deviation amount 111 (ΔXi, ΔYi) is, for example, a difference between the basic coordinate I (Xi1, Yi1) and the coordinate (Xi2, Yi2) of the light spot formed on the Hess chart board 100 by the green light Lg irradiated toward the basic coordinate I.

[0021] The control unit 11 inputs the generated input value Din to the learning model 15, and acquires the disease name corresponding to the Hess chart as a response (output value Dout) to the input value Din from the learning model 15. The control unit 11 generates a video signal including the acquired disease name, and outputs it to the display unit 17.

[0022] The imaging unit 12 is, for example, a camera including a CCD (Charge Coupled Device) image sensor or a CMOS (Complementary metal-oxide-semiconductor) image sensor. The imaging unit 12 captures an image of the Hess chart board 100 under the control of the control unit 11, and outputs the Hess chart image thus obtained to the control unit 11. The imaging unit 12, for example, captures nine Hess chart images I1 to I9, and outputs the nine Hess chart images I1 to I9 thus obtained to the control unit 11.

[0023] The communication unit 13 receives a control signal output from the shutter button 500 , and outputs the received control signal to the control unit 11 .

[0024] The memory 14 stores a program (e.g., an operating system) executed by the control unit 11. The memory 14 is configured with, for example, a RAM (Random Access Memory), a ROM (Read Only Memory), an auxiliary storage device (such as a hard disk), etc. The memory 14 further stores a program 14A.

[0025] When an input value Din is input, the learning model 15 outputs a disease name corresponding to the Hess chart as a response (output value Dout) to the input value Din. The learning model 15 is a model in which machine learning has been performed using the learning input value Din as an explanatory variable and the doctor's diagnosis (disease name) for the Hess chart, which is the original data of the learning input value Din, as a target variable. The learning model 15 is composed of, for example, artificial intelligence (AI).

[0026] The input unit 16 receives instructions from the outside (for example, a user) and outputs the received instructions to the control unit 11. The input unit 16 may be, for example, a mechanical input interface including a button, a dial, etc., or may be a voice input interface including a microphone, etc. The input unit 16 may be, for example, a touch panel provided on the display surface of the display unit 17.

[0027] The display unit 17 is composed of a panel-type display device such as a liquid crystal panel or an organic EL (Electro Luminescence) panel. The display unit 17 displays an image based on a video signal input from the control unit 11. The display unit 17 displays the disease name obtained by the learning model 15.

[0028] [Operation] Next, an example of the examination procedure in the eye movement examination device 10 according to the present embodiment will be described with reference to FIG. 4. FIG. 4 shows an example of the examination procedure in the eye movement examination device 10. First, the subject 400 sits facing the Hess chart board 100 and fixes his / her head on the chin rest. Next, the subject 400 wears the green glass 220 on his / her right eye and the red glass 210 on his / her left eye, and holds the spotlight 300 in his / her hand. Next, the subject 400, while viewing the Hess chart board 100, irradiates the green light Lg emitted from the spotlight 300 to the basic coordinates A to I in the order of the basic coordinates A, B, C, D, E, F, G, H, I. At this time, the medical staff presses the shutter button 500 at the timing when the green light Lg is irradiated to the basic coordinates A to I in the order of the basic coordinates A, B, C, D, E, F, G, H, I. As a result, the imaging unit 12 acquires nine Hess chart images I1-I9 at left eye fixation through imaging, and outputs the acquired nine Hess chart images I1-I9 at left eye fixation to the control unit 11. The control unit 11 acquires the nine Hess chart images I1-I9 at left eye fixation from the imaging unit 12 (step S101). The control unit 11 stores the acquired nine Hess chart images I1-I9 at left eye fixation in the memory 14.

[0029] Next, the subject 400 wears a red glass 210 on his right eye and a green glass 220 on his left eye, and while viewing the Hess chart board 100, the green light Lg emitted from the spotlight 300 is aimed at the basic coordinates A to I, and is irradiated in the order of the basic coordinates A, B, C, D, E, F, G, H, and I. At this time, the medical staff presses the shutter button 500 at the timing when the green light Lg is irradiated in the order of the basic coordinates A to I, and is aimed at the basic coordinates A to I. As a result, the imaging unit 12 acquires nine Hess chart images I1 to I9 at the time of right eye fixation by imaging, and outputs the acquired nine Hess chart images I1 to I9 at the time of right eye fixation to the control unit 11. The control unit 11 acquires the nine Hess chart images I1 to I9 at the time of right eye fixation from the imaging unit 12 (step S101). The control unit 11 stores in the memory 14 the nine Hess chart images I1 to I9 acquired during right eye fixation.

[0030] Next, the control unit 11 reads out nine Hess chart images I1-I9 at left eye fixation from the memory 14, and calculates numerical data 110 (amount of deviation 111) relating to the deviation between the basic coordinates A, B, C, D, E, F, G, H or I and the light spot for each of the basic coordinates A-I in the read out nine Hess chart images I1-I9 at left eye fixation. The control unit 11 calculates the numerical data 110 (amount of deviation 111) in the nine Hess chart images I1-I9 at left eye fixation for each of the basic coordinates A-I (step S102).

[0031] The control unit 11 further reads out nine Hess chart images I1-I9 at right eye fixation from the memory 14, and calculates numerical data 110 (amount of deviation 111) relating to the deviation between the basic coordinates A, B, C, D, E, F, G, H or I and the light spot for each of the basic coordinates A-I in the read nine Hess chart images I1-I9 at right eye fixation. The control unit 11 calculates the numerical data 110 (amount of deviation 111) in the nine Hess chart images I1-I9 at right eye fixation for each of the basic coordinates A-I (step S102).

[0032] Next, the control unit 11 generates a sequence (displacement data) in which the displacement amount 111 for each basic coordinate 112 is arranged according to a predetermined rule. The control unit 11 generates a sequence of the displacement amount 111 at the time of left eye fixation and a sequence of the displacement amount 111 at the time of right eye fixation, and generates a sequence by combining these sequences as an input value Din of the learning model 15. The control unit 11 inputs the generated input value Din to the learning model 15 (step S103). The control unit 11 acquires the disease name corresponding to the Hess chart from the learning model 15 as a response (output value Dout) to the input value Din (step S104). The control unit 11 generates a video signal including the acquired disease name, and outputs it to the display unit 17. The display unit 17 displays the disease name acquired by the learning model 15. In this manner, the examination in the eye movement examination device 10 is performed.

[0033] [effect] Next, the effects of the eye movement testing device 10 according to this embodiment will be described.

[0034] Eye movement disorders are a symptom of various diseases, including not only ophthalmological diseases, but also brain surgical diseases and internal medicine diseases. The Hess chart is very important for understanding the content and cause of eye movement disorders. However, a diagnosis by an experienced doctor is essential to identify what kind of disease may be suspected from the Hess chart.

[0035] On the other hand, in this embodiment, numerical data 110 relating to the deviation between the basic coordinates 112 and the light spot coordinates for each of the basic coordinates 112 in the 18 Hess chart images I obtained by the imaging unit 12 is calculated. In this embodiment, a deviation amount 111, which is a difference between the basic coordinates 112 and the light spot coordinates, is calculated as the numerical data 110. After that, a sequence in which the numerical data 110 (deviation amount 111) for each of the basic coordinates 112 is arranged according to a predetermined rule is input to the learning model 15 as an input value Din, and the disease name corresponding to the Hess chart is acquired from the learning model 15 as a response (output value Dout) to the input value Din. Thus, in this embodiment, the 18 Hess chart images I are not directly input to the learning model 15, but a numerical sequence expressed by numerical values ​​obtained from the 18 Hess chart images I is input to the learning model 15. This makes it possible to estimate the disease name with high accuracy in the learning model 15. As a result, even inexperienced doctors can handle the Hess chart.

[0036] In this embodiment, the imaging unit 12 is a camera, and 18 Hess chart images I are obtained by imaging the Hess chart board 100 with the camera every time a light spot is irradiated onto the Hess chart board 100 toward the basic coordinates 112. This saves medical personnel the trouble of reading the recorded results from the Hess chart recording paper, and allows medical personnel to quickly and accurately learn the name of the disease.

[0037] <Modification of the first embodiment> Next, a modification of the eye movement testing device 10 according to the above embodiment will be described.

[0038] [Variation A] FIG. 5 shows a modified example of input / output values ​​to the learning model 15. In the above embodiment, the control unit 11 may calculate, as the numerical data 110, light spot coordinates 113 for each of the basic coordinates A to I in the nine Hess chart images I1 to I9 at the time of left eye fixation, and light spot coordinates 113 for each of the basic coordinates A to I in the nine Hess chart images I1 to I9 at the time of right eye fixation. In this case, the control unit 11 may generate a sequence including the basic coordinates 112, the light spot coordinates 113 at the time of left eye fixation, and the light spot coordinates 113 at the time of right eye fixation, as the input value Din to be input to the learning model 15. Even in this case, it is possible to estimate the disease name with high accuracy in the learning model 15. As a result, even an inexperienced doctor can handle the Hess chart.

[0039] 6 shows an example of a test procedure in the eye movement testing device 10. First, the control unit 11 acquires nine Hess chart images I1 to I9 at left eye fixation and nine Hess chart images I1 to I9 at right eye fixation in the same manner as in the above embodiment (step S101). The control unit 11 stores the acquired nine Hess chart images I1 to I9 at left eye fixation and nine Hess chart images I1 to I9 at right eye fixation in the memory 14.

[0040] Next, the control unit 11 reads out from the memory 14 the nine Hess chart images I1-I9 at left eye fixation, and calculates the light spot coordinates 113 for each of the basic coordinates A-I in the nine Hess chart images I1-I9 at left eye fixation that have been read out (step S102). The control unit 11 further reads out from the memory 14 the nine Hess chart images I1-I9 at right eye fixation, and calculates the light spot coordinates 113 for each of the basic coordinates A-I in the nine Hess chart images I1-I9 at right eye fixation that have been read out (step S102).

[0041] Next, the control unit 11 generates a sequence (shift data) in which the light spot coordinates 113 for each basic coordinate 112 are arranged according to a predetermined rule. For example, the control unit 11 generates the light spot coordinates 113 (Xa2, Ya2) at the basic coordinate A (Xa1, Ya1), the light spot coordinates 113 (Xb2, Yb2) at the basic coordinate B (Xb1, Yb1), the light spot coordinates 113 (Xc2, Yc2) at the basic coordinate C (Xc1, Yc1), the light spot coordinates 113 (Xd2, Yd2) at the basic coordinate D (Xd1, Yd1), the light spot coordinates 113 (Xd2, Yd2) at the basic coordinate E (Xe1, Ye1), The control unit 11 generates a sequence in which the light spot coordinate 113 (Xe2, Ye2) at the basic coordinate F (Xf1, Yf1), the light spot coordinate 113 (Xf2, Yf2) at the basic coordinate G (Xg1, Yg1), the light spot coordinate 113 (Xg2, Yg2) at the basic coordinate H (Xh1, Yh1), and the light spot coordinate 113 (Xi2, Yi2) at the basic coordinate I (Xi1, Yi1) are arranged in this order. The control unit 11 further generates, for example, a sequence of the basic coordinates A to I, a sequence of the light spot coordinate 113 at the time of left eye fixation, and a sequence of the light spot coordinate 113 at the time of right eye fixation, and generates a sequence by combining these sequences as the input value Din of the learning model 15.

[0042] Next, the control unit 11 inputs the input value Din to the learning model 15 (step S103). The control unit 11 acquires the disease name corresponding to the Hess chart from the learning model 15 as a response (output value Dout) to the input value Din (step S104). The control unit 11 generates a video signal including the acquired disease name and outputs it to the display unit 17. The display unit 17 displays the disease name acquired by the learning model 15. In this manner, the examination in the eye movement examination device 10 is performed.

[0043] In this modification, the light spot coordinates 113 are calculated for each of the basic coordinates 112 in the 18 Hess chart images I obtained by the imaging unit 12. In this modification, the basic coordinates 112 and the light spot coordinates 113 are calculated as the numerical data 110. After that, a sequence in which the basic coordinates 112 and the numerical data 110 (light spot coordinates 113) for each basic coordinate 112 are arranged according to a predetermined rule is input to the learning model 15 as the input value Din, and the disease name corresponding to the Hess chart is acquired from the learning model 15 as a response (output value Dout) to the input value Din. In this manner, in this modification, the 18 Hess chart images I are not directly input to the learning model 15, but a numerical sequence obtained from the 18 Hess chart images I and expressed in numerical values ​​is input to the learning model 15. This makes it possible to estimate the disease name with high accuracy in the learning model 15. As a result, even inexperienced doctors can handle the Hess chart.

[0044] [Variation B] FIG. 7 shows an example of a recording sheet 600 for the Hess screen test. The recording sheet 600 is a Hess chart on which light spots Pa to Pi are plotted for each of the basic coordinates A to I. In the above embodiment and its modified example, the imaging unit 12 may obtain a Hess chart image by imaging the recording sheet 600 at the time of left eye fixation and the recording sheet 600 at the time of right eye fixation, instead of the Hess chart board 100. In this case, the control unit 11 can calculate the amount of deviation 111 or the coordinates of the light spots Pa to Pi from the two recording sheets 600 (the recording sheet 600 at the time of left eye fixation and the recording sheet 600 at the time of right eye fixation). Therefore, even in this modified example, the learning model 15 can estimate the disease name with high accuracy. As a result, even an inexperienced doctor can handle the Hess chart.

[0045] [Variation C] Fig. 8 shows a modified example of the eye movement testing device 10 of Fig. 2. In this modified example, a receiving unit 18 is provided instead of the imaging unit 12, and a light receiving panel 700 is provided instead of the Hess chart board 100.

[0046] The receiving unit 18 receives the received light image Ic output from the light receiving panel 700, and outputs the received received light image Ic to the control unit 11. The light receiving panel 700 captures an image formed by light incident on the surface (light receiving surface) of the light receiving panel 700 according to the control by the control unit 11, and thereby acquires the received light image Ic. The light receiving panel 700 outputs the acquired received light image Ic to the receiving unit 18 as a Hess chart image I. The light receiving panel 700 has, for example, a light receiving element array in which a plurality of light receiving elements are arranged in a matrix directly under the light receiving surface of the light receiving panel 700. The light receiving panel 700 generates the received light image Ic based on, for example, a signal output from the light receiving element array. The light receiving panel 700 acquires, for example, nine received light images Ic1 to Ic9 at the time of left eye fixation by receiving light, and outputs the acquired nine received light images Ic1 to Ic9 at the time of left eye fixation to the control unit 11. The light receiving panel 700, for example, obtains nine light receiving images Ic1 to Ic9 during right eye fixation by receiving light, and outputs the obtained nine light receiving images Ic1 to Ic9 during right eye fixation to the control unit 11.

[0047] In this modification, the control unit 11 calculates the amount of deviation 111 or the coordinates of the light spots Pa to Pi by processing the received light image Ic obtained from the light receiving panel 700 in the same manner as the Hess chart image I according to the above embodiment. The control unit 11 generates an input value Din using the amount of deviation 111 or the coordinates of the light spots Pa to Pi, inputs the generated input value Din to the learning model 15, and obtains the disease name corresponding to the Hess chart from the learning model 15 as a response (output value Dout) to the input value Din. In this manner, in this modification, a numerical sequence obtained from the received light image Ic obtained from the light receiving panel 700 and expressed in numerical values ​​is input to the learning model 15. This makes it possible to estimate the disease name with high accuracy in the learning model 15. As a result, even inexperienced doctors can handle the Hess chart.

[0048] [Variation D] In the above modification C, the light receiving panel 700 may be configured to detect light incident on the surface (light receiving surface) of the light receiving panel 700 under the control of the control unit 11, and calculate the light spot coordinates 113 from the detection result. In this case, the light receiving panel 700 outputs the light spot coordinates 113 to the control unit 11 instead of the received light image Ic. The control unit 11 may generate the input value Din using the light spot coordinates 113 obtained from the light receiving panel 700, or may generate the input value Din using the basic coordinates 112 and the light spot coordinates 113 obtained from the light receiving panel 700. In this manner, in this modification, a numerical sequence obtained from the light spot coordinates 113 obtained from the light receiving panel 700 and expressed in numerical values ​​is input to the learning model 15. This makes it possible to estimate the disease name with high accuracy in the learning model 15. As a result, even an inexperienced doctor can handle the Hess chart.

[0049] <Second embodiment> [composition] 9 shows a schematic configuration example of an eye movement testing system 20 according to an embodiment of the present invention. The eye movement testing system 20 includes an eye movement testing device 30, a server device 40, a Hess chart board 100, a spotlight 300, and a shutter button 500. The eye movement testing device 30 and the server device 40 are connected via an external network 50.

[0050] The external network 50 is, for example, a network that performs communication using a communication protocol (TCP / IP) that is commonly used on the Internet. The external network 50 may be, for example, a secure network that performs communication using a communication protocol unique to the network. The external network 50 is, for example, the Internet, an intranet, or a local area network. The connection between the external network 50 and the eye movement testing device 30 or the server device 40 may be, for example, a wired LAN (Local Area Network) such as Ethernet, a wireless LAN such as Wi-Fi, or a mobile phone line.

[0051] The eye movement testing device 30 is a device obtained by newly providing a communication unit 19 in the eye movement testing device 10 according to the above embodiment and omitting the learning model 15. The communication unit 19 communicates with a server device 40 via an external network 50.

[0052] The server device 40 includes, for example, a control unit 41, a learning model 15, and a communication unit 42. The control unit 41 includes a CPU, a GPU, and the like, and executes, for example, an operating system stored in a memory. When an input value Din is input from the eye movement testing device 30 via an external network 50, the control unit 41 further inputs the input value Din to the learning model 15, and acquires a disease name corresponding to the Hess chart from the learning model 15 as a response (output value Dout) to the input value Din. The control unit 41 outputs the acquired disease name to the eye movement testing device 30 via the external network 50.

[0053] The communication unit 42 communicates with the eye movement testing device 30 via the external network 50. When the input value Din is input from the control unit 41, the learning model 15 outputs the disease name corresponding to the Hess chart to the control unit 41 as a response (output value Dout) to the input value Din.

[0054] In this embodiment, the learning model 15 is provided in a server device 40 on an external network 50. In this case, for example, a plurality of eye movement testing devices 30 can share one learning model 15. This allows a plurality of doctors in a hospital or a plurality of hospitals to use the eye movement testing device 30 to perform diagnosis based on the Hess chart.

[0055] [Variation E] 10 shows a modified example of the schematic configuration of the eye movement testing system 20 according to the second embodiment. In this modified example, a receiving unit 18 is provided instead of the imaging unit 12, and a light receiving panel 700 is provided instead of the Hess chart board 100. That is, in this modified example, the eye movement testing device 30 can perform the same operations as the eye movement testing device 10 according to the modified example C.

[0056] In this modification, the learning model 15 is also provided in a server device 40 on an external network 50. In this case, for example, a plurality of eye movement testing devices 30 can share one learning model 15. This allows a plurality of doctors in a hospital or a plurality of hospitals to use the eye movement testing device 30 to perform diagnosis based on the Hess chart.

[0057] Although the present invention has been described above by way of a number of embodiments and their modified examples, the present invention is not limited to the above-described embodiments, and various modifications are possible. Note that the effects described in this specification are merely examples. The effects of the present invention are not limited to the effects described in this specification. The present invention may have effects other than those described in this specification.

Claims

1. an image acquisition unit for acquiring a plurality of Hess chart images; a data calculation unit that calculates a deviation amount between a basic coordinate and a light spot for each of the Hess chart images from the plurality of Hess chart images obtained by the image acquisition unit; a disease name acquiring unit that inputs numerical data in which the plurality of deviation amounts obtained by the data calculating unit are arranged into a learning model and acquires a disease name corresponding to the Hess chart from the learning model; a display unit that displays the disease name obtained by the disease name acquisition unit; Equipped with The image acquisition unit is a camera, and acquires the plurality of Hess chart images by capturing an image of the Hess chart board with the camera each time the light spot is irradiated onto the Hess chart board toward the basic coordinates. Eye movement testing device.

2. an image acquisition unit for acquiring a plurality of Hess chart images; a data calculation unit that calculates basic coordinates and coordinates of a light spot for each of the Hess chart images from the plurality of Hess chart images obtained by the image acquisition unit; a disease name acquiring unit that inputs numerical data obtained by arranging the plurality of sets of the basic coordinates and the coordinates of the light spots obtained by the data calculating unit into a learning model and acquires a disease name corresponding to the Hess chart from the learning model; a display unit that displays the disease name obtained by the disease name acquisition unit; Equipped with The image acquisition unit is a camera, and acquires the plurality of Hess chart images by capturing an image of the Hess chart board with the camera each time the light spot is irradiated onto the Hess chart board toward the basic coordinates. Eye movement testing device.

3. An image acquisition unit for acquiring a plurality of Hess chart images; a data calculation unit that calculates a deviation amount between a basic coordinate and a light spot for each of the Hess chart images from the plurality of Hess chart images obtained by the image acquisition unit; a disease name acquiring unit that inputs numerical data in which the plurality of deviation amounts obtained by the data calculating unit are arranged into a learning model and acquires a disease name corresponding to the Hess chart from the learning model; a display unit that displays the disease name obtained by the disease name acquisition unit; Equipped with The image acquisition unit is a light receiving panel on which a Hess chart is drawn, and the light receiving panel receives light every time the light spot is irradiated onto the light receiving panel toward the basic coordinates, thereby acquiring the Hess chart images. Eye movement testing device.

4. An image acquisition unit for acquiring a plurality of Hess chart images; a data calculation unit that calculates basic coordinates and coordinates of a light spot for each of the Hess chart images from the plurality of Hess chart images obtained by the image acquisition unit; a disease name acquiring unit that inputs numerical data obtained by arranging the plurality of sets of the basic coordinates and the coordinates of the light spots obtained by the data calculating unit into a learning model and acquires a disease name corresponding to the Hess chart from the learning model; a display unit that displays the disease name obtained by the disease name acquisition unit; Equipped with The image acquisition unit is a light receiving panel on which a Hess chart is drawn, and the light receiving panel receives light every time the light spot is irradiated onto the light receiving panel toward the basic coordinates, thereby acquiring the multiple Hess chart images. Eye movement testing device.

Citation Information

Patent Citations

  • Ophthalmologic examination system

    JP2009045175A

  • Ophthalmologic examination apparatus and hess examination apparatus

    JP2011004930A

  • Cerebral disorder diagnosis support apparatus and method for supporting cerebral disorder diagnosis

    JP2013169375A

  • Eye examination device

    JP2017042301A

  • Medical image processing apparatus, medical image processing method, and program

    JP2019216848A