JP7252295B2 - Ophthalmic device and ophthalmic measurement method - Google Patents

Description

The present invention relates to an ophthalmologic apparatus and an ophthalmologic measurement method.

An ophthalmologic apparatus that performs auto-alignment for automatically adjusting the position of a measurement optical system with respect to an eye to be examined has been proposed (see, for example, Patent Document 1). This ophthalmologic apparatus automatically measures (inspects) eye characteristics (optical characteristics) such as eye refractive power and corneal shape of an eye to be examined by an auto-aligned measurement optical system.

JP 2016-49283 A

However, in the above-described conventional ophthalmologic apparatus, auto-alignment cannot be performed due to a special condition of the eye to be examined due to illness or the like, or even if auto-alignment succeeds, insufficient exposure may occur. There are situations where measurement is not possible with Here, since the conventional ophthalmologic apparatus automatically performs measurement according to a predetermined measurement procedure, there is a high possibility that automatic measurement cannot be performed in the above-described situation no matter how many times the measurement is performed. For this reason, conventional ophthalmologic apparatuses are limited in situations in which the ocular characteristics of an eye to be examined can be automatically measured.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is an object of the present invention to expand the conditions in which eye identification of an eye to be examined can be automatically measured, and to improve measurement performance.

In order to achieve the above object, an ophthalmologic apparatus according to the present application includes an eye characteristic measuring unit that measures the eye characteristics of an eye to be examined, and a measurement condition acquisition unit that acquires measurement conditions when the eye characteristics measuring unit performs measurement. a control unit that controls the eye characteristics measuring unit according to a predetermined measurement procedure; the measurement conditions in a situation where the eye characteristics cannot be measured in the measurement procedure; and a modality of an ophthalmologic apparatus main body, and learning measurement for avoiding a situation in which measurement cannot be performed based on the measurement conditions, the situation data, and the modality in a situation in which the eye characteristics are appropriately measured in the measurement procedure. a learning unit that generates a condition, and when a situation arises in which measurement of the eye characteristics of the subject eye cannot be performed in the measurement procedure, the control unit learns the situation from the learned measurement condition created by the learning unit. is extracted, and based on the learned measurement condition, the eye characteristics measuring unit is controlled to measure the eye characteristics of the subject's eye.

Further, an ophthalmologic measurement method according to the present application is an ophthalmologic measurement method performed by the ophthalmologic apparatus, comprising a step of measuring ocular characteristics of an eye to be examined according to a predetermined measurement procedure, and a step of measuring the ocular characteristics. a step of acquiring situation data relating to the condition of the eye to be examined and the modality of the ophthalmologic apparatus main body; and a step of acquiring the measurement conditions, the situation data, and Based on the modality, the measurement condition, the situation data, and the modality in a situation in which the eye characteristics are appropriately measured in the measurement procedure, a learned measurement condition is generated to avoid a situation in which measurement cannot be performed. and extracting the learning measurement condition corresponding to the situation from the learning measurement condition generated in the step of generating the learning measurement condition when the eye characteristic cannot be measured. and measuring the eye characteristics of the subject's eye based on the learned measurement conditions.

ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to expand the condition which can measure eye identification of a to-be-tested eye automatically, and to improve a measurement performance.

1 is a schematic diagram showing the overall configuration of an ophthalmologic apparatus according to Example 1. FIG. 2 is a functional block diagram showing an example of the functional configuration of the ophthalmologic apparatus of FIG. 1; FIG. 2 is a block diagram showing an example of the system configuration of the ophthalmologic apparatus main body of FIG. 1; FIG. FIG. 2 is an explanatory diagram showing an example of the optical configuration of the main body of the ophthalmologic apparatus of FIG. 1; 2 is a flowchart showing the flow of eye characteristic measurement processing (eye characteristic measurement method) executed by the ophthalmologic apparatus main body of FIG. 1; 2 is a flow chart showing the flow of learning processing (learning method) executed by the cloud server of FIG. 1; 6B is a flow chart showing the flow of data accumulation processing in FIG. 6A; 6B is a flow chart showing the flow of the learning process in FIG. 6A; 6B is a flow chart showing the flow of the artificial intelligence engine delivery process of FIG. 6A; A configuration example of an artificial intelligence engine learned by applying a neural network is shown.

An embodiment of an ophthalmologic apparatus according to the present disclosure will be described below with reference to the drawings.

The ophthalmologic apparatus 100 according to the first embodiment is constructed using a cloud computing service, and as shown in FIG. They are interconnected via a network 1 .

The communication network 1 is not particularly limited as long as it controls connection between the cloud server 3 and the main body 10 of the ophthalmologic apparatus. As the communication network 1, for example, a public telephone line such as the Internet, a dedicated telephone line, an optical communication line, a PAN (Personal Area Network), a LAN (Local Area Network), a MAN (Metropolitan Area Network), a WAN (Wide Area Network), Wimax (Worldwide Interoperability for Microwave Access), a wireless network such as a mobile phone network, or the like can be used.

Next, functions of the cloud server 3 and the ophthalmologic apparatus main body 10 will be described with reference to the functional block diagram of FIG. As shown in FIG. 2, the cloud server 3 is installed in a data center 4, for example.

The cloud server 3 has a storage section 5 and a learning section 6 . The storage unit 5 is updated by a data accumulation unit 7 for accumulating ophthalmologic measurement information (error log data) such as measurement condition data and situation data transmitted from the ophthalmologic apparatus main body 10 via the communication network 1 and a learning unit 6 . and an artificial intelligence engine storage unit 8 for storing the artificial intelligence engine.

The learning unit 6 is composed of so-called artificial intelligence (AI). Based on the error log data accumulated in the data accumulation unit 7, the learning unit 6 generates error avoidance data (learning measurement condition data) for avoiding errors. The error avoidance information is a method (avoidance measure) for measuring the eye characteristics of the subject's eye E in a situation where the eye characteristics cannot be measured by a predetermined measurement sequence (measurement procedure). The learning unit 6 updates the artificial intelligence engine stored in the artificial intelligence engine storage unit 8 based on the created error avoidance information. The updated artificial intelligence engine is sent to the ophthalmologic device main body 10 via the communication network 1 automatically or periodically by an update request from the ophthalmologic device main body 10 .

In the initial state where the ophthalmologic apparatus 100 is not yet in use, a situation in which measurement is not possible has not occurred, and error log data has not been created or transmitted, so learning is not performed by the learning unit 6. , error avoidance information is not accumulated. When the ophthalmologic apparatus 100 is used and a situation occurs in which eye characteristics cannot be measured and error log data is transmitted, learning is started, and error avoidance information corresponding to various situations is accumulated. However, the learning unit 6 is caused to learn in advance by assuming several situations in which measurement is not possible, and an artificial intelligence engine that accumulates error avoidance information is generated, stored in the artificial intelligence engine storage unit 8, and delivered to the ophthalmologic apparatus main body 10. You can also keep

The data accumulation unit 7 and the artificial intelligence engine storage unit 8 can be provided, for example, for each model of the connected ophthalmologic apparatus main body 10 . In this case, the learning unit 6 updates the artificial intelligence engine for each model based on the error log data in the data storage unit 7 for each model. As a result, it is possible to share learning results for situations in which various eye characteristics cannot be measured in each ophthalmologic apparatus main body 10 of the same model. In addition, since learning is performed based on error log data in various situations, the learning performance of the learning unit 6 is improved, and the performance of the artificial intelligence engine generated can also be improved. By using such an artificial intelligence engine in each ophthalmologic apparatus main body 10, it is possible to expand the situation in which the eye identification of the subject's eye E can be automatically measured, and improve the measurement performance.

Also, the data storage unit 7 and the artificial intelligence engine storage unit 8 can be provided for each ophthalmologic apparatus main body 10 . In this case, the learning unit 6 can be configured to analyze and learn the error log data for each ophthalmologic apparatus main body 10 and update the respective artificial intelligence engines.

Note that the device having the storage unit 5 and the learning unit 6 is not limited to the cloud server 3, and may be a general-purpose large-scale computer connected to the ophthalmologic apparatus main body 10 via the communication network 1, cable, or the like. A mainframe) or in-house server can be used, or a personal computer can be used for further simplification. Also, the storage unit 5 and the learning unit 6 can be integrally provided in the ophthalmologic apparatus main body 10 .

The ophthalmologic apparatus main body 10 includes a measuring section 20, a user interface section 30, a control section 40, a data processing section 50, and a storage section 60, as shown in FIG.

The measurement unit 20 has a function of measuring the eye characteristics (optical characteristics) of the eye E to be examined. The measurement unit 20 includes an eye characteristic measurement unit 16 that measures eye characteristics and captures an image of the eye E to be examined using various optical systems, and a measurement condition acquisition unit 17 that acquires measurement condition data.

The measurement condition data are parameters for the operation of the eye characteristic measuring unit 16, such as alignment and eye characteristic measurement. For example, there are control parameters (operation history data) such as the moving speed of the gantry during alignment, and measurement parameters such as the focal length, gain, measurement position, and exposure amount of the optical system. The measurement condition data also includes a measurement mode. A measurement mode is a set of data in which various measurement parameters such as focal length, gain, measurement position, and exposure amount are determined in advance according to the measurement site of eye characteristics and the purpose of measurement. be. The measurement modes include, for example, an optic disc measurement mode and a macular measurement mode. If a measurement mode is specified when measuring eye characteristics, various measurement parameters are automatically set and measurement is performed. It is also possible to measure eye characteristics by individually inputting measurement parameters without specifying a measurement mode. In the ophthalmologic apparatus main body 10, alignment and measurement are performed by operating the optical system and driving members under the control of the control unit 40 according to various parameters input and operations performed on the user interface unit 30. Acquisition of condition data is possible.

The user interface unit 30 has a function of inputting and outputting various data, an operation unit 31 for operating the measurement unit 20, a display unit 32 for displaying images taken by the measurement unit 20, and an input unit 33 for inputting various data such as parameters to be used.

The storage unit 60 includes internal memory such as ROM and RAM, and external memory such as HDD and flash memory. The storage unit 60 stores a program for operating the ophthalmologic apparatus main body 10 and various data used for measurement. In addition, the storage unit 60 includes an error log data storage unit 61 that stores error log data, and an artificial intelligence engine storage unit 62 that stores an update artificial intelligence engine distributed from the cloud server 3. I have.

The control unit 40 comprehensively controls the operation of the ophthalmologic apparatus main body 10 based on the programs stored in the storage unit 60 . The control unit 40 includes an input/output control unit 41, a measurement sequence control unit 42, and an error processing unit 43, as shown in FIG.

The input/output control unit 41 controls data input/output within the ophthalmologic apparatus main body 10 and data input/output between the ophthalmologic apparatus main body 10 and the cloud server 3 . The input/output control unit 41 receives operation data from the operation unit 31 and input data from the input unit 33 and transmits the data to the measurement sequence control unit 42 and the error processing unit 43 . The input/output control unit 41 also inputs the error log data to the artificial intelligence engine 51 . Further, the input/output control unit 41 communicates with the cloud server 3 via the communication network 1 and transmits error log data from the error processing unit 43 to the cloud server 3 . In addition, a request for updating the artificial intelligence engine is transmitted to the cloud server 3 at a predetermined timing, an updated artificial intelligence engine distributed from the cloud server 3 is received in response to the update request, and an artificial intelligence engine storage unit 62.

The transmission of the error log data can be performed, for example, at the timing when the power of the ophthalmologic apparatus main body 10 is turned off. Also, it can be sent each time error log data is created, or it can be sent every predetermined time or at a predetermined time. Also, the input unit 33 can be used to instruct transmission. The update request and reception of the updated artificial intelligence engine can be performed, for example, when the power of the ophthalmologic apparatus main body 10 is turned on, or can be performed at predetermined time intervals or predetermined times. Also, the input unit 33 can be used to issue an update request.

The measurement sequence control unit 42 controls the measurement unit 20 in accordance with a predetermined measurement sequence stored in the storage unit 60 based on the operation data from the operation unit 31 and the input data from the input unit 33, and performs alignment and alignment. The eye characteristic measurement of optometry E is executed. In addition, when a situation arises in which eye characteristics cannot be measured by a predetermined measurement sequence, the measurement sequence control unit 42 inputs ophthalmologic measurement information such as measurement condition data and situation data in that situation to the data processing unit 50 to obtain data. According to the learning measurement sequence output from the processing section 50, the measurement section 20 is controlled to execute the measurement process again.

The situation data indicates the situation of the subject's eye E, and examples thereof include subject (subject's eye E) data and images (still images and moving images) of the subject's eye E acquired by the measuring unit 20. . The subject data includes, for example, subject's age, sex, subject ID, race, medical history of subject eye E (glaucoma, cataract, etc.), ocular characteristics of subject eye E (keratoconus, roughness of corneal apex, etc.). etc.). These can be input using the operation unit 31 and the input unit 33, or acquired from an external device or the like connected to the ophthalmologic apparatus main body 10. FIG. Examples of images include an anterior segment image E′ captured by the eye characteristic measuring unit 16, a fundus image, a fundus tomographic image (OCT image), and the like.

The error processing unit 43 collects various data when eye characteristics cannot be measured in a predetermined measurement sequence, creates error log data, and stores the error log data in the error log data storage unit 61 . The error log data includes, for example, control parameters (operation history data) such as the moving speed of the pedestal 13 and the head unit 14 during alignment in situations where measurement is not possible, and measurement condition data such as optical system measurement parameters during eye characteristic measurement. , an image such as the anterior segment image E′, situation data such as subject data, and measurement mode are linked with modality (device information for identifying the ophthalmologic apparatus main body 10), error code, and the like. is. Also, the error processing unit 43 transmits the created error log data to the input/output control unit 41 .

The data processing unit 50 functions as one of the control units that control the operation of the eye characteristic measurement unit 16 and includes an artificial intelligence engine 51 and an error avoidance processing unit 52 . The latest artificial intelligence engine stored in the artificial intelligence engine storage unit 62 is uploaded to the artificial intelligence engine 51 . The artificial intelligence engine 51 analyzes error log data in situations where measurement cannot be performed with a predetermined measurement sequence, and extracts learning measurement conditions (error avoidance information) that are highly likely to avoid such situations. The error avoidance information includes, for example, a measurement mode according to the situation data (hereinafter referred to as "error avoidance measurement mode"), a control parameter such as the movement speed of the gantry 13 for alignment (hereinafter referred to as "error avoidance control parameter"). ), focal length, gain, measurement position, exposure amount, and other measurement parameters (hereinafter referred to as “error avoidance measurement parameters”) for eye characteristic measurement. The error avoidance processing section 52 outputs the extracted error avoidance information to the measurement sequence control section 42 of the control section 40 .

Examples of the ophthalmologic apparatus main body 10 as described above include an eye characteristic measuring device (such as an autokeratometer) that measures eye refractive power and the radius of curvature of the eye, an intraocular pressure measuring device that measures intraocular pressure, and a visual acuity measuring device. an optometric apparatus for observing the fundus, a fundus camera apparatus for observing the fundus and photographing the fundus image, and a three-dimensional fundus image photographing apparatus for photographing the fundus image in three dimensions. For these ophthalmologic apparatus main bodies 10, one type or a plurality of types can be used, and a plurality of types can also be used for each one or a plurality of types.

As an example of the ophthalmologic apparatus main body 10, an eye characteristic measuring apparatus for measuring eye refractive power and eye curvature radius will be described below. As shown in FIGS. 1, 3, and 4, the ophthalmologic apparatus main body 10 includes a base 11, a drive section 12, a pedestal 13, a head section 14, a face receiving section 15, an operation section 31 as a user interface section 30, It has a display section 32 and an input section 33 . In this ophthalmologic apparatus main body 10 , a pedestal 13 is provided on a base 11 via a drive section 12 , and the pedestal 13 can be moved forward, backward, left, right, up and down with respect to the base 11 by the drive section 12 .

The base 11 is provided with a face receiving portion 15 for fixing the subject's face. A head portion 14 is provided on the base 13 . The operation unit 31 is provided on the pedestal 13, and when it is tilted, it moves the head unit 14 in the front, rear, left, and right directions, and when it rotates around the axis, it moves the head unit 14 in the vertical direction. The display unit 32 is provided in the head unit 14, and has a touch panel type display screen 32a (see FIG. 4) configured by, for example, a liquid crystal display device (LCD monitor). This display screen 32a functions as an input unit 33 for inputting subject data and the like.

The head unit 14 is provided with the aforementioned control unit 40 , data processing unit 50 and storage unit 60 . The control unit 40 comprehensively controls the operation of the ophthalmologic apparatus main body 10 based on a program stored in a storage unit 60 including a connected storage device, internal memory, and the like. The control unit 40 is connected with the light sources of the respective optical systems (target light source 22a, glare light source 22n, reflector measurement light source 23g, alignment light source 25a, alignment light source 26a, and keratling light source 27b), which will be described later. turn off the light. The control unit 40 is connected to operation units of each optical system (imaging element 21g, turret unit 22d, focusing lens 22h, VCC lens 22k, driving unit for the reflex light source unit 23a and focusing lens 23t), which will be described later. They are driven (including movement) as appropriate.

The drive unit 12, the operation unit 31, the display unit 32, and the input unit 33 are connected to the control unit 40, and according to the operation of the operation unit 31, the input from the input unit 33, and the above program, the image captured by the imaging device 21g The image is displayed on the display screen 32a of the display unit 32 as appropriate.

The head unit 14 is provided with an optical system for examining an eye to be examined. This optical system functions as an eye characteristic measuring unit 16 that measures the eye characteristic of the eye E to be examined. As shown in FIG. 3, the optical system as the eye characteristics measurement unit 16 includes an observation system 21, a target projection system 22, an eye refractive power measurement system 23, a subjective examination system 24, alignment systems 25 and 26, and a keratometric system. 27. The structures of the observation system 21, target projection system 22, eye refractive power measurement system 23, subjective inspection system 24, alignment system 25, alignment system 26, kerato system 27, etc., eye refractive power (ref), and subjective inspection Since the principle of measuring the corneal shape (keratometry) and the like are well known, a detailed description thereof will be omitted and will be briefly described below.

The observation system 21 has a function of observing the anterior segment of the subject's eye E, and includes an objective lens 21a, a dichroic filter 21b, a half mirror 21c, a relay lens 21d, a dichroic filter 21e, an imaging lens 21f, and an imaging device (CCD). 21 g. The observation system 21 forms an image of the luminous flux reflected by the subject's eye E (anterior segment) on the imaging device 21g. The control unit 40 causes the display screen 32a of the display unit 32 to display an anterior segment image E′ based on the image signal output from the imaging device 21g. A kerat system 27 is provided in front of the objective lens 21a.

The visual target projection system 22 has a function of presenting a visual target to the eye E to be examined. The subjective examination system 24 has a function of performing a subjective examination and presenting an optotype to the eye E to be examined, and shares an optical element constituting an optical system with the optotype projection system 22 .

The visual target projection system 22 and the subjective inspection system 24 include a visual target light source 22a, a color correction filter 22b, a collimator lens 22c, a turret section 22d, a half mirror 22e, a relay lens 22f, a reflecting mirror 22g, a focusing lens 22h, and a relay lens. 22i, a field lens 22j, a variable cross cylinder lens (VCC lens) 22k, a reflecting mirror 22l, and a dichroic filter 22m. The visual target projection system 22 and the subjective inspection system 24 share the dichroic filter 21b and the objective lens 21a with the observation system 21 . In addition, the subjective examination system 24 has at least two glare light sources 22n for irradiating the subject's eye E with glare light on an optical path different from the optical path from the target light source 22a.

The turret unit 22d switches the target projected by the target projection system 22 onto the fundus Ef of the eye E to be examined (presented to the eye E to be examined). The visual target projection system 22 projects the fixation target indicated by the turret portion 22d onto the fundus oculi Ef of the subject's eye E via the optical member described above. The examiner or the control unit 40 performs alignment with the examinee fixating the presented fixation target, moves the focusing lens 22h to the far point of the eye E to be examined, and then adjusts the cloudy state. Resting ocular refractive power is measured.

In the subjective examination system 24, the focusing lens 22h and the VCC lens 22k are appropriately set under the control of the control unit 40, and the target indicated by the turret unit 22d according to the measurement content is directed to the subject's eye E via the optical members described above. Project on the fundus Ef. The examiner or the control unit 40 asks the examinee how the presented optotype looks, and repeats the selection of the optotype according to the response and the question, thereby determining the prescription value. Here, when performing a glare inspection (glare test), the glare light source 22n is turned on under the control of the control section 40 .

The eye refractive power measurement system 23 has a function of measuring eye refractive power. The eye refractive power measurement system 23 detects (receives) a ring-shaped light beam projection system 23A that projects a ring-shaped measurement pattern onto the fundus Ef of the eye to be examined E, and the reflected light of the ring-shaped measurement pattern from the fundus Ef. and a ring-shaped light receiving system 23B.

The ring-shaped luminous flux projection system 23A has a reflector light source unit 23a, a relay lens 23b, a pupil ring diaphragm 23c, a field lens 23d, a perforated prism 23e, and a rotary prism 23f. Ring-shaped beam projection system 23A shares dichroic filter 22m with target projection system 22 (subjective inspection system 24), and shares dichroic filter 21b and objective lens 21a with observation system 21. The reflector light source unit 23a has a reflector measuring light source 23g for reflector measurement using an LED, a collimator lens 23h, a conical prism 23i, and a ring pattern forming plate 23j. The measurement system 23 is integrally movable along the optical axis. In the ring-shaped light beam projection system 23A, the reflector light source unit 23a emits a ring-shaped measurement pattern, which is guided to the objective lens 21a via the above-described optical members, thereby projecting the ring-shaped measurement pattern on the fundus Ef of the eye E to be examined. Project.

The ring-shaped light receiving system 23B has a hole 23p of a holed prism 23e, a field lens 23q, a reflecting mirror 23r, a relay lens 23s, a focusing lens 23t, and a reflecting mirror 23u. The ring-shaped light receiving system 23B shares the objective lens 21a, the dichroic filter 21b, the dichroic filter 21e, the imaging lens 21f, and the imaging device 21g with the observation system 21, and the dichroic filter 22m is used as the target projection system 22 (a subjective inspection system). 24), and the rotary prism 23f and perforated prism 23e are shared with the ring beam projection system 23A. The ring-shaped light receiving system 23B causes the ring-shaped measurement pattern formed on the fundus oculi Ef to form an image on the imaging device 21g via the optical member described above. As a result, the image sensor 21g detects the image of the ring-shaped measurement pattern, and the control unit 40 displays the image of the measurement pattern on the display screen 32a. Spherical power, cylindrical power, and axial angle as eye refractive power are measured by well-known methods.

The kerat system 27 has a kerat plate 27a and a kerat ring light source 27b. The keratometric system 27 projects a keratling light flux (a ring-shaped index for corneal curvature measurement) for measuring the shape of the cornea onto the cornea Ec of the eye E to be examined. The keratling luminous flux is reflected by the cornea Ec and is imaged by the observation system 21 on the imaging element 21g. Based on this, the corneal shape (curvature radius) is measured by a well-known technique. An alignment system 25 is provided behind the kerato system 27 .

The alignment systems 25 and 26 have a function of aligning the optical system with the eye E to be examined. The alignment system 25 performs alignment along the optical axis of the observation system 21 (vertical direction), and the alignment system 26 performs alignment in directions perpendicular to the optical axis (vertical direction and lateral direction).

The alignment system 25 has a pair of alignment light sources 25a and a projection lens 25b. A light beam from each alignment light source 25a is projected onto the cornea Ec as a parallel light beam by each projection lens 25b. Image on 21g. The control unit 40 or the examiner moves the head unit 14 in the front-rear direction so that the ratio of the distance between the two point images by the alignment light source 25a on the imaging element 21g and the diameter of the keratling image is within a predetermined range. Thus, alignment in the direction (front-rear direction) along the optical axis of the observation system 21 is performed. Note that the alignment in the front-rear direction may be performed by adjusting the position of the head section 14 so that the bright point image Br is focused by the alignment light source 26a, which will be described later.

Alignment system 26 is provided in observation system 21 . The alignment system 26 has an alignment light source 26a and a projection lens 26b, and shares the half mirror 21c, the dichroic filter 21b and the objective lens 21a with the observation system 21. The alignment system 26 projects the luminous flux from the alignment light source 26a as a parallel luminous flux onto the cornea Ec, and the observation system 21 forms an image on the imaging element 21g. The control unit 40 or the examiner moves the head unit 14 in the vertical and horizontal directions based on the luminescent spots (luminescent spot images) projected on the cornea Ec, thereby moving the head unit 14 in a direction orthogonal to the optical axis of the observation system 21 (vertical direction). , horizontal direction). At this time, the control unit 40 causes the display screen 32a to display an alignment mark AL, which serves as a guide for the alignment mark, in addition to the anterior segment image E′ formed with the bright spot image Br.

An ophthalmologic measurement process (ophthalmologic measurement method) performed by the ophthalmologic apparatus 100 according to the first embodiment configured as described above will be described. In the ophthalmologic apparatus 100 , the ophthalmologic apparatus body 10 performs eye characteristic measurement processing, and the cloud server 3 performs learning processing.

First, an example of eye characteristic measurement processing performed by the ophthalmologic apparatus main body 10 will be specifically described with reference to the flowchart of FIG. The flowchart of FIG. 5 is started when an operation is performed in the ophthalmologic apparatus main body 10 to start measurement of eye characteristics.

First, in step S<b>1 , the measurement condition acquisition unit 17 starts acquisition of measurement condition data in the eye characteristics measurement process in the eye characteristics measurement unit 16 . Next, in step S2, the input/output control unit 41 acquires subject data such as age input by the examiner from the operation unit 31, the input unit 33, or the like, and the measurement mode.

Next, in step S3, the measurement sequence control unit 42 of the control unit 40 controls alignment control parameters and eye characteristics measurement according to a predetermined measurement sequence according to subject data and measurement mode. The initial setting of the eye characteristic measuring unit 16 is performed by setting measurement parameters and the like.

After that, the process proceeds to step S4, and alignment is performed by the alignment systems 25 and 36 of the eye characteristic measurement unit 16. FIG. As the measurement condition data, the measurement condition acquisition unit 17 acquires control parameters (operation history data) such as the movement speed and movement coordinates of the gantry 13 at the time of alignment. In addition, an anterior segment image E′ of the subject's eye E, a fundus image, a fundus tomographic image, and the like are acquired by the imaging element 21g of the observation system 21 as situation data.

If the alignment systems 25 and 26 cannot perform the automatic alignment, the examiner may perform manual alignment. Manual control parameters (operation history data) at this time are also acquired by the measurement condition acquisition unit 17 as measurement condition data.

After the alignment is completed, the process proceeds to step S5. In step S5, the measurement sequence control section 42 controls the eye characteristic measuring section 16 according to a predetermined measurement sequence, so that the eye characteristic measuring section 16 measures the eye characteristic of the subject's eye E by the operation described above. Measurement parameters such as focal length, gain, measurement position, and exposure amount at this time are acquired by the measurement condition acquisition unit 17 as measurement condition data.

In the next step S6, it is determined whether or not the eye characteristics have been properly measured. If the measurement is properly performed and the determination is YES, the process proceeds to step S7, the acquisition of the measurement condition data is terminated, and the eye characteristic measurement process is terminated. The measurement condition data at this time is stored in the storage unit 60 as data indicating the situation in which the measurement was properly performed, and can be used for learning as a successful example. can also be discarded.

On the other hand, if it is determined NO in step S6 and the eye characteristics cannot be appropriately measured, the process proceeds to step S8. Such situations include, for example, eyelids and eyelashes covering the anterior segment (cornea Ec) of the subject's eye E, ocular tremor (nystagmus), frequent blinking, and keratoconus. As a result, the alignment cannot be properly performed automatically or manually, and as a result, the eye characteristics cannot be measured properly. In addition, there is a situation in which the eye characteristics cannot be properly measured due to insufficient exposure in the optical system, out-of-focus, or the like, even though the alignment has been completed. In addition, there is a situation in which, when calculating a measurement value such as a spherical power based on an image signal, the numerical value cannot be calculated due to an overflow or the like.

In step S8, the error processing unit 43 creates error log data (ophthalmologic measurement information). In this embodiment, the error log data includes the modality of the ophthalmologic apparatus main body 10, measurement mode, error code, measurement condition data including measurement parameters and control parameters, images such as the anterior ocular segment image E′, age, sex, and person. Circumstance data consisting of subject data such as species, medical history, and the like.

In the next step S<b>9 , the error processing unit 43 stores the created error log data in the error log data storage unit 61 and outputs it to the input/output control unit 41 . The input/output control unit 41 transmits the error log data to the cloud server 3 at a predetermined timing, and proceeds to step S10.

Then, in step S10, it is determined whether or not it is the first measurement, and if it is determined as YES (first measurement), the process proceeds to step S11.

In step S<b>11 , the input/output control unit 41 inputs error log data (ophthalmic measurement information) to the artificial intelligence engine 51 in order to avoid situations in which eye characteristics cannot be measured in a predetermined measurement sequence. In the next step S12, the artificial intelligence engine 51 analyzes the error log data and extracts error avoidance information such as an error avoidance measurement mode, an error avoidance measurement parameter, and an error avoidance control parameter with a high possibility of avoiding the situation. The extracted error avoidance information is output to the measurement sequence controller 42 by the error avoidance processor 52 .

Upon receiving the error avoidance information, the measurement sequence control unit 42 initializes the eye characteristic measurement unit 16 based on the error avoidance measurement mode, error avoidance measurement parameters, and error avoidance control parameters in step S13. After that, the process returns to step S4, and the alignment in step S4 and the eye characteristic measurement in step S5 are performed based on the initial settings. If the eye characteristics can be appropriately measured (YES in step S6), acquisition of measurement condition data ends (step S8), and the measurement process ends. By extracting the error avoidance information in this way, the eye characteristics can be appropriately measured, and the measurement performance of the ophthalmologic apparatus main body 10 can be improved.

In addition, when the measurement can be properly performed by the error avoidance information in this way, the error avoidance information, the ophthalmologic measurement information at that time, and the like may be transmitted to the cloud server 3 . As a result, the cloud server 3 can collect examples of successful workarounds, and the learning performance can be further improved.

On the other hand, even if the error avoidance information is used, if the eye characteristics cannot be properly measured (determination in step S6 is NO), the error log data is created in step S8, the error log data is stored in step S9, and the cloud server Send to 3. At the time of this storage and transmission, information may be added indicating that the error log data is in a situation where the eye characteristics cannot be measured even in the second measurement using the error avoidance information. As a result, the range of learning in the cloud server 3 can be expanded, and the learning performance can be improved.

After that, by proceeding to step S10, it is determined whether or not it is the first measurement. Since this measurement is the second time, NO is determined in step S10, and the process advances to step S7 to end acquisition of the measurement condition data and end the measurement process.

Next, an example of the ophthalmologic measurement learning process performed by the cloud server 3 will be described with reference to the flowcharts of FIGS. 6A to 6D. As shown in FIG. 6A, in the cloud server 3, the data accumulation process in step S20, the learning process in step S30, and the artificial intelligence engine delivery process in step S40 are performed independently while the cloud server 3 is running. It is executed repeatedly under a predetermined condition.

The data accumulation processing in step S20 will be described using the flowchart of FIG. 6B. The data accumulation process is executed at the timing when error log data is received from the ophthalmologic apparatus main body 10 . As shown in FIG. 6B, the cloud server 3 receives the error log data transmitted from the ophthalmologic apparatus main body 10 in step S21. Next, in step S22, the received error log data is separated for each item. In the first embodiment, the items include the modality of the ophthalmologic apparatus main body 10, the measurement mode, the error code, the measurement parameters and control parameters of the measurement condition data, the anterior ocular segment image E' of the situation data, the subject data, and the like. .

The cloud server 3 stores the separated data in the data storage unit 7 by modality, for example (step S23).

Next, the learning process of step S30 will be described using the flowchart of FIG. 6C. The learning process is executed, for example, every predetermined period when a predetermined amount of data is accumulated in the data accumulation unit 7 . Further, the learning process can be executed for each model of the ophthalmologic apparatus main body 10 using modality as a key, and the learning result can be shared by a plurality of ophthalmologic apparatus main bodies 10 of the same model.

First, in step S31, error log data accumulated for a predetermined period is obtained from the data accumulation unit 7. FIG. By inputting the acquired error log data to the learning unit 6, which is an artificial intelligence engine (step S32), learning by artificial intelligence is executed (step S33).

The learning unit 6 uses, for example, a deep learning algorithm to generate workarounds (error avoidance information) for situations where measurement is impossible. FIG. 7 shows a configuration example of an artificial intelligence engine learned by applying a neural network. As shown in FIG. 7, artificial intelligence connects nodes by learning based on input data, changes the connection strength to optimize the number of nodes, and finally selects an error avoidance measurement mode and an error avoidance measurement mode. Output error avoidance information (learning measurement condition data) such as parameters and error avoidance control parameters.

In addition, the learning method by artificial intelligence is not limited to the learning method of Example 1, and any of machine learning methods such as expert systems, case-based reasoning, Bayesian networks, fuzzy theory, evolutionary calculation, etc. may be used.

Next, in step S34, it is determined whether or not the predetermined function has been exhibited. Exhibiting a predetermined performance here means that a workaround (error avoidance information) for avoiding a situation in which learning progresses and eye characteristics cannot be measured appropriately has been acquired to some extent.

If the determination in step S34 is YES, it is determined that the learning has been sufficiently performed, and the process proceeds to step S35. On the other hand, if the determination is NO, it is determined that the learning is insufficient, and the process returns to step S33 to continue the learning by the artificial intelligence.

The determination in step S34 can also be automatically determined by artificial intelligence through simulation or the like. Alternatively, the developer or the like of the ophthalmologic device main body 10 conducts a test on an actual device using the learning results of artificial intelligence, and the function is exhibited when the measurement is performed appropriately or when the success rate reaches or exceeds a predetermined value. It is also possible to make an input to the cloud server 3 to the effect that it has been done. When receiving this input, the cloud server 3 can determine YES in step S34 and proceed to step S35.

In step S35, the artificial intelligence engine that has undergone sufficient learning in this manner is stored in the artificial intelligence engine storage unit 8 as the latest version of the artificial intelligence engine.

Finally, the artificial intelligence engine distribution processing in step S40 will be described using the flowchart of FIG. 6D. Update requests for the artificial intelligence engine are sent to the cloud server 3 from a plurality of ophthalmologic apparatus main bodies 10 via the communication network 1 . As shown in FIG. 6D, in step S41, when the cloud server 3 receives the update request of this artificial intelligence engine, in step S42, the latest artificial intelligence engine corresponding to each ophthalmologic apparatus main body 10 is updated using modality or the like as a key. is obtained from the artificial intelligence engine storage unit 8. Then, in step S43, each of the acquired artificial intelligence engines is delivered to the main body 10 of the ophthalmologic apparatus that made the request. As a result, the latest artificial intelligence engine can be used in each ophthalmologic apparatus main body 10, the possibility of automatically measuring eye characteristics is expanded, and the measurement performance can be improved. In addition, it is also possible to manage the version of the distributed AI engine, and if the AI engine has not been updated since the last update request, reply to that effect and not distribute the AI engine. can. Alternatively, it is also possible to omit version management and automatically deliver the artificial intelligence engine of the artificial intelligence engine storage unit 8 to the ophthalmologic apparatus main body 10 each time an update request is made.

As described above, in the ophthalmologic apparatus 100 of the first embodiment, when the eye characteristics cannot be appropriately measured by the predetermined measurement sequence in the eye characteristics measuring unit 16, the situation data and the measurement condition data Error log data (ophthalmic measurement information) is stored and accumulated, and workarounds (learning measurement conditions) are learned by the learning unit 6, which is artificial intelligence. The learned workaround (learned measurement condition) is reflected in the ophthalmologic apparatus main body 10 . Therefore, even in a situation where eye characteristics cannot be appropriately measured in a predetermined measurement sequence, the ophthalmologic apparatus main body 10 extracts a workaround corresponding to the situation from the learned workarounds. The possibility of being able to measure appropriately can be expanded. Therefore, the situation in which the eye characteristics of the eye to be examined E can be automatically measured can be expanded, and the measurement performance of the ophthalmologic apparatus 100 can be improved.

Further, the ophthalmologic apparatus 100 of the first embodiment includes a plurality of ophthalmologic apparatus main bodies 10 , which are connected to the learning unit 6 via the communication network 1 . Furthermore, in Example 1, the learning unit 6 is provided in the cloud server 3 . The learning unit 6 learns based on error log data from a plurality of ophthalmologic apparatus main bodies 10 and generates error avoidance information. Therefore, each ophthalmologic apparatus main body 10 can share learning results for various situations that cannot be measured by a predetermined measurement sequence, and learning performance can be improved. Further, since more error log data can be collected by the cloud server 3, the learning performance of the learning unit 6 can be improved. Therefore, it is possible to expand the situation in which the eye identification of the subject's eye E can be automatically measured in each ophthalmologic apparatus main body 10, and to further improve the measurement performance.

In addition, since learning results can be shared by a plurality of ophthalmic apparatus main bodies 10 of the same model, for example, an ophthalmologic apparatus main body 10 that mainly measures the eye characteristics of elderly people and an ophthalmological device main body 10 that mainly measures the eye characteristics of children An artificial intelligence engine can be generated by collecting error log data in the device body 10 . By distributing this artificial intelligence engine to each ophthalmologic apparatus main body 10, it is possible to share experience of errors in each ophthalmologic apparatus main body 10 and workarounds therefor, so that which ocular characteristics can be used by the ophthalmologic apparatus main body 10 and which patient Even when measuring the eye characteristics of optometry E, by measuring based on the error avoidance information corresponding to various situations, the possibility of automatic measurement can be expanded, and the measurement performance can be improved. can.

In addition, in the ophthalmologic apparatus 100 of the first embodiment, the measurement condition data includes control parameters such as a movement speed and movement coordinates for moving the eye characteristics measuring unit 16 with respect to the eye to be examined E, and It includes any of the measurement parameters such as focal length, gain, measurement position, and exposure when measuring the characteristics. Based on these measurement condition data, the learning unit 6 can generate error avoidance control parameters and error avoidance measurement parameters that are highly likely to be measured depending on the situation, and can return them to each ophthalmologic apparatus main body 10 . Therefore, when the ocular characteristics cannot be measured by each ophthalmologic apparatus main body 10, the possibility of automatic measurement can be expanded by setting these parameters and performing the measurement. Automated measurements can be made more quickly.

Further, the ophthalmologic apparatus 100 of the first embodiment includes the degree of exposure of the anterior segment and the ocular characteristics of the eye E to be examined in the situation (situation data) of the eye E to be examined. For this reason, the ophthalmologic apparatus main body 10 may be used in a situation where the eyelid or eyelashes are covered with the anterior segment (cornea Ec) and the degree of exposure is low, a situation where the subject's eye E has keratoconus, or a corneal vertex is rough. Even in situations, the possibility of automatic measurement can be expanded by measuring based on the error avoidance information.

The ophthalmic apparatus of the present application has been described above based on Example 1, but the specific configuration is not limited to Example 1, and does not depart from the gist of the invention according to each claim. Design changes, additions, etc. are permitted as long as

For example, in Example 1, an observation system 21 , a target projection system 22 , an eye refractive power measurement system 23 , a subjective examination system 24 , and a keratometry system 27 are provided as the eye characteristic measurement unit 16 . However, the eye characteristic measurement unit 16 is not limited to the configuration of the first embodiment as long as it can automatically measure the eye characteristics of the eye to be examined E after automatically performing alignment with the eye to be examined E.

Further, in the first embodiment, alignment and eye characteristics are first measured according to a predetermined measurement sequence, and if the measurement is not possible, a workaround is extracted by the artificial intelligence engine, and based on the workaround, alignment and eye characteristics are again measured. are performing measurements. However, the present application is not limited to this, and situation data such as the input subject data and the captured anterior segment image E′ is analyzed in advance to determine whether or not appropriate measurement can be performed. However, if it is determined that the situation cannot be measured, the artificial intelligence engine can extract workarounds and perform measurement. Therefore, the time required for automatic measurement can be shortened. This is because measurement can be prevented in vain as compared to the case where measurement is performed according to a predetermined measurement sequence regardless of the situation, and a workaround is extracted when measurement cannot be performed.

Further, a speaker or the like may be provided as the user interface unit 30, and as a workaround, the subject may be notified by voice that he/she should refrain from blinking or the timing of opening the eyelids. As a result, it is possible to avoid a situation in which the eyelids or eyelashes overlap the anterior segment of the eye and cannot be measured, or a situation in which measurement is impossible due to blinking.

6 Learning Unit 10 Ophthalmic Apparatus Main Body 16 Eye Characteristic Measuring Unit 17 Measurement Condition Acquisition Unit 40 Control Unit 50 Data Processing Unit (Control Unit) 100 Ophthalmic Apparatus

Claims (4)

an eye characteristic measuring unit that measures the eye characteristic of the subject's eye;
a measurement condition acquisition unit that acquires measurement conditions when the measurement is performed by the eye characteristic measurement unit;
a control unit that controls the eye characteristic measurement unit according to a predetermined measurement procedure;
Said measurement conditions in a situation in which eye characteristics cannot be measured in said measurement procedure, situation data relating to the condition of said eye to be examined, modality of an ophthalmologic apparatus main body, and a situation in which eye characteristics are properly measured in said measurement procedure a learning unit that generates a learning measurement condition that avoids a situation in which measurement cannot be performed, based on the measurement condition, the situation data, and the modality in
The control unit extracts a learning measurement condition corresponding to the situation from the learning measurement conditions created by the learning unit when the measurement procedure cannot measure the eye characteristics of the eye to be examined. controlling the eye characteristics measuring unit based on the learning measurement conditions to measure the eye characteristics of the eye to be inspected ;
When the control unit controls the eye characteristic measurement unit based on the learning measurement condition, and the eye characteristic measurement is appropriately performed, the learning unit determines whether the eye characteristic measurement is properly performed. 1. An ophthalmologic apparatus, characterized in that the learned measurement conditions are collected as examples of successful cases, and referred to when generating the learned measurement conditions for avoiding a situation in which measurement cannot be performed. an eye characteristic measuring unit that measures the eye characteristic of the subject's eye;
a measurement condition acquisition unit that acquires measurement conditions when the measurement is performed by the eye characteristic measurement unit;
a control unit that controls the eye characteristic measurement unit according to a predetermined measurement procedure;
Said measurement conditions in a situation in which eye characteristics cannot be measured in said measurement procedure, situation data relating to the condition of said eye to be examined, modality of an ophthalmologic apparatus main body, and a situation in which eye characteristics are properly measured in said measurement procedure a learning unit that generates a learning measurement condition that avoids a situation in which measurement cannot be performed, based on the measurement condition, the situation data, and the modality in
The control unit extracts a learning measurement condition corresponding to the situation from the learning measurement conditions created by the learning unit when the measurement procedure cannot measure the eye characteristics of the eye to be examined. controlling the eye characteristics measuring unit based on the learning measurement conditions to measure the eye characteristics of the eye to be inspected;
When the eye characteristics cannot be measured even if the eye characteristics measuring unit is controlled based on the learning measurement conditions, the control unit changes the conditions in which the measurement cannot be performed under the learning measurement conditions. Additional information indicating that it is an error measurement condition is added, and the learning unit converts the learning measurement condition to which the additional information indicating that it is an error measurement condition is a learning measurement condition that avoids a situation in which measurement cannot be performed. An ophthalmologic device that is referred to when generating. an eye characteristic measuring unit that measures the eye characteristic of the subject's eye ;
a measurement condition acquisition unit that acquires measurement conditions when the measurement is performed by the eye characteristic measurement unit ;
a plurality of ophthalmologic apparatus main bodies each having a control unit that controls the eye characteristics measuring unit according to a predetermined measurement procedure; The measurement can be performed based on the situation data related to the optometry situation, the modality of the main body of the ophthalmologic apparatus, and the measurement conditions, the situation data, and the modality in the situation where the eye characteristics are appropriately measured in the measurement procedure. a server having a learning unit that generates a learning measurement condition for avoiding a situation where the modality does not exist, and a storage unit that stores the measurement condition and the situation data in association with the modality;
each ophthalmologic apparatus body and the server are connected to each other via a communication network,
The control unit extracts a learning measurement condition corresponding to the situation from the learning measurement conditions created by the learning unit when the measurement procedure cannot measure the eye characteristics of the eye to be examined. controlling the eye characteristics measuring unit based on the learning measurement conditions to measure the eye characteristics of the eye to be inspected;
Each ophthalmologic apparatus main body has the measurement conditions and the modality in a situation in which the eye characteristics cannot be measured in the measurement procedure, and the measurement conditions and the modality in a situation in which the eye characteristics are appropriately measured in the measurement procedure. , the measurement conditions and the modalities in a situation where the eye characteristics are appropriately measured under the learning measurement conditions, and the measurement conditions and the modalities in a situation where the eye characteristics cannot be measured under the learning measurement conditions. , through the communication network to the server;
The server acquires the situation data from the main body of the ophthalmologic apparatus or an external device via the communication network using the modality as a key, storing the received measurement conditions in the storage unit in association with the modality;
The learning unit acquires the measurement condition and the situation data from the storage unit using the modality as a key, generates the learning measurement condition for each modality, and transmits the generated learning measurement condition to the communication network. An ophthalmologic device characterized by transmitting to each ophthalmologic device main body via. An ophthalmologic measurement method performed by the ophthalmologic apparatus according to claim 1 ,
a step of measuring ocular characteristics of an eye to be examined according to a predetermined measurement procedure;
obtaining measurement conditions for measuring the eye characteristics;
a step of obtaining situation data about the situation of the eye to be examined and the modality of the ophthalmologic apparatus main body;
The measurement conditions, the situation data, and the modality in a situation in which the eye characteristics cannot be measured, and the measurement conditions, the situation data, and the above in a situation in which the eye characteristics are appropriately measured in the measurement procedure. generating learning measurement conditions that avoid situations in which measurements cannot be performed, based on the modality;
When a situation arises in which the eye characteristics cannot be measured, the learning measurement condition corresponding to the situation is extracted from the learning measurement conditions generated in the step of generating the learning measurement condition, and the learning measurement condition is extracted. and measuring the ocular characteristics of the subject eye based on
The step of generating a learning measurement condition for avoiding a situation in which the measurement cannot be performed includes measuring the eye characteristics appropriately when the step of measuring the eye characteristics of the subject's eye is appropriately performed based on the learning measurement conditions. Collecting the learning measurement conditions when performed at the time as examples of successful cases, and referring to when generating the learning measurement conditions,
An ophthalmologic measurement method characterized by:

Images