Eye examination system and eye examination program

Abstract

To provide an eye examination system and an eye examination program for allowing an examiner to easily determine the quality validity of a photographed ophthalmologic image.SOLUTION: An eye examination system includes photographing means for acquiring an ophthalmologic image by photographing an eye of a subject, quality target setting means for setting a quality target being a quality target in the ophthalmologic image on the basis of medical examination information of the subject, quality evaluation means for acquiring a quality evaluation result obtained by evaluating the quality of the ophthalmologic image of the subject which is actually photographed, and display control means for displaying both the quality target and the quality evaluation result or a quality comparison result being a comparison result of both of them.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to an eye examination system and an eye examination program. [Background technology]

[0002] Fundus imaging devices such as fundus cameras and OCT devices are widely used in the field of ophthalmology. When the quality of an image captured by the imaging device is not good, the image may be re-captured (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2017-104708 Summary of the Invention [Problem to be solved by the invention]

[0004] However, there may be cases where it is difficult to improve the quality of an image even if the image is re-photographed.

[0005] Here, the quality of the image may depend on factors such as the presence or absence of abnormalities in the subject's eye, medical history, and individual differences. For example, myopic and hyperopic eyes are more likely to have reflection artifacts in the optical system. Eyes with small pupils are more likely to have vignetting (a type of artifact) and uneven brightness in the image. Furthermore, for example, if the optic body of the subject's eye is opaque, the opacity may interfere with the projection and reception of imaging light by the fundus imaging device, making it difficult to adjust the optical system and, further, making it impossible to capture fundus images with good brightness or contrast. Furthermore, if there is a disorder in the optic nerve system, unstable fixation during imaging may degrade the image quality.

[0006] However, it may be difficult for the examiner to grasp the above reasons in advance when taking an image. Even if the examiner were to grasp the above reasons in advance, it would be difficult for the examiner to determine the appropriate level of image quality. Therefore, if the quality of the taken ophthalmologic image is not evaluated as good, it is likely that the examiner will have to take the image again, which will result in more rework than necessary.

[0007] Furthermore, for example, the level of proficiency in testing varies from examiner to examiner, and the acceptable image quality may also differ between examiners. For examiners with low proficiency, it may be difficult to improve the quality by retaking images.

[0008] Furthermore, image quality can sometimes be improved with assistance and adjustments. For example, good images may not be obtained if the eyelids are not open enough during imaging, blinking occurs, or alignment or focus is shifted due to head or eye movement. These can be improved with eyelid opening assistance, face support assistance, alignment adjustment, and adjustment of imaging timing.

[0009] The present disclosure has been made based on at least some of the problems of the prior art, and has as its technical objective the provision of an eye examination system and an eye examination program that allow the examiner to easily judge the validity of the quality of the captured ophthalmic images. [Means for solving the problem]

[0010] The eye examination system according to the first aspect of the present disclosure includes an imaging means for acquiring ophthalmic images by photographing the subject's eyes, a quality target setting means for setting a quality target, which is a target for the quality of the ophthalmic images, based on the subject's medical information, a quality evaluation means for acquiring a quality evaluation result that evaluates the quality of the ophthalmic images actually captured of the subject, and a display control means for displaying both the quality target and the quality evaluation result, or a quality comparison result that is a comparison result of both. An eye examination program according to a second aspect of the present disclosure is executed by a processor of an eye examination device, causing the eye examination device to acquire ophthalmic images by photographing the subject's eyes, set a quality target that is a target for the quality of the ophthalmic images based on the subject's medical information, acquire a quality evaluation result that evaluates the quality of the ophthalmic images actually photographed of the subject, and display both the quality target and the quality evaluation result, or a quality comparison result that is a comparison result of both. An eye examination system according to a third aspect of the present disclosure includes an imaging means for acquiring ophthalmic images by photographing the subject's eye, an examination condition acquisition means for acquiring examination conditions for the imaging means based on the subject's medical information, and an examination condition setting means for setting the examination conditions acquired by the examination condition acquisition means for the imaging means. An eye examination program according to a fourth aspect of the present disclosure is an eye examination program that, when executed by a processor of an eye examination device, causes the eye examination device to acquire examination conditions for the imaging means based on the subject's medical information, and set the examination conditions acquired by the examination condition acquisition means for the imaging means. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram showing a schematic configuration of an eye examination system in an embodiment. [Figure 2] 10 is a flowchart showing the flow of a photographing operation. [Figure 3] FIG. 2 is a diagram showing input training data and output training data used in training the first mathematical model. [Figure 4] FIG. 10 is a diagram showing input training data and output training data used to train a second mathematical model. [Figure 5] 10 is an observation screen in the embodiment. [Figure 6] 10 is a confirmation screen according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] [overview] Embodiments according to the present disclosure will be described below, and each embodiment may be applied to part or all of the other embodiments.

[0013] The eye examination system according to the first embodiment includes at least an imaging unit, a quality target setting unit, a quality evaluation unit, and a display control unit. The imaging unit is provided in an ophthalmic imaging device. Any of the components of the eye examination system other than the imaging unit may be provided in the ophthalmic imaging device or in a computer connected to the ophthalmic imaging device.

[0014] The imaging unit acquires ophthalmologic images by capturing images of the subject's eye, and may be any of an OCT device, a frontal fundus image capturing device, a Scheimpflug camera, a specular microscope, etc.

[0015] The quality target setting unit sets a quality target, which is a quality target for the ophthalmologic image, based on the medical information of the subject. The medical information here refers to at least one of the subject's physical condition, medical condition, treatment, etc., and may be, for example, text or numerical data related to treatment, progress, diagnosis, or the like. Specific examples include the subject's age, gender, symptoms, medical history, family history, progress, diagnosis, test results, etc. The medical information of the subject may be acquired, for example, from an external ophthalmologic device, an external storage device such as a server that stores electronic medical records and ophthalmologic information, or the like. Furthermore, predetermined keywords may be extracted from free text written in the electronic medical record, and the medical information of the subject.

[0016] The quality of an ophthalmologic image may be evaluated based on at least one of an index indicating signal quality (e.g., SSI (Signal Strength Index) or SQI (SlO Quality Index)), a ratio of image signal level to noise level (SNR (Signal to Noise Ratio)), background noise level, image contrast, etc. However, the quality is not necessarily limited to these, and may be evaluated based on other indexes.

[0017] Furthermore, for example, the quality target may be another ophthalmic image (hereinafter referred to as a target image) having the target image quality. A plurality of target images of different qualities may be stored in an accessible database. Each target image is associated with medical information, and a target image having medical information similar to that of the subject may be acquired as the quality target. The target image does not necessarily have to be an image stored in a database in advance; for example, an ophthalmic image of the subject or the like modified to have the target image quality may be used as the target image. The modification may be performed by appropriately using known image processing.

[0018] The quality evaluation unit obtains a quality evaluation result obtained by evaluating the quality of an ophthalmologic image of an actual subject. If the quality target is an index indicating image quality such as SSI, the quality evaluation result may also be set using the same type of index.

[0019] The display control unit displays both the quality target and the quality evaluation result, or a quality comparison result that is a comparison result of both. Here, if the quality target is a target image, the captured ophthalmic image may be displayed side by side with the target image. This allows the examiner to intuitively judge the quality of the captured ophthalmic image by visually comparing the images. For example, the quality comparison result may indicate the result of a comparison between the captured ophthalmic image and the target image through image processing. For example, the similarity between the images or the difference in the results of a predetermined analysis process for each image can be used as the quality comparison result.

[0020] According to the first embodiment, a quality target for each ophthalmologic image is appropriately set based on the patient's medical information. Furthermore, the examiner can check both the quality target and the quality evaluation result, or the quality comparison result, which is a comparison of both. This makes it easier to reduce rework such as retaking images.

[0021] In this embodiment, the display control unit displays both the quality target and the quality evaluation result, or the quality comparison result, on the same screen as the ophthalmic image of the subject's eye showing the imaging result. In this case, the screen may be capable of accepting an operation input for re-imaging. When an operation input for re-imaging is accepted, re-imaging is performed. A UI element for accepting an operation input for re-imaging and the ophthalmic image of the subject's eye may be displayed simultaneously. When the examiner confirms that the quality evaluation result is not sufficient relative to the quality target, re-imaging can be performed promptly.

[0022] In this embodiment, the quality target setting unit may have a pre-trained mathematical model that receives input data containing clinical information about a subject and outputs a quality target corresponding to the input data. This mathematical model is pre-trained using the clinical information about multiple subjects as input training data and the quality evaluation results obtained based on ophthalmologic images of the multiple subjects as output training data. Because clinical information contains a wide variety of information, the parameters that affect image quality can be diverse and complex. In response to this, appropriate quality targets can be set by using the above-described mathematical model.

[0023] In this embodiment, the quality target setting unit may further acquire proficiency information regarding the proficiency of the examiner, such as the examiner's job title, the cumulative number of examinations, and the quality evaluation results of the examinations that the examiner has performed.

[0024] In this case, the quality target setting unit may set the quality target based on the proficiency information and the medical information of the subject. For example, the quality target may be adjusted so that an examiner with low proficiency has a lower quality target than an examiner with high proficiency.

[0025] Furthermore, the proficiency information may be used for purposes other than setting quality targets. For example, the proficiency information may be used to determine whether an examiner is capable of capturing images that satisfy the quality targets. In this case, the quality targets are set based on the medical information of the examinee, and whether an examiner has the proficiency corresponding to the quality targets may be determined based on the proficiency information. If an examiner does not have sufficient proficiency for the quality targets, a notification may be output urging a change of examiner. Since the examiner is more likely to be performed by an examiner with sufficient proficiency, rework of the examination is less likely to occur.

[0026] The eye examination system according to this embodiment may further include an examination condition setting unit. The examination condition setting unit acquires examination conditions for the imaging unit based on the medical information of the subject. In this case, the examination conditions may be associated with the medical information in advance. The examination conditions may be, for example, conditions related to the adjustment of the optical system in the imaging unit. Examples include focus, light intensity, and sensitivity. The examination conditions may also be, for example, conditions related to at least one of the alignment state of the imaging unit with respect to the subject's eye and the presentation position of the fixation target. The examination conditions may also be, for example, conditions related to the on / off of various automatic adjustment controls such as auto-alignment and auto-optimization.

[0027] By setting the examination conditions based on the medical information of the subject, even an inexperienced examiner can perform imaging under examination conditions appropriate for the subject. Therefore, when imaging a subject for whom it is difficult to capture high-quality ophthalmic images, even if the quality evaluation result of the ophthalmic image is not high, it is considered to be an image of reasonable quality, thereby reducing unnecessary re-imaging where quality improvement is not expected.

[0028] Here, the examination condition setting unit may have a second mathematical model. The second mathematical model is pre-trained so that, when medical information of a subject is input as input data, the second mathematical model outputs examination conditions corresponding to the input data. The second mathematical model may be pre-trained with medical information about multiple subjects as input training data and with examination conditions used when imaging multiple subjects as output training data. There is a wide variety of medical information, and the examination conditions vary from device to device, resulting in diverse and complex parameters. In response to this, appropriate examination conditions can be set by using the above-described mathematical model.

[0029] [Example] Examples according to embodiments of the present disclosure will be described with reference to the drawings.

[0030] 1 shows a schematic configuration of an eye examination system 1 in the first embodiment. The eye examination system 1 includes an ophthalmologic imaging device 10, a PC (personal computer) 20, and a medical information system 30. Each device may be connected to each other via a network.

[0031] <Ophthalmic imaging device> In this embodiment, an OCT device 10 is used as an example of an ophthalmic imaging device. The OCT device 10 acquires OCT data of the subject's eye via an imaging optical system (not shown). For example, the imaging optical system includes an OCT light source, a scanning unit for scanning the OCT light, an optical system for irradiating the subject's eye with the OCT light, and a light-receiving element for receiving light reflected by the tissue of the subject's eye. Unless otherwise specified, in this embodiment, OCT data of the fundus is acquired. For example, the OCT data may be two-dimensional OCT data (B-scan data) or three-dimensional OCT data. A two-dimensional tomographic image is generated based on the two-dimensional OCT data. Furthermore, a three-dimensional image and an OCT front image (also referred to as an en-face image) are acquired based on the three-dimensional OCT data. Motion contrast data may also be used. The motion contrast may be information capturing, for example, blood flow in the subject's eye, changes in retinal tissue, etc. MC data is acquired by processing multiple OCT data acquired at the same position but at different times.

[0032] The OCT device 10 may further include an observation optical system. The observation optical system is an optical system different from the above-described photographing optical system, and acquires a front image of the fundus as an observation image. The observation optical system may be a fundus camera optical system, an SLO optical system, or other optical systems.

[0033] <Ophthalmology Computer> In this embodiment, a PC 20 is used as an example of an ophthalmologic computer. In this embodiment, the PC 20 also serves as a control unit that controls the operation of the OCT device 10. The PC 20 also acquires at least ophthalmologic images of the subject's eye (various OCT images and observation images) captured via the OCT device 10 as examination results. The PC 20 includes at least a processor 21 (processing device) and a memory 22.

[0034] In this embodiment, the memory 22 pre-stores a control program for the OCT device 10, an image processing program for processing OCT data, an ophthalmologic information processing program for generating and transferring an examination report, fixed value data, etc. The various programs are read and executed by the processor 21. The memory 22 may also store examination results acquired via the OCT device 10. For example, examination results of the subject's eye acquired in a previous examination may be stored in the memory 22 in advance. An operation unit 23 such as a mouse and keyboard may be connected to the PC 20. Various operations are input via the operation unit 23.

[0035] A monitor 25 is also connected to the PC 20. The monitor 25 displays ophthalmologic images captured by the OCT device 10. It also displays various GUIs used for imaging.

[0036] <Medical Information System> The medical information system 30 digitizes medical information and manages it centrally using a server. The managed medical information includes at least ophthalmologic images and medical information. The medical information includes information on the subject's physical condition, medical condition, treatment, etc. For example, the medical information may be text or numerical data related to treatment, progress, diagnosis, etc. Specific examples include the subject's age, sex, symptoms, medical history, family history, progress, diagnosis, test results, etc. The medical information is managed in association with the subject's identification information. During the examination, at least some of this clinical information is acquired by the PC 20 and used to set quality targets for the ophthalmic images.

[0037] Various types of medical information managed by the medical information system 30 can be displayed on a diagnostic terminal (not shown), thereby enabling the division of labor between examination and diagnosis.

[0038] <Operation description> Next, the flow of operations performed when examining an eye to be examined in the eye examination system 1 will be described with reference to the flowchart of FIG.

[0039] First, the PC 20 acquires the subject's identification information (S1). By registering the subject's name and ID in advance, the subject can be searched for on the search screen and the subject's identification information can be acquired. This allows, for example, the subject's information to be associated with the test results to be acquired.

[0040] Next, the PC 20 acquires medical information of the identified subject from the medical information system 30 (S2). For example, in this embodiment, the acquired medical information includes age (Age), refractive power (Ref), axial length (AL), fixation quality (Fixation), and presence or absence of cataract (Cataract). The PC 20 acquires and sets a quality target for the ophthalmologic image to be captured based on the medical information (S3).

[0041] In this embodiment, a first mathematical model is used in which the relationship between clinical information and quality targets is trained by a machine learning algorithm.

[0042] A mathematical model refers to, for example, a data structure for predicting the relationship between input data and output data. The mathematical model is constructed by training using a training data set. In this embodiment, the training data set is a set of input training data and output training data. For example, the correlation data (e.g., weights) between each input and output is updated through training. In this embodiment, a program and data for realizing the constructed mathematical model are installed in the PC 20.

[0043] Commonly known machine learning algorithms include neural networks, random forests, boosting, and support vector machines (SVMs).

[0044] In this embodiment, a multi-layer neural network is used as the machine learning algorithm. The neural network includes an input layer for inputting data, an output layer for generating data of the analysis result to be predicted, and one or more hidden layers between the input layer and the output layer. A plurality of nodes (also referred to as units) are arranged in each layer. In particular, in this embodiment, a convolutional neural network (CNN), which is a type of multi-layer neural network, is used. Note that other machine learning algorithms may also be used. For example, generative adversarial networks (GAN), which utilize two competing neural networks, may be adopted as the machine learning algorithm.

[0045] In this embodiment, the first mathematical model is pre-trained to output a quality target for an ophthalmic image of a subject's eye for the subject's medical information.

[0046] 3 schematically shows input training data and output training data in the first mathematical model. Medical information for each of multiple (n) subjects (subjects P1, P2, ..., Pn) is used as the input training data. For example, the medical information may include age (Age), refractive power (Ref), axial length (AL), fixation status (Fixation), and the presence or absence of cataracts (Cataract).

[0047] The quality assessment results of the OCT images taken of these subjects are used as the training data for output. One example of the quality assessment results is the SSI of the OCT images, which can be calculated using a predetermined formula based on the pixel information of the OCT images.

[0048] Furthermore, in this embodiment, recommended examination conditions are acquired (S4). In this embodiment, the relationship between the medical information and the recommended examination conditions is acquired and set using a second mathematical model that has been previously trained. The second mathematical model is trained by a machine learning algorithm so that recommended examination conditions are output for the subject's medical information. That is, using the previously trained second mathematical model, examination conditions that have a high correlation with the input medical information are output as recommended examination conditions.

[0049] For example, input training data and output training data in the second mathematical model are shown schematically in Fig. 4. The input training data may be common to the first mathematical model, and a description thereof will be omitted.

[0050] Information indicating the examination conditions when the subjects were imaged is used as output training data for the second mathematical model. For example, information such as the presence or absence of auto-alignment (Auto Alignment), the presence or absence of auto-optimization (Auto Optimization), OCT sensitivity, and the presence or absence of eyelid opening assistance may be used. Note that auto-optimization here includes adjustment of at least one of focus, OPL, and polarization in the OCT optical system.

[0051] The PC 20 displays an observation screen on the monitor 25 (S5). As shown in FIG. 5, the observation screen displays an observation image 120 and an OCT image 130 acquired in real time. On the observation screen, an SQI, which is the quality of the observation image 120, is displayed near the observation image 120. In addition, an SSI, which is the quality of the OCT image, is displayed near the OCT image. The SQI and SSI are updated sequentially based on images acquired at any time. In addition, a quality target and recommended inspection conditions are displayed on the monitor 25 (S6).

[0052] In this embodiment, an SSI target value 135 is displayed as an example of a display mode of the quality target. The SSI target value 135 may be a representative value of the tolerance range, or may be a value with a range that indicates the entire tolerance range. By displaying the quality target simultaneously with the observation image and the real-time OCT image, the examiner can easily determine whether or not the inspection conditions should be adjusted.

[0053] As an example of a display mode of recommended examination conditions, in this embodiment, a predetermined number of recommended examination conditions that have a higher probability of improving the quality evaluation of ophthalmic images are displayed. In this embodiment, three recommended examination conditions are displayed in the window 140. In this embodiment, the second mathematical model outputs the probability for each examination condition. The three recommended examination conditions displayed in the window 140 are the top three with the highest probabilities. The examiner sets the examination conditions of the OCT device 10 to the same conditions as the recommended examination conditions, for example, by manually adjusting the examination conditions via the operation unit 23.

[0054] The PC 20 can accept operation inputs for adjusting the inspection conditions via various GUI widgets on the observation screen (S7). For example, auto-optimization is performed by operating button 223. In addition, an optical path length adjustment unit 224, a focus adjustment unit 225, and a sensitivity adjustment unit 226 are provided as GUI widgets, and the inspection conditions can be manually changed to desired states by operating each GUI widget.

[0055] Next, the PC 20 controls the OCT device 10 to capture an image of the subject's eye (S8). This allows OCT data of the subject to be acquired. At this time, the OCT device 10 captures the image based on preset examination conditions.

[0056] Next, the screen transitions to a confirmation screen (S9). The captured OCT image is displayed on the monitor 25. An operation input for proceeding to recapture may be accepted via the confirmation screen. The PC 20 also performs a predetermined process on the displayed OCT image and calculates the SSI as a quality evaluation result. The calculated SSI is displayed on the monitor 25 together with the target value 135 (S10). This allows the examiner to easily understand whether the quality evaluation result of the displayed OCT image is an appropriate value for the subject. If the quality of the OCT image does not reach a desired level, recapture can be performed (S11: Yes). The confirmation screen may include a recapture button 230 for accepting an instruction to recapture. If an OCT image of appropriate quality is confirmed, the OCT image is saved based on the operation of the save button 231, and the series of imaging operations ends (S11: No → S12).

[0057] Although the present disclosure has been described above based on the embodiments, it should be understood that the present disclosure is not necessarily limited to the above embodiments.

[0058] For example, in the above embodiment, a mathematical model is provided in PC 20, and the quality target of the ophthalmic image and the recommended examination conditions are obtained using the mathematical model, but this is not necessarily limited to this. A mathematical model may be provided in medical information system 30, and PC 20 may obtain the quality target of the ophthalmic image and the recommended examination conditions identified in medical information system 30 based on medical information. [Explanation of symbols]

[0059] 10 OCT device 20 PC 30 Medical Information Systems

Claims

1. an imaging means for capturing an ophthalmologic image by imaging the eye of a subject; a quality target setting means for setting a quality target for the ophthalmologic image based on medical information of a subject; a quality evaluation unit that acquires a quality evaluation result that evaluates the quality of an ophthalmologic image of the subject that has actually been captured; a display control means for displaying both the quality target and the quality evaluation result, or a quality comparison result which is a comparison result of both the quality target and the quality evaluation result; An eye examination system comprising:

2. the display control means displays both the quality target and the quality evaluation result, or the quality comparison result, on the same screen as the ophthalmologic image of the subject's eye showing the photographing result, Furthermore, 2. The eye examination system according to claim 1, further comprising a re-imaging unit that executes re-imaging when an operation input for re-imaging is received from the screen.

3. The quality evaluation means A mathematical model trained by a machine learning algorithm to receive the clinical information of a subject as input data and output the quality target corresponding to the input data, 2. The eye examination system of claim 1, wherein the medical information of a plurality of subjects serves as input training data, and the quality assessment results obtained based on the ophthalmologic images of a plurality of subjects serves as output training data, and the system has a pre-trained mathematical model.

4. an examination condition acquisition means for acquiring examination conditions for the imaging means based on medical information of a subject; an inspection condition setting means for setting the inspection conditions acquired by the inspection condition acquisition means for the imaging means; 4. The eye examination system according to claim 1, further comprising:

5. The inspection condition acquisition means a second mathematical model trained by a machine learning algorithm to receive the medical information of a subject as input data and output the test conditions corresponding to the input data, 5. The eye examination system of claim 4, further comprising a second mathematical model that has been pre-trained, the second mathematical model having the medical information of a plurality of subjects as input training data and the examination conditions when photographing the plurality of subjects as output training data.

6. 1. An eye examination program comprising: When executed by a processor of an eye examination device, the eye examination device: Acquire ophthalmic images by photographing the subject's eyes; setting a quality target that is a quality target for the ophthalmologic image based on medical information of a subject; obtaining a quality evaluation result that evaluates the quality of an ophthalmologic image of the subject that has actually been captured; An eye examination program that displays both the quality target and the quality evaluation result, or a quality comparison result that is a comparison result of both.

7. an imaging means for capturing an ophthalmologic image by imaging the eye of a subject; an examination condition acquisition means for acquiring examination conditions for the imaging means based on medical information of a subject; an examination condition setting means for setting the examination conditions acquired by the examination condition acquisition means for the photographing means.

8. 1. An eye examination program comprising: When executed by a processor of an eye examination device, the eye examination device: acquiring examination conditions for the imaging means based on medical information of the subject; The eye examination system sets the examination conditions acquired by the examination condition acquisition means for the photographing means.

Images