Method for visualizing ocular disease lesions based on optical coherence tomography images, apparatus therefor, and system for predicting ocular disease prognosis comprising the same

Abstract

A chronic eye disease monitoring device based on optical coherence tomography images is disclosed. The monitoring device recognizes multiple lesions corresponding to each of multiple eye diseases based on an acquired optical coherence tomography image of the retinal layer, and can display the acquired optical coherence tomography image and the recognized multiple lesions through multiple split screens. By providing the monitoring device, the convenience of analysis and diagnostic accuracy of eye diseases and lesions can be improved.

Description

Technology Field

[0001] The present disclosure relates to computing technology in the medical field. More specifically, the present disclosure relates to a method for monitoring chronic eye disease based on optical coherence tomography (OCT) images, an apparatus for the same, and an apparatus for determining the prognosis of the monitored eye disease and determining factors for determining the prognosis of the eye disease. Background Technology

[0002] The following description merely provides background information related to the present embodiment and does not constitute prior art.

[0003] Optical coherence tomography (OCT) images, which non-invasively capture three-dimensional cross-sectional images of a patient's eye, are used for ophthalmic examinations or ophthalmic surgeries for retinal diseases. In this case, a measurement light (e.g., near-infrared light) is transmitted through the retina being examined, and an OCT image can be generated by detecting the reflected light reflected from within the retina.

[0004] However, monitoring and analyzing patients' eye diseases through OCT images was difficult to perform without expertise.

[0005] Therefore, there is a need for a method to intuitively and effectively monitor eye diseases using acquired OCT images, and furthermore, a method to automatically diagnose eye diseases by frequently monitoring the acquired OCT images. Prior art literature

[0006] Republic of Korea Published Patent Application No. 10-2019-0128292 (Publication Date: Nov. 18, 2019) The problem to be solved

[0007] The problem that the present disclosure aims to solve is to provide a method for monitoring chronic eye diseases based on OCT images and diagnosing them based on an artificial intelligence algorithm.

[0008] In addition, another problem that the present disclosure aims to solve is to provide a method for determining the prognosis of a monitored chronic eye disease and determining the factors (variables) for determining the prognosis of the eye disease.

[0009] The problems that this disclosure aims to solve are not limited to those mentioned above, and other unmentioned problems will be clearly understood by a person skilled in the art from the description below. means of solving the problem

[0010] To solve the aforementioned problem, an optical coherence tomography (OCT) image-based chronic eye disease monitoring device according to the present disclosure comprises: a display; a memory for storing one or more instructions; and one or more processors for executing the instructions stored in the memory.

[0011] The processor recognizes multiple lesions corresponding to each of multiple eye diseases based on an acquired optical coherence tomography (OCT) image of the retinal layer, and can output the acquired optical coherence tomography image and the recognized multiple lesions through the display.

[0012] The processor can display the display by dividing it into a plurality of regions, wherein the display setting screen is displayed in the first region, the target area of ​​the retinal layer is displayed in the second region, one of a plurality of longitudinal images of the target area is displayed in the third region, numerical data of each of a plurality of lesions corresponding to each of the plurality of eye diseases is displayed in the fourth region, and an interface in which one of the plurality of longitudinal images of the target area can be selected is displayed in the fifth region.

[0013] The above processor can output different colors of the target area according to the thickness of the lesion in the second area through the display.

[0014] The processor can output a line matching each of the plurality of end images to the second area through the display, and when one of the plurality of end images is selected in the second area, output the selected end image to the third area.

[0015] The processor can use a deep learning-based model to segment the internal limiting membrane (ILM) located above the target area or the Bruch membrane (BM) located below the target area in each of the plurality of longitudinal images, and segment a plurality of lesions corresponding to each of the plurality of eye diseases.

[0016] The processor can output the divided area by highlighting it through the display.

[0017] The processor can output numerical data corresponding to each of a plurality of lesions corresponding to one of an eye disease among age-related macular degeneration (AMD), diabetic macular edema (DME), and retinal vein occlusion (RVO) to the fourth area through the display.

[0018] Additionally, a method for monitoring chronic eye disease based on optical coherence tomography (OCT) images performed by a process according to the present disclosure comprises: a step of acquiring an optical coherence tomography (OCT) image of a retinal layer; and a step of recognizing a plurality of lesions corresponding to each of a plurality of eye diseases based on the acquired optical coherence tomography (OCT) image of the retinal layer, and displaying the acquired optical coherence tomography image and the recognized plurality of lesions.

[0019] The above-mentioned displaying step may include displaying a display setting screen in a first area, a target area of ​​the retinal layer in a second area, one of a plurality of longitudinal images of the target area in a third area, numerical data of each of a plurality of lesions corresponding to each of the plurality of eye diseases in a fourth area, and an interface in which one of the plurality of longitudinal images of the target area can be selected in a fifth area.

[0020] The step of displaying in the second area above may include the step of displaying the color of the target area differently depending on the thickness of the lesion.

[0021] The step of displaying in the second area includes the step of displaying a line matching each of the plurality of end images in the second area, and the step of displaying in the third area may include the step of displaying the selected end image in the third area when one of the plurality of end images is selected in the second area.

[0022] The above method may further include the step of using a deep learning-based model to segment the internal limiting membrane (ILM) positioned above the target area or the Bruch membrane (BM) positioned below it in each of the plurality of longitudinal images, segmenting a plurality of lesions corresponding to each of the plurality of eye diseases, and highlighting and outputting the segmented areas.

[0023] The step of displaying in the fourth area may include displaying numerical data corresponding to each of a plurality of lesions corresponding to one of an eye disease among age-related macular degeneration (AMD), diabetic macular edema (DME), and retinal vein occlusion (RVO) in the fourth area.

[0024] Additionally, an apparatus for determining the prognosis of an eye disease and determining factors for determining the prognosis of an eye disease according to the present disclosure comprises: a memory; and a processor electrically connected to said memory.

[0025] The processor collects at least one of the following information for each of a plurality of lesions corresponding to an eye disease: patient medical record information, demographic information, the patient's time-series visual acuity information, information on the history of drugs administered to the patient, thickness information of the plurality of lesions, and volume information of the plurality of lesions, and can determine the prognosis of the eye disease based on the collected information.

[0026] The above processor can predict the patient's vision at a preset future point in time.

[0027] The above processor can determine the factors for determining the prognosis of the eye disease using statistical methods.

[0028] In addition to this, a computer program stored on a computer-readable recording medium may be further provided to carry out a method for implementing the present disclosure.

[0029] In addition, a computer-readable recording medium for recording a computer program for executing a method for implementing the present disclosure may be further provided. Effects of the invention

[0030] The method for monitoring and diagnosing chronic eye diseases based on OCT images disclosed in this disclosure is provided in a visual and intuitive manner, thereby improving the convenience of analysis and diagnostic accuracy for eye diseases.

[0031] The effects of the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0032] FIG. 1a is a diagram illustrating the schematic process of an OCT image-based chronic eye disease monitoring device according to the present disclosure, and FIG. 1b is a diagram showing the differences between the OCT technology according to the present disclosure and the prior art. FIG. 2 is a block diagram showing the configuration of an OCT image-based chronic eye disease monitoring device according to the present disclosure. FIG. 3 is a sequence diagram illustrating a method for monitoring chronic eye disease based on OCT images according to the present disclosure. FIG. 4 is a sequence diagram illustrating a screen output method of an OCT image-based chronic eye disease monitoring device according to the present disclosure. FIG. 5 is a diagram showing the screen output configuration of an OCT image-based chronic eye disease monitoring device according to the present disclosure. FIGS. 6(a) and FIGS. 6(b) are drawings for explaining the highlight output function according to the lesion thickness according to the present disclosure. FIG. 7 shows a screen that outputs different colors according to the thickness of the lesion according to the present disclosure. FIGS. 8 and 9 show lesion-related calculated numerical data of an OCT image-based chronic eye disease monitoring device according to the present disclosure. FIG. 10 shows a screen that outputs the lesion thickness according to the present disclosure in a separate area. FIG. 11 shows a screen that outputs a plurality of numerical data for lesions corresponding to each eye disease according to the present disclosure. FIG. 12 shows data representing lesions corresponding to a patient's eye disease in a time series according to the present disclosure. FIGS. 13 to 16 illustrate a method for determining the prognosis of a monitored eye disease according to the present disclosure and deriving factors for determining the prognosis of the eye disease. Specific details for implementing the invention

[0033] Throughout this disclosure, the same reference numerals denote the same components. This disclosure does not describe all elements of the embodiments, and general content in the art to which this disclosure pertains or content that overlaps between embodiments is omitted. The terms 'part, module, component, block' as used in the specification may be implemented in software or hardware, and depending on the embodiments, a plurality of 'parts, modules, components, blocks' may be implemented as a single component, or a single 'part, module, component, block' may include a plurality of components.

[0034] Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are directly connected but also cases where they are indirectly connected, and indirect connections include connections made via a wireless communication network.

[0035] Furthermore, when it is stated that a part "includes" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0036] Throughout the specification, when it is stated that a component is located "on" another component, this includes not only cases where a component is in contact with another component, but also cases where another component exists between the two components.

[0037] The terms first, second, etc. are used to distinguish one component from another, and the components are not limited by the aforementioned terms.

[0038] Singular expressions include plural expressions unless there is an obvious exception in the context.

[0039] In each step, identification codes are used for convenience of explanation and do not describe the order of the steps; the steps may be performed differently from the specified order unless a specific order is clearly indicated in the context.

[0040] The operating principles and embodiments of the present disclosure will be described below with reference to the attached drawings.

[0041] In this specification, the 'OCT image-based chronic eye disease monitoring device and device for determining the prognosis of an eye disease and determining factors for determining the prognosis of an eye disease according to the present disclosure' may be implemented in various devices capable of performing computational processing and providing results to a user, as well as in cloud and cluster systems. For example, the OCT image-based chronic eye disease monitoring device according to the present disclosure may include a computer, a server device, and a portable terminal, or may take the form of any one of these.

[0042] Here, the computer may include, for example, a notebook, desktop, laptop, tablet PC, slate PC, etc. equipped with a web browser.

[0043] The above server device is a server that processes information by communicating with an external device, and may include an application server, a computing server, a database server, a file server, a game server, a mail server, a proxy server, and a web server.

[0044] The above portable terminal may include, for example, all types of handheld-based wireless communication devices such as PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminals, smartphones, etc., as well as wearable devices such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted devices (HMDs).

[0045] FIG. 1a is a diagram illustrating the schematic process (S1 to S3) of an OCT image-based chronic eye disease monitoring device (100 in FIG. 2, hereinafter referred to as the 'monitoring device') according to the present disclosure, and FIG. 1b is a diagram showing the differences between the OCT technology according to the present disclosure and the prior art.

[0046] First, the monitoring device (100) can acquire an OCT image of the retinal layer (S1).

[0047] The retinal layer OCT image may be an OCT image that includes the internal limiting membrane (ILM) of the retinal layer and the Bruch membrane (BM) located underneath, and the monitoring device (100) may acquire multiple OCT images in the form of image files. The OCT image may be used to analyze the structure of biological tissue and may be an image generated by measuring the reflection and absorption of the tissue using optical wavelengths.

[0048] Next, the monitoring device (100) can divide the lesion corresponding to the eye disease (S2).

[0049] Here, the eye disease may be one of age-related macular degeneration (AMD), diabetic macular edema (DME), and retinal vein occlusion (RVO), but the examples are not limited thereto.

[0050] After that, the monitoring device (100) can perform visualization of the segmented lesion (S3).

[0051] The monitoring device (100) can visualize and display existing segmented lesions, unlike the existing OCT technology in [Table 1] below, for the convenience of analysis and diagnosis of a user (e.g., medical staff, administrator, etc.) who performs diagnosis and analysis of lesions corresponding to eye diseases (differentiation from existing OCT technology).

[0052] [Table 1]

[0053]

[0054] A monitoring device (100) applying OCT technology according to the present disclosure is designed to enhance convenience and efficiency by enabling diagnostic assistance and analysis, as it allows both patient information and image information to be checked in a single device, thereby enabling the verification of information that would otherwise need to be checked separately from existing patient information and OCT devices in relation to analysis and diagnostic convenience.

[0055] Referring to FIG. 1b below, a monitoring device (100) to which OCT technology according to the present disclosure is applied may use and include both patient personal information and OCT images for user diagnosis and analysis convenience, and the patient personal information may include EMR (electronic medical record), images may be extracted from OCT images, and the extracted images may be used for diagnosis and analysis (differentiation from existing OCT device medical systems).

[0056] FIG. 2 is a block diagram showing the configuration of a monitoring device (100) according to the present disclosure. The components of the monitoring device (100) shown in FIG. 2 are not essential for implementing the monitoring device (100) according to the present disclosure, so the monitoring device (100) described in this specification may have more or fewer components than the components listed above.

[0057] Among the above components, the communication unit (110) may include one or more components that enable communication with various devices equipped with a communication module, for example, may include at least one of a broadcast reception module, a wired communication module, a wireless communication module, a short-range communication module, and a location information module. The short-range communication module may include a module for recognizing the approach of an external device (e.g., a UWB (Ultra-Wideband) communication module).

[0058] The input unit (120) is for inputting video information (or signal), audio information (or signal), data, or information input from a user, and may include at least one of at least one camera, a touch input unit provided on a touch screen, at least one microphone, and a user input interface. Voice data or image data collected by the input unit (120) may be analyzed and processed into a user control command.

[0059] The display (130) displays (outputs) information processed by the device (100). For example, the display (130) can display information on the execution screen of an application (e.g., an application) running on the device (100), UI (User Interface) and GUI (Graphic User Interface) information based on such execution screen information.

[0060] The memory (150) can store data supporting various functions of the monitoring device (100) and programs for the operation of the processor, and can store input / output data (e.g., music files, still images, videos, etc.), and can store a number of application programs (or applications) running on the monitoring device (100), data for the operation of the device, and instructions. At least some of these application programs can be downloaded from an external server via wireless communication.

[0061] The memory (150) can store one or more instructions and can store a deep learning-based neural network model. The memory (150) may include a storage medium corresponding to at least one type among flash memory type, hard disk type, SSD type (Solid State Disk type), SSD type (Silicon Disk Drive type), multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, and optical disk. Additionally, the memory (150) may be a database that is separated from the monitoring device (100) but connected via wired or wireless means, and may be implemented as a database system.

[0062] The processor (190) may include one or more processors and may include at least one core. The processor (190) may execute instructions stored in memory (150). The processor (190) may be implemented as memory that stores data for an algorithm or a program that reproduces the algorithm for controlling the operation of components within the monitoring device (100), and at least one processor (not shown) that performs the aforementioned operation using the data stored in memory. In this case, the memory and the processor may each be implemented as separate chips. Alternatively, the memory and the processor may be implemented as a single chip.

[0063] At least one component may be added or removed in response to the performance of the components shown in FIG. 2. Additionally, it will be readily understood by those skilled in the art that the relative positions of the components may be changed in response to the performance or structure of the system.

[0064] Meanwhile, each component illustrated in Figure 2 refers to a software and / or hardware component such as a Field Programmable Gate Array (FPGA) and an Application Specific Integrated Circuit (ASIC).

[0065] Additionally, the device (500) for determining the prognosis of an eye disease and determining the factors for determining the prognosis of an eye disease, which will be described later, may be the same device as the monitoring device (100) described above, or it may be implemented as a separate device, but its configuration may be similar to that of the monitoring device (100).

[0066] FIG. 3 is a sequence diagram illustrating a method for monitoring chronic eye diseases based on OCT images according to the present disclosure. FIG. 4 is a sequence diagram illustrating a method for displaying a screen of a monitoring device (100) according to the present disclosure. FIG. 5 is a diagram illustrating the configuration of the screen output of a monitoring device (100) according to the present disclosure. FIG. 6(a) and FIG. 6(b) are diagrams illustrating a highlight output function according to lesion thickness according to the present disclosure. FIG. 7 shows a screen that outputs different colors according to lesion thickness according to the present disclosure. FIG. 8 and FIG. 9 show numerical data related to lesions calculated by the monitoring device (100) according to the present disclosure. FIG. 10 shows a screen that outputs lesion thickness in a separate area according to the present disclosure. FIG. 11 shows a screen that outputs multiple numerical data for lesions corresponding to each eye disease according to the present disclosure. While describing FIG. 3, FIG. 4 to FIG. 11 will be referenced together where necessary.

[0067] In step S310, the processor (190) can acquire an OCT image of the retinal layer.

[0068] The processor (190) can acquire an OCT (Optical Coherence Tomography) image of the retinal layer through the communication unit (110), the input unit (120), etc. The acquired OCT image may include a plurality of OCT B-scan images. The plurality of OCT B-scan images may be OCT end images, which are images acquired by sending light to a sample through an optical fiber and measuring the reflection intensity at each position by measuring the time it takes for the light to be reflected back.

[0069] In step S320, the processor (190) can recognize multiple lesions corresponding to each of the multiple eye diseases based on the acquired OCT image of the retinal layer.

[0070] Here, the eye disease may include one of age-related macular degeneration (AMD), diabetic macular edema (DME), and retinal vein occlusion (RVO), but the examples are not limited thereto.

[0071] Age-related macular degeneration (AMD) is a disease in which the macula in the center of the retina is damaged due to aging and is a major cause of vision loss; diabetic macular edema (DME) is a disease in which edema occurs in the macula due to damage to the blood vessels of the retina caused by diabetes; and retinal vein occlusion (RVO) is an eye disease in which blood cannot flow to the retina as the retinal veins become blocked.

[0072] Multiple lesions corresponding to eye diseases may include IRF (intraretinal fluid), SRF (subretinal fluid), PED (pigment epithelial detachment), SHRM (subretinal hyperreflective material), etc., but the examples are not limited thereto.

[0073] Here, IRF can be used as a term representing intraretinal sclerotic material, which refers to changes caused by diseases associated with abnormal vascular growth or the leakage of abnormal blood and fluids in the retina. This can occur as abnormal blood vessels formed beneath the retinal neuroepithelial layer, corresponding to the posterior part of the retina, or the resulting abnormal fluid, leak into the retina. SRF represents intraretinal servopathic fluid, which is a condition resulting from fluid leakage into the servopathic layer of the retina. The servopathic layer is located in the outer part of the retina, and fluid leakage can cause problems with the structure and function of the retina. PED is a condition in which the retinal pigment epithelium separates from the upper part of the retina and can primarily occur in the posterior part of the retina. SHRM is subretinal hyperreflective material and can be identified in retinal eye diseases.

[0074] In step S330, the processor (190) can display the acquired OCT image and the recognized plurality of lesions, and the processor (190) can divide the display (130) into a plurality of divided screens (e.g., first to fifth regions) and provide them.

[0075] In an embodiment, the processor (190) can match possible lesions in each eye disease, and when each eye disease is selected, it can output information about the possible lesions in the selected eye disease through the display (130).

[0076] Referring to FIGS. 4 and 5, the processor (190) can control the display (130) to display a display setting screen in a first area (A1 in FIG. 5), a target area of ​​the retinal layer in a second area (A2), one of a plurality of longitudinal images of the target area in a third area (A3), numerical data of each of a plurality of lesions corresponding to each of a plurality of eye diseases in a fourth area (A4), and an interface in which one of a plurality of longitudinal images of the target area can be selected in a fifth area (A5).

[0077] The processor (190) can output a lesion corresponding to each of the multiple eye diseases and an ETDRS Ring corresponding to the lesion together in the fourth region (A4) (A4a~A4F). Here, the ETDRS Ring points to a specific point located at a certain distance from the center of the retina, and this point can be used to evaluate and track retinal diseases, and can be used to evaluate visual acuity and fundus vascular conditions, etc.

[0078] Additionally, the processor (190) may also place an ETDRS Ring in a second area (A2) that displays a target area of ​​the retinal layer. In an embodiment, the processor (190) may display an ETDRS Ring (B1b) in the second area (A2) when the corresponding item (B1a) of the first area (A1) is selected. At this time, when the processor (190) displays the ETDRS Ring in the second area (A2), it may display it in the second area (A2) by reflecting the thickness and volume of the actual lesion.

[0079] Referring to FIG. 4, in step S331, the processor (190) can display a setting screen in the first area (A1).

[0080] The processor (190) can output items corresponding to various setting functions to the first area (A1). For example, the processor (190) can display a button for loading an OCT B-scan image, a button for displaying the top and bottom of an OCT B-scan image, a button for displaying the left and right range of an OCT B-scan image, a button for displaying grid lines on an OCT image, a button for turning the color map of an OCT B-scan image on / off, a button for adjusting the transparency of the OCT B-scan image color map, a button for turning the visualization image of the division thickness of an OCT B-scan image on / off, a menu related to ETDRS Ring, a color map on / off button, etc., but the embodiment is not limited to the example described above.

[0081] In step S333, the processor (190) can display the target area of ​​the retinal layer in the second area.

[0082] Referring to FIGS. 5, FIGS. 6(a) and FIGS. 6(b), the processor (190) can output a target area of ​​the retinal layer to a second region (A2), and can output the color of the target area in a contour line form differently depending on the thickness of the lesion (e.g., B2a, B2b, etc., outputting different colors depending on the thickness). Here, the target area can be said to be the entire target area of ​​the retinal layer for analyzing and diagnosing eye diseases.

[0083] The processor (190) can display the thickness in color only when a menu (B4a) that sets a different color according to the thickness in the second area (A2) is selected among the menus (buttons) placed in the first area (A1). Referring to FIG. 6(a), the menu is activated (ON), and FIG. 6(b) shows the menu being deactivated (OFF).

[0084] Referring to FIG. 7, when the corresponding menu is activated, the processor (190) may display in red for thicknesses greater than 200 mm, in green for thicknesses less than 100 mm, and in the indicated colors for thicknesses in between. However, the thickness range and color display may vary depending on the embodiment. Here, the color may vary depending on the embodiment.

[0085] The processor (190) outputs lines (multiple green horizontal lines) that match each of the multiple end images to the second area (A2), and when one of the multiple end images is selected in the second area (A2), the selected end image can be output to the third area (A3). Additionally, the processor (190) can output a line corresponding to the selected end image by highlighting it in the second area (A2), and can also output the corresponding end image by highlighting it in the fifth area (A5).

[0086] In an embodiment, the processor (190) can be synchronized with each other when a line is selected in the second area (A2) or a specific end image is selected in the fifth area (A5).

[0087] For example, when a specific line (horizontal line) is selected in the second area (A2), the processor (190) can also display the corresponding end image in the third area (A3) and the fifth area (A5). Additionally, when a specific end image is selected in the fifth area (A5), the processor (190) can select the specific line (horizontal line) in the second area (A2) (e.g., highlight it) and output the corresponding end image in the third area (A3).

[0088] In step S335, the processor (190) can display one of a plurality of end images of the target area in the third area (A3). That is, the processor (190) can map one end image to one horizontal green line displayed in the second area (A2) and map one end image displayed in the fifth area (A5) to display an OCT end image in the third area (A3).

[0089] The processor (190) can use a deep learning-based model to divide the internal limiting membrane (ILM) located above the target area in each of the plurality of longitudinal images to the bruch membrane (BM) located below the target area, and divide the plurality of lesions corresponding to each of the plurality of eye diseases.

[0090] The processor (190) can divide each retinal layer between the inner limiting membrane and Bruch's membrane of the target area based on various CNN-based algorithms based on labels corresponding to each layer of the retinal layer. Additionally, the processor (190) can divide each retinal layer between the inner limiting membrane and Bruch's membrane of the target area based on various CNN-based algorithms based on labels corresponding to lesions (IRF, SRF, PED, SHRM, etc.) corresponding to each eye disease. During the learning phase, the processor (190) can divide the retinal side and the region corresponding to the lesion using each label until a predetermined accuracy is secured (e.g., the value of the loss function satisfies a preset value).

[0091] In an embodiment, the processor (190) may display the boundary of the inner boundary film to the Bruck film of the target area in a third region (A3). However, it may be displayed by selecting an item placed in the first region (A1), but the present disclosure is not limited thereto.

[0092] In an embodiment, the processor (190) receives a prior input of possible lesions for each eye disease, and when a specific eye disease is selected, it can display possible lesions corresponding to the selected eye disease. Accordingly, rather than the diagnostician checking multiple lesion screens, if only a specific eye disease is selected in advance, images, analysis information, etc. regarding the related lesions can be displayed.

[0093] The processor (190) can output a divided area through the display (130) by highlighting it.

[0094] In step S337, the processor (190) can output numerical data corresponding to each of a plurality of lesions corresponding to one of the eye diseases of age-related macular degeneration (AMD), diabetic macular edema (DME), and retinal vein occlusion (RVO) to the fourth region (A4).

[0095] With reference to FIGS. 5, FIGS. 8 and FIGS. 9 together, the processor (190) can display the ETDRS Ring value for the corresponding lesion in the retinal layer and each lesion.

[0096] The processor (190) can calculate the number of pixels (CNT) corresponding to each of the center zone, upper, left, lower, and right parts of the divided retinal layer (e.g., ILM~BM) and the total sum of thickness values ​​(SUM) at each pixel location, and when the total sum of thickness values ​​(SUM) is divided by the number of pixels (CNT), the average thickness value (AVG) can be calculated.

[0097] Additionally, the processor (190) can output the volume (VOL) of each lesion using the thickness values ​​of the pixels included in the lesion. Referring to FIG. 9, the processor (190) has a basic volume calculation method (910) that calculates an average volume value by summing the thickness values ​​of the pixel locations included in the center zone, upper, left, lower, and right parts of the ETDRS Ring and dividing the result by the number of pixels in the corresponding area.

[0098] In addition, there is an optimal volume calculation method (920) that calculates the average volume value by dividing the sum of the thickness values ​​of the pixel locations by the number of pixels excluding the number of pixels for which the thickness value is invalid (value is 0).

[0099] In the embodiment, when the processor (190) calculates the average volume value of each lesion, if the optimal volume calculation method (920) is used, the average volume value may increase (156%, 167%, etc.) compared to the basic volume calculation method (910), and this has the effect of highlighting each lesion.

[0100] The processor (190) can visually display the lesion by placing the average thickness value, minimum thickness value, maximum thickness value, center thickness value, thickness value at a specific pixel location, volume value, etc., corresponding to each lesion (e.g., IRF, PED, etc.) in the ETDRS Ring.

[0101] For example, the processor (190) can display a calculated value corresponding to an IRF lesion, and can calculate the average value (122.99~) and volume value (0.09~) of the center zone for the IRF lesion, and performance improvement can be expected compared to the value obtained by dividing the total area of ​​the existing ETDRS Ring.

[0102] The processor (190) can display output values ​​corresponding to PED lesions, and can calculate and display average values ​​and volume values ​​for the center zone, upper, left, lower, and right parts.

[0103] In an embodiment, if a lesion is not recognized, the processor (190) may set the output value corresponding to the lesion to 0 (NULL) and then display it (e.g., SRF, SHRM, etc.).

[0104] In the embodiments, the processor (190) may display only the value corresponding to the lesion corresponding to each eye disease, but the present disclosure is not limited thereto.

[0105] In an embodiment, the processor (190) can output the thickness / area of ​​each eye disease lesion to the display (130) in chronological order, and can provide the diagnostician (e.g., medical staff) with the changes in the patient's condition due to treatment (e.g., injection treatment) in chronological order.

[0106] In step S339, the processor (190) can display an interface in the fifth region (A5) where one of a plurality of end images of the target area can be selected.

[0107] Table 2 below compares the equivalence between a Heidelberg device, which is one of the conventional OCT imaging devices, and a Hyper Vision OCT with the OCT technology according to the present disclosure. Referring to Table 2, the Hyper Vision OCT with the OCT technology according to the present disclosure can be seen to have a range of equivalence to the conventional OCT technology.

[0108] [Table 2]

[0109]

[0110] Referring to FIG. 10, the processor (190) can display retinal thickness in a third area (A3) (Fig. 10(a)). The processor (190) can display retinal thickness (B6) if a menu (button) that activates the corresponding function is selected in the first area (A1). If the processor (190) does not display retinal thickness if a menu that activates the corresponding function is not selected in the first area (A1) (Fig. 10(b)).

[0111] Additionally, the processor (190) can display the boundary of the inner boundary membrane (ILM) to the Bruck membrane (BM) in the first region (A1) (when an item is selected in the first region). Additionally, if the processor (190) wants to display only a specific layer (between the inner boundary membrane and the Bruck membrane), it can display only that specific layer.

[0112] Referring to FIG. 11, the processor (190) can display lesion-related data such as IRF, SRF, PED, SHRM corresponding to each disease through the ETDRS Ring. However, the relevant data can be provided only when the activation options for IRF, SRF, PED, SHRM, etc. are 'On'.

[0113] In an embodiment, the processor (190) can enable / disable the lesion activation option through the On / Off menu of the fourth area (A4), and when set to enabled, the activated lesion can be visually segmented and displayed when displaying the OCT longitudinal image in the third area (A3) (e.g., B3a in FIG. 5).

[0114] When the activation On / Off button in the fourth area (A4) is selected for multiple lesions, the processor (190) can divide and display the areas corresponding to the lesions in different colors in the third area (A3).

[0115] In an embodiment, when a lesion corresponding to each eye disease is recognized, the processor (190) may display it on the corresponding OCT longitudinal image in the fifth area (A5) so that the user can analyze the image. Accordingly, the user's convenience of analysis may be improved.

[0116] As described above, the monitoring device (100) can clearly provide the diagnostician with the thickness / volume of the lesion corresponding to each eye disease, thereby providing convenience for diagnosis and analysis.

[0117] FIG. 12 shows data representing lesions corresponding to a patient's eye disease in a time series according to the present disclosure.

[0118] A device (500, hereinafter referred to as the 'device for determining factors for determining eye disease prognosis') according to the present disclosure for determining the prognosis of an eye disease and determining factors for determining the prognosis of an eye disease can output the thickness of a lesion for an eye disease to a display (530) through a plurality of diagnoses.

[0119] The ocular disease prognosis determination factor determination device (500) can first display the thickness of the internal limiting membrane (ILM) to Bruch's membrane (BM) of the retinal layer and the thickness of each lesion (IRF, SRF, PED, SHRM) over time. Accordingly, the diagnostician can check the data in FIG. 12, verify the patient's condition according to diagnosis / treatment, and use it for future medical / treatment.

[0120] FIGS. 13 to 16 illustrate a method for determining the prognosis of a monitored eye disease according to the present disclosure and for deriving factors for determining the prognosis of the eye disease.

[0121] As illustrated in Fig. 13, ophthalmology is one of the medical specialties where continuous visits are necessary to make decisions regarding treatment methods (e.g., medication, injections, follow-up, etc.) depending on the disease, so it is necessary to assist in confirming this effectively in a short period of time. However, as it is a medical specialty where radiology specialists do not provide interpretation support, this device, which enables ophthalmologists to unify diagnosis, interpretation, and decision-making, can support effective prognosis assessment in clinical situations.

[0122] Precision medicine can be realized as doctors can determine personalized treatment by applying differential treatment cycles and types of medications for each patient, or by verifying the effectiveness of treatment more precisely. In addition, it can help ensure the reliability of medical care by allowing patients to check for overtreatment.

[0123] The ophthalmic disease prognosis determination factor determination device (500) may collect and utilize multivariable structured data such as patient medical record information including information on lesions related to the patient's ophthalmic disease (e.g., DME, RVO, AMD, etc.), demographic information, time-series visual acuity information (one or more examinations), and injection information (one or more medication administrations), but the collected information is not limited to the above. LogMAR (logarithm of the minimum angle of resolution) may be used for visual acuity information, but the embodiments are not limited thereto.

[0124] At this time, the ophthalmic disease prognosis determination factor determination device (500) can collect patient information by dividing it into the initial examination time (1 time) before administering the drug (e.g., injection), 3 months (2 times), 6 months (3 times), and 12 months (4 times) after the initial examination (4 times or more are possible depending on the embodiment), and when data is divided by period, if an excessive number of missing values ​​occurs, the data may not be used.

[0125] Lesion information may include the estimated thickness, volume, etc. of the lesion corresponding to each eye disease, and may include information such as the IRF, SRF, PED, SHRM, and ILM-BM thickness of the ETDRS Ring. Demographic information may include age and gender (removed if there are many missing values), and binocular visual acuity is assumed to be unrelated to changes in visual acuity. Patient visual acuity information may include visual acuity information at the time of diagnosis, visual acuity information after several months, differences in visual acuity over several months, deviation (variance), and differences between good and poor visual acuity states. Information on medication history administered to the patient via oral administration or injection includes injection information, and injection information includes injection administration information according to the lesion, which may include the number of Eylea injections (e.g., 3 times for AMD, every 3 months for DME and RVO), the number of 3 times Lucentis injections, and the number of 3 times Avastin injections.

[0126] The ophthalmic disease prognosis determination factor determination device (500) can collect information such as special diagnosis and treatment information for the 3rd time (in the case of AMD, 3 months in the case of DME and RVO), whether OCT is present in the 3rd time (in the case of AMD, 3 months in the case of DME and RVO), the total number of injections in the 6th month, the difference between the total number of injections in the 6th month and the total number of injections in the 3rd month, the total number of injections in the 12th month, and the difference between the total number of injections in the 12th month and the total number of injections in the 6th month.

[0127] The ophthalmic disease prognosis determination factor determination device (500) can perform various predictions by collecting various data in addition to the data described above.

[0128] The ophthalmic disease prognosis determination factor determination device (500) can perform quarterly vision prediction of a patient, vision improvement (whether vision has worsened, whether vision is good or bad, etc.) using collected multivariable structured data, and can derive key factors (variables) for prediction.

[0129] Referring to FIG. 14, the ophthalmic disease prognosis judgment factor determination device (500) can perform exploratory data analysis. As a result of the analysis, the change in visual acuity after N (3, 6, 12) months is proportional to the initial visual acuity information, and the data can confirm that there is a tendency for the change in visual acuity after N months to be greater as the initial visual acuity is higher.

[0130] The box plot on the left shows the age corresponding to the change in visual acuity after N months (e.g., 12 months). Through this, it is observed that when initial visual acuity is good and visual acuity improves after 3 months, the age group tends to be higher, and when initial visual acuity is good and visual acuity improves after 3 months, it can be observed that the age group is significantly higher after 12 months, just as it was after 3 months.

[0131] In the box plot on the right, initial visual acuity information corresponding to changes in visual acuity after N months (e.g., 12 months) is displayed, and when visual acuity improves, a tendency for the initial visual acuity to be low is observed, when visual acuity deteriorates, a tendency for the initial visual acuity to be high is observed, and when the initial visual acuity is low, a tendency for improvement or maintenance can be observed.

[0132] Referring to FIG. 15, the ocular disease prognosis judgment factor determination device (500) can derive key variables for determining the ocular disease prognosis based on a decision tree.

[0133] The ophthalmic disease prognosis determination factor determination device (500) can generate a decision tree, can repeatedly divide data using features, select the best feature in each division to divide the data into subsets, and repeat this process to construct a tree.

[0134] The eye disease prognosis determination factor determination device (500) can generate all conditions using features, calculate the average impurity for each condition, and determine the condition with the lowest average impurity as the good condition. At this time, the eye disease prognosis determination factor determination device (500) can repeat the process until the impurity no longer decreases.

[0135] In addition, the ocular disease prognosis determination factor determining device (500) can perform classification using 10 subsets (folds), through which it can be confirmed that in the case of a specific lesion (DME), age, initial visual acuity information, and change in visual acuity after 3 months are important variables in predicting whether visual acuity will improve after 6 months from the point in time 3 months prior. Also, in the case of AMD, the ocular disease prognosis determination factor determining device (500) can be confirmed that change in visual acuity after 6 months, injection information at 12 months prior, and initial visual acuity information are important variables in predicting whether visual acuity will improve after 12 months from the point in time 12 months prior.

[0136] Referring to FIG. 16, the ocular disease prognosis judgment factor determination device (500) can identify important factors for judging the prognosis of ocular disease through OLS (ordinary least squares) regression analysis.

[0137] The ophthalmic disease prognosis determination factor determination device (500) can calculate the significance of the model, the significance of the variable, and the coefficient of determination, etc., by targeting the difference in visual acuity around N months. As a result of the model significance test, the p-value may be calculated to be less than 0.05, and as a result of the variable significance test, all variables are significant, but the p-value may be calculated to be less than 0.05 for the change in visual acuity after 6 months, the difference in visual acuity after 6 months (good vs. bad), and visual acuity information after 3 months. In addition, it can be observed that the model explains 64% of the total data variance with a coefficient of determination of 0.64. The ophthalmic disease prognosis determination factor determination device (500, controlled by a processor) collects at least one of the following information for each of the multiple lesions corresponding to the ophthalmic disease: patient medical record information, demographic information, the patient's time-series visual acuity information, information on the history of drugs administered to the patient, thickness information of the multiple lesions, and volume information of the multiple lesions, and can determine the prognosis of the ophthalmic disease based on the collected information.

[0138] In addition, the ophthalmic disease prognosis judgment factor determination device (500) can predict the patient's vision at a preset future point in time and can determine various key factors for judging the ophthalmic disease prognosis based on statistics and machine learning.

[0139] Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium that stores instructions executable by a computer. The instructions may be stored in the form of program code and, when executed by a processor, may generate a program module to perform the operation of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

[0140] Computer-readable recording media include all types of recording media that store instructions that can be decoded by a computer. Examples include ROM (Read Only Memory), RAM (Random Access Memory), magnetic tape, magnetic disk, flash memory, optical data storage devices, etc.

[0141] As described above, the disclosed embodiments have been explained with reference to the attached drawings. Those skilled in the art will understand that the present disclosure may be practiced in forms different from the disclosed embodiments without changing the technical spirit or essential features of the present disclosure. The disclosed embodiments are illustrative and should not be interpreted restrictively. Explanation of the symbols

[0142] 100: Monitoring device, 190: Processor.

Claims

Claim 1 A device for visualizing chronic eye disease lesions based on optical coherence tomography (OCT) images, comprising: a display; a memory for storing one or more instructions; and one or more processors for executing the instructions stored in the memory, wherein the processor recognizes a plurality of lesions corresponding to each of a plurality of eye diseases based on an optical coherence tomography (OCT) image of an acquired retinal layer, divides the display into a plurality of regions to output the acquired optical coherence tomography image and the recognized plurality of lesions through the display, wherein the processor is configured to display an ETDRS Ring for each of the retinal layer and the plurality of lesions in at least some of the plurality of regions, calculates an average value corresponding to each region of the ETDRS Ring, places the calculated value on the ETDRS Ring and displays it, wherein if there is a lesion among the plurality of lesions that is not recognized, the calculated value corresponding to the lesion is displayed as a preset value indicating the undetected state of the lesion. Claim 2 A device according to claim 1, wherein the plurality of lesions comprises IRF (intraretinal fluid), SRF (subretinal fluid), PED (pigment epithelial detachment), and SHRM (subretinal hyperreflective material). Claim 3 In paragraph 2, the device is configured such that the processor performs a prediction related to the prognosis of each of the plurality of eye diseases based on lesion information including a calculated value corresponding to each of the plurality of lesions. Claim 4 In paragraph 3, the device is configured such that the processor calculates the average value corresponding to each region based on the number and thickness of pixels for each of the center zone, upper, left, lower, and right parts of the ETDRS Ring. Claim 5 In paragraph 4, the device is configured such that the processor performs at least one of quarterly vision prediction of the patient, prediction of whether vision will improve, and prediction of the patient's vision at a preset future point in time. Claim 6 In claim 5, the device is configured such that the processor collects at least one of patient medical record information, demographic information, patient time-series visual acuity information and drug history information administered to the patient, and derives key variables for prognosis determination based on at least one of a decision tree and ordinary least squares (OLS) regression analysis. Claim 7 A device according to claim 1, wherein the processor divides the display into a plurality of regions and displays the display, wherein the display setting screen is displayed in the first region, the target area of ​​the retinal layer is displayed in the second region, one of a plurality of longitudinal images of the target area is displayed in the third region, numerical data of each of a plurality of lesions corresponding to each of the plurality of eye diseases is displayed in the fourth region, and an interface in which one of the plurality of longitudinal images of the target area is selectable is displayed in the fifth region. Claim 8 The device according to claim 1, wherein the processor is configured to use a deep learning-based model to divide the internal limiting membrane (ILM) located above the target area in each of the plurality of longitudinal images to the Bruch membrane (BM) located below the target area, divide the plurality of lesions corresponding to each of the plurality of eye diseases, and output the divided areas by highlighting them through the display. Claim 9 A method for visualizing chronic eye disease lesions based on optical coherence tomography (OCT) images performed by a processor of a device, comprising: a step of recognizing a plurality of lesions corresponding to each of a plurality of eye diseases based on an optical coherence tomography (OCT) image of an acquired retinal layer; and a step of dividing a display into a plurality of regions and outputting the acquired optical coherence tomography image and the recognized plurality of lesions through the display; wherein the outputting step comprises: a step of displaying an ETDRS Ring for each of the retinal layer and the plurality of lesions in at least some of the plurality of regions; a step of calculating an average value corresponding to each region of the ETDRS Ring; and a step of placing and displaying the calculated value on the ETDRS Ring, wherein if there is a lesion among the plurality of lesions that is not recognized, a step of displaying the calculated value corresponding to the lesion as a preset value indicating the undetected state of the lesion. Claim 10 An eye disease prognosis prediction system comprising: a device for acquiring optical coherence tomography (OCT) images of a retinal layer; and a device for performing visualization of lesions corresponding to eye diseases based on the optical coherence tomography (OCT) images; wherein the device for performing visualization recognizes a plurality of lesions corresponding to each of a plurality of eye diseases based on the acquired optical coherence tomography (OCT) images of a retinal layer, divides a display into a plurality of regions to output the acquired optical coherence tomography images and the recognized plurality of lesions through the display, and is configured to display an ETDRS Ring for each of the retinal layer and the plurality of lesions in at least some of the plurality of regions, calculates an average value corresponding to each region of the ETDRS Ring, places the calculated value on the ETDRS Ring and displays it, and if there is a lesion among the plurality of lesions that is not recognized, displays the calculated value corresponding to the lesion as a preset value indicating the undetected state of the lesion, and is configured to perform a prediction related to the course or prognosis of each of the plurality of eye diseases based on lesion information including the calculated value corresponding to each of the plurality of lesions.

Images