SYSTEM AND METHOD FOR ANALYZING CORNEAL LESION USING ANTERIOR OCULAR SEGMENT IMAGE, AND COMPUTER-READABLE RECORDING MEDIUM

• Main Authors: HAN, Gyu Le, RO, Yong Man, LEE, Hye Bin, KIM, Young Jun, LIM, Dong Hui
• Format: Patent
• Language: English; French; Korean
• Published: 05.08.2021
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DIAGNOSIS; HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA; HUMAN NECESSITIES; HYGIENE; IDENTIFICATION; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS; MEDICAL OR VETERINARY SCIENCE; PHYSICS; SURGERY

The present invention relates to a system and method for analyzing a corneal lesion using an anterior ocular segment image, and more particularly, to a system and method for analyzing the location and cause of a disease in an anterior ocular segment image by machine learning clinical information about a subject on the basis of deep learning. The present invention provides a system for analyzing a corneal lesion using an anterior ocular segment image, and a corneal lesion analysis method using same. The system comprises: an image acquisition unit which acquires an anterior ocular segment image from an eyeball of a subject; a feature extraction unit which extracts feature information about the location and cause of a lesion in the cornea from the anterior ocular segment image by applying a convolution layer with the anterior ocular segment image through machine learning on the basis of a database in which clinical information obtained by analyzing pre-acquired locations and causes of lesions in corneas of subjects is stored; and a result determination unit which checks the location of the cornea in the anterior ocular segment image by using the feature information, and analyzes and determines the location and cause of the lesion in the cornea from the location of the cornea. According to the present invention, there are advantages in that the rate of misdiagnosis is lowered through accurate and rapid diagnosis by a diagnosis model trained on the basis of the clinical information, and a suitable diagnosis in line with the trends at the time of medical treatment can be presented on the basis of an updated database. La présente invention concerne un système et un procédé pour analyser une lésion cornéenne au moyen d'une image de segment oculaire antérieur, et, plus particulièrement, un système et un procédé pour analyser l'emplacement et la cause d'une maladie dans une image de segment oculaire antérieur par apprentissage machine d'informations cliniques concernant un sujet sur la base d'un apprentissage profond. La présente invention concerne un système pour analyser une lésion cornéenne au moyen d'une image de segment oculaire antérieur, et un procédé d'analyse de lésion cornéenne l'utilisant. Le système comprend : une unité d'acquisition d'image qui acquiert une image de segment oculaire antérieur à partir d'un globe oculaire d'un sujet ; une unité d'extraction de caractéristique qui extrait des informations de caractéristique concernant l'emplacement et la cause d'une lésion dans la cornée à partir de l'image de segment oculaire antérieur en appliquant une couche de convolution avec l'image de segment oculaire antérieur par apprentissage machine sur la base d'une base de données dans laquelle sont stockées des informations cliniques obtenues par analyse d'emplacements et de causes pré-acquis de lésions dans des cornées de sujets ; et une unité de détermination de résultat qui vérifie l'emplacement de la cornée dans l'image de segment oculaire antérieur à l'aide des informations de caractéristique, et analyse et détermine l'emplacement et la cause de la lésion dans la cornée à partir de l'emplacement de la cornée. Selon la présente invention, il y a des avantages dans la mesure où le taux de diagnostic erroné est abaissé par l'intermédiaire d'un diagnostic précis et rapide par un modèle de diagnostic entraîné sur la base des informations cliniques, et un diagnostic approprié conforme aux tendances au moment du traitement médical peut être présenté sur la base d'une base de données mise à jour. 본 발명은, 전안부 이미지를 이용한 각막 병변 분석 시스템 및 방법에 관한 것으로서, 특히 피검사자의 임상 정보를 딥러닝 기반으로 기계 학습하여 전안부 영상에 대한 질환 위치와 원인을 분석하는 시스템 및 방법에 관한 것이다. 본 발명은 피검사자의 안구로부터 전안부 이미지를 획득하는 이미지획득부; 기 획득된 피검사자들의 각막의 병변 위치 및 원인을 분석한 임상 정보가 저장된 데이터베이스를 기초로 머신러닝에 의해 상기 전안부 이미지와의 컨벌루션 레이어(Convolution Layer)를 적용하여 상기 전안부 이미지에서 상기 각막의 병변 위치 및 원인에 관한 특징 정보를 추출하는 특징추출부; 및 상기 특징 정보를 이용하여 상기 전안부 이미지에서 상기 각막의 위치를 확인하고 상기 각막의 위치로부터 상기 각막의 병변 위치 및 원인을 분석 및 판단하는 결과판단부를 포함하는 전안부 이미지를 이용한 각막 병변 분석 시스템 및 이를 이용한 각막 병변 분석 방법을 제공한다. 본 발명에 따르면, 임상 정보를 바탕으로 학습된 진단 모델에 의한 정확하고 빠른 진단을 통해 오진율을 낮추고 갱신되는 데이터베이스에 기반하여 진료 당시 추세에 맞는 적절한 진단을 제시할 수 있는 이점이 있다.