미만성 간질성 폐질환 진단을 위한 임상 의사결정지원시스템과 컴퓨터 시각모듈과의 통합
이윤희*1), 채영문1), 김종효2), 한대희3), 전성완4), 한헌5)
2)서울대학교 의과대학 의공학교실
Abstract : In order to manage large medical data and improve work process quality, hospitals have increasingly introduced the Picture Archive and Communication System (PACS) and Electronic Medical Records (EMR). As a result, the Clinical Decision Support System (CDSS) is considered to be an essential medical knowledge management system that helps clinicians make better and effective decisions for diagnosis. The purpose of this study was to study computer vision module for automatic Ground Gross Opacity (GGO) detection and Honeycombing, to develop artificial intelligence-based CDSS integrated with hospital information systems for diagnosis of diffuse interstitial lung disease (DILD), and to validate CDSS. In order to diagnose DILD using HRCT for the rule-based CDSS the system was developed based on 12 diseases, 254 HRCT findings and 295 rules (237 in "With" and 58 in "Not"). The computer visual module for automatic GGO and Honeycombing detection from the HRCT image data was developed by Texture Features and Geometric Features. Also, Datawarehouse was developed to integrate the CDSS with Hospital Information Systems to collect the patient information and diagnosis knowledge for diagnosis. The results of validation examinations showed that the CDSS also featured interfaces managing the rules, finding list, and disease list. The score of the examiners were higher ( p = 0.0078) when they used the CDSS (167) than they didn't (110). keyword : Clinical decision support system(CDSS), Artificial intelligence(AI), Radiologic, Diffuse interstitial lung disease(DILD), Computer vision, Hospital Information Systems Integration
투고내용을 삭제하시려면 "삭제"버튼을 누르시고 다시 접수하여 주시기 바랍니다.
Copyright (c) 2002 by The Korean Society of Medical Informatics