AI-Powered De-identification System for Electronic Medical Records Developed in South Korea
В Южной Корее появился ИИ для обезличивания медзаписей - ВШОУЗ
Medical center researchers, in collaboration with the Ministry of Health and Welfare, developed an AI system for de-identifying medical records in South Korea.
Summary
Researchers at Asan Medical Center in Seoul developed an AI system that automatically de-identifies electronic medical records. This system can remove personal patient information (names, dates of birth, etc.) while preserving clinically vital data such as diagnoses and surgical plans.
Key Players
Details
Researchers from Asan Medical Center in Seoul have developed a de-identification system for electronic medical records using artificial intelligence (AI). The system automatically removes personal identifying information (such as names, dates of birth, addresses) while retaining clinically important details like diagnoses, prescriptions, and operation plans. It demonstrated high reliability, reportedly accurately masking personal data in over 95 out of 100 cases. A key feature is that the solution does not require cloud services or external connections, allowing it to be installed locally on standard hospital equipment. This adherence is crucial for complying with South Korea's Personal Information Protection Act (PIPA) and related data laws. Furthermore, the researchers are collaborating with the Ministry of Health and Welfare to prepare for the implementation of FHIR templates, an international standard format for exchanging medical data. This effort aims to establish a unified format for information sharing between different clinics.
Technology Note
FHIR（Fast Healthcare Interoperability Resources）は医療データ交換の国際標準。このエントリの関連技術: FHIR
Original content copyright by respective publishers