미숙아 질환 진단 AI 모델 개발
전북대학교 의생명연구소와의 공동 연구로, 미숙아 질환 진단을 보조하는 AI 모델을 개발한다. 생체신호·의료영상·검사결과·EMR 등 멀티모달 데이터를 결합하여 조기 진단과 위험 예측을 수행하며, 설명 가능한 추론과 병상에서 활용 가능한 수준의 효율적인 모델을 지향한다.
- This project is a joint research project with the Biomedical Research Institute (의생명연구소) of Jeonbuk National University.
- Period: Aug. 2026 – Jan. 2028.
미숙아 질환 진단 AI 모델 개발
The goal of this project is to develop AI models that assist in diagnosing diseases in premature infants. Premature infants in the NICU are monitored through a heterogeneous mix of continuous vital signs, medical images, laboratory results, and EMR records, and the clinically important conditions — respiratory disorders, neurologic injury, infection, and growth complications — often show only subtle early signals across these modalities.
Working with clinicians from the Biomedical Research Institute, this project builds multimodal diagnostic models that combine these signals for early detection and risk prediction, with an emphasis on explainable inference and on models efficient enough to be deployed at the bedside. The resulting models are evaluated against clinical data to validate their usefulness as decision support.