AInternDOCTor: Medical Image Analysis Agent for Coronary Optical Coherence Tomography

Abstract

Coronary Optical Coherence Tomography (OCT) plays a pivotal role in shaping clinical decisions for cardiovascular disease management. While recent innovations have led to specialized models for various coronary OCT image analysis tasks, these solutions often operate in isolation, limiting their practical utility in clinical practice. To bridge this gap, we present AInternDOCTor, the first OCT image analysis agent that seamlessly integrates state-of-the-art coronary OCT analysis tools and multi-modal large language models into a unified and extensible framework. Experiment results on our constructed coronary OCT visual question answering dataset demonstrate that the answer accuracy of the proposed agent is significantly better than that of GPT-4o.

Ähnliche Arbeiten