New framework enables AI to capture and scale the diagnostic expertise of master physicians through lightweight language models

Med-Shicheng framework allows LLMs to systematically learn and transfer distinguished physicians' medical knowledge and decision-making methodologies

Addresses the challenge of scaling master physicians' expertise, which traditionally develops slowly and is difficult to transmit widely

Built on Tianyi model and designed to standardize case-dependent adaptation rules across clinical practice

Research focuses on five National Masters of Chinese Medicine, curating multi-source materials to train a single model to internalize their expertise