Aligning Clinical Needs and AI Capabilities: A Survey on LLMs for Medical Reasoning
A survey of medical reasoning from clinical and computational perspectives, with a five-level benchmark evaluating 18 models.
Citation
Qi Peng, Jiatong Li, Sirui Huang, Yiyang Jiang, Kaisong Gong, Ronger Ding, Shijie Ye, Changmeng Zheng, Yi Cai, Xiaobo Yang, Jin Huang, Xiao-Yong Wei, and Qing Li. (2026). Aligning Clinical Needs and AI Capabilities: A Survey on LLMs for Medical Reasoning. Machine Intelligence Research.
About this work
This survey connects clinical practice with computational methods through a dual-view framework. On the clinical side, it develops a five-level competency scheme based on Miller’s Pyramid. On the computational side, it links deductive, inductive, and abductive reasoning to common medical goals and tasks.
The work also introduces a five-level medical reasoning benchmark and evaluates 18 state-of-the-art models, highlighting complementary strengths and persistent gaps in diagnosis, decision support, dialogue, hallucination, and grounding.
Co-first author.