Reinforcement learning improves LLM accuracy and reasoning in disease classification from radiology reports
Abstract
Accurate disease classification from radiology reports is essential for many applications. While supervised fine-tuning (SFT) of lightweight LLMs improves accuracy, it can degrade reasoning. We propose a two-stage approach: SFT on disease labels followed by Group Relative Policy Optimization (GRPO) to refine predictions by optimizing accuracy and format without reasoning supervision. Across three radiologist-annotated datasets, SFT outperformed baselines and GRPO further improved classification and enhanced reasoning recall and comprehensiveness.
Fields of Study & Subjects
Artificial Intelligence in Healthcare and Education
Radiology practices and education
Machine Learning in Healthcare
Publication details
- Journal / source
- npj Digital Medicine
- Publisher
- Springer Science and Business Media LLC
- Volume
- 9
- Issue
- 1
- Publication date
- 30 April 2026
- Language
- EN