Teaching AI Stepwise Diagnostic Reasoning with Report-Guided Chain-of-Thought Learning
This addresses the need for interpretable and competent AI systems in radiology by providing a scalable method to enhance diagnostic accuracy, though it is incremental as it builds on existing vision-language models.
This study tackled the problem of improving AI diagnostic reasoning in radiology by developing DiagCoT, a framework that uses free-text reports to train vision-language models, resulting in significant gains: zero-shot disease classification AUC increased from 0.52 to 0.76, pathology grounding mIoU from 0.08 to 0.31, and report generation BLEU from 0.11 to 0.33.
This study presents DiagCoT, a multi-stage framework that applies supervised fine-tuning to general-purpose vision-language models (VLMs) to emulate radiologists' stepwise diagnostic reasoning using only free-text reports. DiagCoT combines contrastive image-report tuning for domain alignment, chain-of-thought supervision to capture inferential logic, and reinforcement tuning with clinical reward signals to enhance factual accuracy and fluency. On the MIMIC-CXR benchmark, DiagCoT improved zero-shot disease classification AUC from 0.52 to 0.76 (absolute gain of 0.24), pathology grounding mIoU from 0.08 to 0.31 (absolute gain of 0.23), and report generation BLEU from 0.11 to 0.33 (absolute gain of 0.22). It outperformed state-of-the-art models including LLaVA-Med and CXR-LLAVA on long-tailed diseases and external datasets. By converting unstructured clinical narratives into structured supervision, DiagCoT offers a scalable approach for developing interpretable and diagnostically competent AI systems for radiology.