Preference fine-tuning for factuality in chest X-ray interpretation models without human feedback

Dennis Hein, Zhihong Chen, Sophie Ostmeier, Justin Xu, Maya Varma, Eduardo Pontes Reis, Arne Edward Michalson MD, Christian Bluethgen, Hyun Joo Shin, Curtis Langlotz, Akshay S Chaudhari

Keywords: LLMs, VLMs, preference fine-tuning, RLHF, DPO, DAAs, preference alignment, radiology, chest X-rays

Abstract: Radiologists play a crucial role by translating medical images into actionable reports. However, the field faces staffing shortages and increasing workloads. While automated approaches using vision-language models (VLMs) show promise as assistants, they require exceptionally high accuracy. Most current VLMs in radiology rely solely on supervised fine-tuning (SFT). Meanwhile, in the general domain, additional preference fine-tuning has become standard practice. The challenge in radiology lies in the prohibitive cost of obtaining radiologist feedback. To address this challenge, we propose an automated pipeline for preference feedback, focusing on chest X-ray (CXR) report generation. Our method leverages publicly available datasets containing pairs of images and radiologist-written reference reports with an LLM-as-a-Judge mechanism, eliminating the need for *additional radiologist feedback*. We evaluate and benchmark five direct alignment algorithms. Our results show up to a 57.4\% improvement in average GREEN scores, a LLM-based metric for evaluating CXR reports, compared to the SFT baseline. We study reward overoptimization via length exploitation, with reports lengthening by up to 3.2x. To assess a potential alignment tax, we benchmark on six additional diverse tasks, finding no significant degradations. A reader study involving four board-certified radiologists indicates win rates of up to 0.62 over the SFT baseline, and macro-averaged F1 scores improve by up to 6.7\%, highlighting the utility of our approach.

Primary Area: foundation or frontier models, including LLMs

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 10713