# Direct Preference Optimization for Suppressing Hallucinated Prior Exams in Radiology Report Generation

arXiv: https://arxiv.org/abs/2406.06496 · HTML: https://mlsanity.com/paper/2406.06496/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 17 of its 29 references that are in the map.
- First posted on arXiv 10 Jun 2024.

## Placed in

- Multimodal LLMs / vision-language models / [Medical vision-language models](https://mlsanity.com/mllm/medical-vision-language-models/)
- LLM alignment / RLHF / preference optimisation / [Medical LLM alignment](https://mlsanity.com/alignment/medical-llm-alignment/)

## Builds on (most cited first, 17 of 17)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [RadGraph: Extracting Clinical Entities and Relations from Radiology Reports](https://mlsanity.com/paper/2106.14463/) (2021)
- [Aligning Modalities in Vision Large Language Models via Preference Fine-tuning](https://mlsanity.com/paper/2402.11411/) (2024)
- [Improving Factual Completeness and Consistency of Image-to-Text Radiology Report Generation](https://mlsanity.com/paper/2010.10042/) (2020)
- [Direct Preference Optimization with an Offset](https://mlsanity.com/paper/2402.10571/) (2024)
- [LiPO: Listwise Preference Optimization through Learning-to-Rank](https://mlsanity.com/paper/2402.01878/) (2024)
- [Improving Radiology Report Generation Systems by Removing Hallucinated References to Non-existent Priors](https://mlsanity.com/paper/2210.06340/) (2022)
- [Multimodal Image-Text Matching Improves Retrieval-based Chest X-Ray Report Generation](https://mlsanity.com/paper/2303.17579/) (2023)
- [A Critical Evaluation of AI Feedback for Aligning Large Language Models](https://mlsanity.com/paper/2402.12366/) (2024)
- [Style-Aware Radiology Report Generation with RadGraph and Few-Shot Prompting](https://mlsanity.com/paper/2310.17811/) (2023)
- [Retrieval Augmented Chest X-Ray Report Generation using OpenAI GPT models](https://mlsanity.com/paper/2305.03660/) (2023)
- [Multimodal Large Language Model driven Radiology Report Generation with Clinical Knowledge Enhancement](https://mlsanity.com/paper/2403.06728/) (2024)
- [Integrating Physician Diagnostic Logic into Large Language Models: Preference Learning from Process Feedback](https://mlsanity.com/paper/2401.05695/) (2024)
- [NOTE: Notable generation Of patient Text summaries through Efficient approach based on direct preference optimization](https://mlsanity.com/paper/2402.11882/) (2024)

## Built on by (most cited first, 7 of 7)

- [MMedPO: Aligning Medical Vision-Language Models with Clinical-Aware Multimodal Preference Optimization](https://mlsanity.com/paper/2412.06141/) (2024)
- [ReXErr: Synthesizing Clinically Meaningful Errors in Diagnostic Radiology Reports](https://mlsanity.com/paper/2409.10829/) (2024)
- [CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding](https://mlsanity.com/paper/2509.23379/) (2025)
- [Anatomy-R1: Enhancing Anatomy Reasoning in Multimodal Large Language Models via Anatomical Similarity Curriculum and Group Diversity Augmentation](https://mlsanity.com/paper/2512.19512/) (2025)
- [Benchmark Success, Clinical Failure: When Reinforcement Learning Optimizes for Benchmarks, Not Patients](https://mlsanity.com/paper/2512.23090/) (2025)
- [Foundation Models for Medical Imaging: Status, Challenges, and Directions](https://mlsanity.com/paper/2602.15913/) (2026)
- [Linguistically-Aligned and Visually-Grounded Preference Optimization for Clinically-Augmented Medical Report Generation](https://mlsanity.com/paper/2608.08494/) (2026)

## Abstract (arXiv)

Recent advances in generative vision-language models (VLMs) have exciting potential implications for AI in radiology, yet VLMs are also known to produce hallucinations, nonsensical text, and other unwanted behaviors that can waste clinicians' time and cause patient harm. Drawing on recent work on direct preference optimization (DPO), we propose a simple method for modifying the behavior of pretrained VLMs performing radiology report generation by suppressing unwanted types of generations. We apply our method to the prevention of hallucinations of prior exams, addressing a long-established problem behavior in models performing chest X-ray report generation. Across our experiments, we find that DPO fine-tuning achieves a 3.2-4.8x reduction in lines hallucinating prior exams while maintaining model performance on clinical accuracy metrics. Our work is, to the best of our knowledge, the first work to apply DPO to medical VLMs, providing a data- and compute- efficient way to suppress problem behaviors while maintaining overall clinical accuracy.