MMedExpert-R1: Strengthening Multimodal Medical Reasoning via Domain-Specific Adaptation and Clinical Guideline Reinforcement
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912831546327040 |
|---|---|
| author | Ding, Meidan Zhang, Jipeng Wang, Wenxuan Zhong, Haiqin Luo, Xiaoling Chen, Wenting Shen, Linlin |
| author_facet | Ding, Meidan Zhang, Jipeng Wang, Wenxuan Zhong, Haiqin Luo, Xiaoling Chen, Wenting Shen, Linlin |
| contents | Medical Vision-Language Models (MedVLMs) excel at perception tasks but struggle with complex clinical reasoning required in real-world scenarios. While reinforcement learning (RL) has been explored to enhance reasoning capabilities, existing approaches face critical mismatches: the scarcity of deep reasoning data, cold-start limits multi-specialty alignment, and standard RL algorithms fail to model clinical reasoning diversity. We propose MMedExpert-R1, a novel reasoning MedVLM that addresses these challenges through domain-specific adaptation and clinical guideline reinforcement. We construct MMedExpert, a high-quality dataset of 10K samples across four specialties with step-by-step reasoning traces. Our Domain-Specific Adaptation (DSA) creates specialty-specific LoRA modules to provide diverse initialization, while Guideline-Based Advantages (GBA) explicitly models different clinical reasoning perspectives to align with real-world diagnostic strategies. Conflict-Aware Capability Integration then merges these specialized experts into a unified agent, ensuring robust multi-specialty alignment. Comprehensive experiments demonstrate state-of-the-art performance, with our 7B model achieving 27.50 on MedXpert-MM and 83.03 on OmniMedVQA, establishing a robust foundation for reliable multimodal medical reasoning systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_10949 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MMedExpert-R1: Strengthening Multimodal Medical Reasoning via Domain-Specific Adaptation and Clinical Guideline Reinforcement Ding, Meidan Zhang, Jipeng Wang, Wenxuan Zhong, Haiqin Luo, Xiaoling Chen, Wenting Shen, Linlin Computer Vision and Pattern Recognition Medical Vision-Language Models (MedVLMs) excel at perception tasks but struggle with complex clinical reasoning required in real-world scenarios. While reinforcement learning (RL) has been explored to enhance reasoning capabilities, existing approaches face critical mismatches: the scarcity of deep reasoning data, cold-start limits multi-specialty alignment, and standard RL algorithms fail to model clinical reasoning diversity. We propose MMedExpert-R1, a novel reasoning MedVLM that addresses these challenges through domain-specific adaptation and clinical guideline reinforcement. We construct MMedExpert, a high-quality dataset of 10K samples across four specialties with step-by-step reasoning traces. Our Domain-Specific Adaptation (DSA) creates specialty-specific LoRA modules to provide diverse initialization, while Guideline-Based Advantages (GBA) explicitly models different clinical reasoning perspectives to align with real-world diagnostic strategies. Conflict-Aware Capability Integration then merges these specialized experts into a unified agent, ensuring robust multi-specialty alignment. Comprehensive experiments demonstrate state-of-the-art performance, with our 7B model achieving 27.50 on MedXpert-MM and 83.03 on OmniMedVQA, establishing a robust foundation for reliable multimodal medical reasoning systems. |
| title | MMedExpert-R1: Strengthening Multimodal Medical Reasoning via Domain-Specific Adaptation and Clinical Guideline Reinforcement |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.10949 |