# MedVol-R1: Reward-Driven Evidence Grounding for Volumetric Reasoning Segmentation

arXiv: https://arxiv.org/abs/2605.26621 · HTML: https://mlsanity.com/paper/2605.26621/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 22 references that are in the map.
- First posted on arXiv 26 May 2026.

## Builds on (most cited first, 18 of 18)

- [LISA: Reasoning Segmentation via Large Language Model](https://mlsanity.com/paper/2308.00692/) (2023)
- [Visual-RFT: Visual Reinforcement Fine-Tuning](https://mlsanity.com/paper/2503.01785/) (2025)
- [Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement](https://mlsanity.com/paper/2503.06520/) (2025)
- [M3D: Advancing 3D Medical Image Analysis with Multi-Modal Large Language Models](https://mlsanity.com/paper/2404.00578/) (2024)
- [SAM-Med2D](https://mlsanity.com/paper/2308.16184/) (2023)
- [MedSAM2: Segment Anything in 3D Medical Images and Videos](https://mlsanity.com/paper/2504.03600/) (2025)
- [SegVol: Universal and Interactive Volumetric Medical Image Segmentation](https://mlsanity.com/paper/2311.13385/) (2023)
- [VideoEspresso: A Large-Scale Chain-of-Thought Dataset for Fine-Grained Video Reasoning via Core Frame Selection](https://mlsanity.com/paper/2411.14794/) (2024)
- [Large-Vocabulary Segmentation for Medical Images with Text Prompts](https://mlsanity.com/paper/2312.17183/) (2023)
- [VILA-M3: Enhancing Vision-Language Models with Medical Expert Knowledge](https://mlsanity.com/paper/2411.12915/) (2024)
- [Lenna: Language Enhanced Reasoning Detection Assistant](https://mlsanity.com/paper/2312.02433/) (2023)
- [Deep Interactive Segmentation of Medical Images: A Systematic Review and Taxonomy](https://mlsanity.com/paper/2311.13964/) (2023)
- [MedGround-R1: Advancing Medical Image Grounding via Spatial-Semantic Rewarded Group Relative Policy Optimization](https://mlsanity.com/paper/2507.02994/) (2025)
- [Reinforcing Video Reasoning Segmentation to Think Before It Segments](https://mlsanity.com/paper/2508.11538/) (2025)
- [MedReasoner: Reinforcement Learning Drives Reasoning Grounding from Clinical Thought to Pixel-Level Precision](https://mlsanity.com/paper/2508.08177/) (2025)
- [MedGround: Bridging the Evidence Gap in Medical Vision-Language Models with Verified Grounding Data](https://mlsanity.com/paper/2601.06847/) (2026)
- [Mask-Enhanced Segment Anything Model for Tumor Lesion Semantic Segmentation](https://mlsanity.com/paper/2403.05912/) (2024)
- [Boltzmann Attention Sampling for Image Analysis with Small Objects](https://mlsanity.com/paper/2503.02841/) (2025)

## Abstract (arXiv)

Volumetric Reasoning Segmentation (VRS) aims to segment a target region in a 3D medical scan from a free-form clinical query, where the referent is often implicit and requires both medical knowledge and volume-grounded reasoning. Existing methods typically rely on specialized segmentation tokens to connect language with mask decoding, but this coupling collapses the decision process into opaque latent representations, limiting interpretability and generalization to diverse narrative expressions. In this paper, we present MedVol-R1, a reinforcement learning-based framework for VRS that explicitly decouples evidence grounding from volumetric delineation: the LVLM grounds clinical reasoning to a verifiable 2D evidence anchor (key axial slice and 2D bounding boxes), which is then propagated into a coherent 3D mask by a frozen MedSAM2 module. We train MedVol-R1 with cold-start supervised fine-tuning followed by GRPO, guided by a multi-component reward that encourages informative evidence selection, accurate 2D spatial grounding, and cross-slice volumetric coherence, without requiring costly chain-of-thought annotations. Experiments on CT-ORG, AbdomenCT-1K, and KiTS23 from the M3D-Seg benchmark demonstrate that MedVol-R1 consistently outperforms strong baselines and achieves state-of-the-art performance, with reinforcement learning providing clear gains over pure supervised fine-tuning.