# MedGround-R1: Advancing Medical Image Grounding via Spatial-Semantic Rewarded Group Relative Policy Optimization

arXiv: https://arxiv.org/abs/2507.02994 · HTML: https://mlsanity.com/paper/2507.02994/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning.
- Builds on 25 of its 29 references that are in the map.
- First posted on arXiv 1 Jul 2025.

## Placed in

- LLM reasoning / [Clinical and biomedical applications](https://mlsanity.com/lr/clinical-and-biomedical-applications/)

## Builds on (most cited first, 25 of 25)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases](https://mlsanity.com/paper/1705.02315/) (2017)
- [TotalSegmentator: robust segmentation of 104 anatomical structures in CT images](https://mlsanity.com/paper/2208.05868/) (2022)
- [MedCLIP: Contrastive Learning from Unpaired Medical Images and Text](https://mlsanity.com/paper/2210.10163/) (2022)
- [Making the Most of Text Semantics to Improve Biomedical Vision--Language Processing](https://mlsanity.com/paper/2204.09817/) (2022)
- [M3D: Advancing 3D Medical Image Analysis with Multi-Modal Large Language Models](https://mlsanity.com/paper/2404.00578/) (2024)
- [LViT: Language meets Vision Transformer in Medical Image Segmentation](https://mlsanity.com/paper/2206.14718/) (2022)
- [A Closer Look at the Explainability of Contrastive Language-Image Pre-training](https://mlsanity.com/paper/2304.05653/) (2023)
- [SlideChat: A Large Vision-Language Assistant for Whole-Slide Pathology Image Understanding](https://mlsanity.com/paper/2410.11761/) (2024)
- [RainMamba: Enhanced Locality Learning with State Space Models for Video Deraining](https://mlsanity.com/paper/2407.21773/) (2024)
- [Ariadne's Thread:Using Text Prompts to Improve Segmentation of Infected Areas from Chest X-ray images](https://mlsanity.com/paper/2307.03942/) (2023)
- [Medical Phrase Grounding with Region-Phrase Context Contrastive Alignment](https://mlsanity.com/paper/2303.07618/) (2023)
- [OphCLIP: Hierarchical Retrieval-Augmented Learning for Ophthalmic Surgical Video-Language Pretraining](https://mlsanity.com/paper/2411.15421/) (2024)
- [Cross-Modal Conditioned Reconstruction for Language-guided Medical Image Segmentation](https://mlsanity.com/paper/2404.02845/) (2024)
- [Interpretable Bilingual Multimodal Large Language Model for Diverse Biomedical Tasks](https://mlsanity.com/paper/2410.18387/) (2024)
- [Serp-Mamba: Advancing High-Resolution Retinal Vessel Segmentation with Selective State-Space Model](https://mlsanity.com/paper/2409.04356/) (2024)
- [ChEX: Interactive Localization and Region Description in Chest X-rays](https://mlsanity.com/paper/2404.15770/) (2024)
- [CausalCLIPSeg: Unlocking CLIP's Potential in Referring Medical Image Segmentation with Causal Intervention](https://mlsanity.com/paper/2503.15949/) (2025)
- [A Refer-and-Ground Multimodal Large Language Model for Biomedicine](https://mlsanity.com/paper/2406.18146/) (2024)
- [Uncertainty-aware Medical Diagnostic Phrase Identification and Grounding](https://mlsanity.com/paper/2404.06798/) (2024)
- [Language-Driven Interactive Shadow Detection](https://mlsanity.com/paper/2408.08543/) (2024)
- [Metal Artifact Reduction in 2D CT Images with Self-supervised Cross-domain Learning](https://mlsanity.com/paper/2109.13483/) (2021)

## Built on by (most cited first, 23 of 23)

- [A Survey of Reinforcement Learning for Large Reasoning Models](https://mlsanity.com/paper/2509.08827/) (2025)
- [From AI for Science to Agentic Science: A Survey on Autonomous Scientific Discovery](https://mlsanity.com/paper/2508.14111/) (2025)
- [A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontiers](https://mlsanity.com/paper/2508.21148/) (2025)
- [UniMedVL: Unifying Medical Multimodal Understanding and Generation through Observation-Knowledge-Analysis](https://mlsanity.com/paper/2510.15710/) (2025)
- [Medical Reasoning in the Era of LLMs: A Systematic Review of Enhancement Techniques and Applications](https://mlsanity.com/paper/2508.00669/) (2025)
- [MedQ-Bench: Evaluating and Exploring Medical Image Quality Assessment Abilities in MLLMs](https://mlsanity.com/paper/2510.01691/) (2025)
- [MedSAM-Agent: Empowering Interactive Medical Image Segmentation with Multi-turn Agentic Reinforcement Learning](https://mlsanity.com/paper/2602.03320/) (2026)
- [When Does RL Help Medical VLMs? Disentangling Vision, SFT, and RL Gains](https://mlsanity.com/paper/2603.01301/) (2026)
- [Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language Models](https://mlsanity.com/paper/2510.18303/) (2025)
- [Surg-R1: A Hierarchical Reasoning Foundation Model for Scalable and Interpretable Surgical Decision Support with Multi-Center Clinical Validation](https://mlsanity.com/paper/2603.12430/) (2026)
- [From Failure to Feedback: Group Revision Unlocks Hard Cases in Object-Level Grounding](https://mlsanity.com/paper/2605.15951/) (2026)
- [UniReason-Med: A Shared Grounded Reasoning Interface for 2D-to-3D Transfer in Medical VQA](https://mlsanity.com/paper/2606.11740/) (2026)
- [MedFact-R1: Towards Factual Medical Reasoning via Pseudo-Label Augmentation](https://mlsanity.com/paper/2509.15154/) (2025)
- [Topo-R1: Detecting Topological Anomalies via Vision-Language Models](https://mlsanity.com/paper/2603.13054/) (2026)
- [Enhancing Medical Visual Grounding via Knowledge-guided Spatial Prompts](https://mlsanity.com/paper/2604.01915/) (2026)
- [MedProbeBench: Systematic Benchmarking at Deep Evidence Integration for Expert-level Medical Guideline](https://mlsanity.com/paper/2604.18418/) (2026)
- [S2-UniSeg: Fast Universal Agglomerative Pooling for Scalable Segment Anything without Supervision](https://mlsanity.com/paper/2508.06995/) (2025)
- [EventRR: Event Referential Reasoning for Referring Video Object Segmentation](https://mlsanity.com/paper/2508.07171/) (2025)
- [CheXPO-v2: Preference Optimization for Chest X-ray VLMs with Knowledge Graph Consistency](https://mlsanity.com/paper/2512.17213/) (2025)
- [MedLoc-R1: Performance-Aware Curriculum Reward Scheduling for GRPO-Based Medical Visual Grounding](https://mlsanity.com/paper/2603.28120/) (2026)
- [MedVol-R1: Reward-Driven Evidence Grounding for Volumetric Reasoning Segmentation](https://mlsanity.com/paper/2605.26621/) (2026)
- [Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning](https://mlsanity.com/paper/2606.31599/) (2026)
- [SAM Meets VLM: Parameter-Decoupled Full-Parameter Training for Unified Medical Reasoning and Segmentation](https://mlsanity.com/paper/2609.37283/) (2026)

## Abstract (arXiv)

Medical Image Grounding (MIG), which involves localizing specific regions in medical images based on textual descriptions, requires models to not only perceive regions but also deduce spatial relationships of these regions. Existing Vision-Language Models (VLMs) for MIG often rely on Supervised Fine-Tuning (SFT) with large amounts of Chain-of-Thought (CoT) reasoning annotations, which are expensive and time-consuming to acquire. Recently, DeepSeek-R1 demonstrated that Large Language Models (LLMs) can acquire reasoning abilities through Group Relative Policy Optimization (GRPO) without requiring CoT annotations. In this paper, we adapt the GRPO reinforcement learning framework to VLMs for Medical Image Grounding. We propose the Spatial-Semantic Rewarded Group Relative Policy Optimization to train the model without CoT reasoning annotations. Specifically, we introduce Spatial-Semantic Rewards, which combine spatial accuracy reward and semantic consistency reward to provide nuanced feedback for both spatially positive and negative completions. Additionally, we propose to use the Chain-of-Box template, which integrates visual information of referring bounding boxes into the <think> reasoning process, enabling the model to explicitly reason about spatial regions during intermediate steps. Experiments on three datasets MS-CXR, ChestX-ray8, and M3D-RefSeg demonstrate that our method achieves state-of-the-art performance in Medical Image Grounding. Ablation studies further validate the effectiveness of each component in our approach. Code, checkpoints, and datasets are available at https://github.com/bio-mlhui/MedGround-R1