Research map / Video / Video-language understanding
MedScope: Incentivizing "Think with Videos" for Clinical Reasoning via Coarse-to-Fine Tool Calling
Wenjie Li, Yujie Zhang, Haoran Sun, Xingqi He, Hongcheng Gao, Chenglong Ma, Ming Hu, Guankun Wang, Shiyi Yao, Renhao Yang, Hongliang Ren, Lei Wang and 2 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add SFT on visual-CoT tool trajectories on top of warm-up (ablation) | MedScope · SVU-31K perception reasoning | CIDEr: 4.54 -> 5.21 | Improved |
| add RL (GA-GRPO) on top of Warm-up + SFT (ablation) | MedScope · SVU-31K perception reasoning | CIDEr: 5.21 -> 10.93 | Improved |
| add SFT cold-start before RL (vs RL directly after warm-up) (ablation) | MedScope · SVU-31K perception reasoning | CIDEr: 5.40 -> 10.93 | Improved |
| add VCoT trajectories in SFT (ablation) | MedScope · SVU-31K temporal reasoning | ROUGE-L: 25.40 -> 34.83 | Improved |
| add VCoT trajectories in SFT before RL (ablation) | MedScope · SVU-31K perception reasoning | CIDEr: 5.80 -> 10.93 | Improved |
| add evidence reward (ablation) | MedScope GA-GRPO · ClinVideo-Eval temporal grounding | R@0.5: 33.2 -> 40.1 | Improved |
| add IoU bonus to evidence reward (ablation) | MedScope GA-GRPO · ClinVideo-Eval temporal grounding | R@0.7: 20.3 -> 22.4 | Improved |
| modulate advantages for both positive and negative trajectories vs positive-only (ablation) | MedScope GA-GRPO · ClinVideo-Eval temporal grounding | R@0.7 gain over baseline: 6.6% -> 12.9% | Improved |
| apply evidence reward only when answer correct (conditional) (ablation) | MedScope GA-GRPO · ClinVideo-Eval temporal grounding | grounding quality | Hurt |
| train with warm-up only (no tool SFT or RL) (ablation) | MedScope · SVU-31K temporal reasoning | CIDEr: 1.55 -> 2.16 | No clear effect |
| add RL after SFT on temporal reasoning CIDEr (ablation) | MedScope · SVU-31K temporal reasoning | CIDEr: 4.70 (Warm-up+RL) vs 4.56 (full) | Mixed |
Closest papers by meaning
- MedHorizon: Towards Long-context Medical Video Understanding in the Wild (NeurIPS 2026 · similarity 0.92)
- Med-CMR: A Fine-Grained Benchmark Integrating Visual Evidence and Clinical Logic for Medical Complex Multimodal Reasoning (CVPR 2026 · similarity 0.89)
- Medical thinking with multiple images (ICLR 2026 · similarity 0.89)
- CARE: Towards Clinical Accountability in Multi-Modal Medical Reasoning with an Evidence-Grounded Agentic Framework (ICLR 2026 · similarity 0.88)
- SurgCoT: Advancing Spatiotemporal Reasoning in Surgical Videos through a Chain-of-Thought Benchmark (CVPR 2026 · similarity 0.87)
- MedEyes: Learning Dynamic Visual Focus for Medical Progressive Diagnosis (AAAI 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
