S-Chain: Structured Visual Chain-of-Thought For Medicine
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911233961099264 |
|---|---|
| author | Le-Duc, Khai Nguyen, Duy M. H. Trinh, Phuong T. H. Nguyen, Tien-Phat Diep, Nghiem T. Ngo, An Vu, Tung Vuong, Trinh Nguyen, Anh-Tien Nguyen, Mau Hoang, Van Trung Nguyen, Khai-Nguyen Nguyen, Hy Ngo, Chris Liu, Anji Ho, Nhat Hauschild, Anne-Christin Nguyen, Khanh Xuan Nguyen-Tang, Thanh Xie, Pengtao Sonntag, Daniel Zou, James Niepert, Mathias Nguyen, Anh Totti |
| author_facet | Le-Duc, Khai Nguyen, Duy M. H. Trinh, Phuong T. H. Nguyen, Tien-Phat Diep, Nghiem T. Ngo, An Vu, Tung Vuong, Trinh Nguyen, Anh-Tien Nguyen, Mau Hoang, Van Trung Nguyen, Khai-Nguyen Nguyen, Hy Ngo, Chris Liu, Anji Ho, Nhat Hauschild, Anne-Christin Nguyen, Khanh Xuan Nguyen-Tang, Thanh Xie, Pengtao Sonntag, Daniel Zou, James Niepert, Mathias Nguyen, Anh Totti |
| contents | Faithful reasoning in medical vision-language models (VLMs) requires not only accurate predictions but also transparent alignment between textual rationales and visual evidence. While Chain-of-Thought (CoT) prompting has shown promise in medical visual question answering (VQA), no large-scale expert-level dataset has captured stepwise reasoning with precise visual grounding. We introduce S-Chain, the first large-scale dataset of 12,000 expert-annotated medical images with bounding boxes and structured visual CoT (SV-CoT), explicitly linking visual regions to reasoning steps. The dataset further supports 16 languages, totaling over 700k VQA pairs for broad multilingual applicability. Using S-Chain, we benchmark state-of-the-art medical VLMs (ExGra-Med, LLaVA-Med) and general-purpose VLMs (Qwen2.5-VL, InternVL2.5), showing that SV-CoT supervision significantly improves interpretability, grounding fidelity, and robustness. Beyond benchmarking, we study its synergy with retrieval-augmented generation, revealing how domain knowledge and visual grounding interact during autoregressive reasoning. Finally, we propose a new mechanism that strengthens the alignment between visual evidence and reasoning, improving both reliability and efficiency. S-Chain establishes a new benchmark for grounded medical reasoning and paves the way toward more trustworthy and explainable medical VLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_22728 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | S-Chain: Structured Visual Chain-of-Thought For Medicine Le-Duc, Khai Nguyen, Duy M. H. Trinh, Phuong T. H. Nguyen, Tien-Phat Diep, Nghiem T. Ngo, An Vu, Tung Vuong, Trinh Nguyen, Anh-Tien Nguyen, Mau Hoang, Van Trung Nguyen, Khai-Nguyen Nguyen, Hy Ngo, Chris Liu, Anji Ho, Nhat Hauschild, Anne-Christin Nguyen, Khanh Xuan Nguyen-Tang, Thanh Xie, Pengtao Sonntag, Daniel Zou, James Niepert, Mathias Nguyen, Anh Totti Machine Learning Computer Vision and Pattern Recognition Faithful reasoning in medical vision-language models (VLMs) requires not only accurate predictions but also transparent alignment between textual rationales and visual evidence. While Chain-of-Thought (CoT) prompting has shown promise in medical visual question answering (VQA), no large-scale expert-level dataset has captured stepwise reasoning with precise visual grounding. We introduce S-Chain, the first large-scale dataset of 12,000 expert-annotated medical images with bounding boxes and structured visual CoT (SV-CoT), explicitly linking visual regions to reasoning steps. The dataset further supports 16 languages, totaling over 700k VQA pairs for broad multilingual applicability. Using S-Chain, we benchmark state-of-the-art medical VLMs (ExGra-Med, LLaVA-Med) and general-purpose VLMs (Qwen2.5-VL, InternVL2.5), showing that SV-CoT supervision significantly improves interpretability, grounding fidelity, and robustness. Beyond benchmarking, we study its synergy with retrieval-augmented generation, revealing how domain knowledge and visual grounding interact during autoregressive reasoning. Finally, we propose a new mechanism that strengthens the alignment between visual evidence and reasoning, improving both reliability and efficiency. S-Chain establishes a new benchmark for grounded medical reasoning and paves the way toward more trustworthy and explainable medical VLMs. |
| title | S-Chain: Structured Visual Chain-of-Thought For Medicine |
| topic | Machine Learning Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.22728 |