arXiv cs.AI· Zhenghao Zhu, Chuxue Cao, Sirui Han, Yuanfeng Song, Xing Chen, Caleb Chen Cao, Yike Guo·· 12 小时前AI 评分31
MedInsightBench：评估多模态医疗数据分析智能体的多步洞察发现基准
MedInsightBench: Evaluating Medical Analytics Agents Through Multi-Step Insight Discovery in Multimodal Medical Data
AI 导读
研究团队提出 MedInsightBench，首个用于评估大型多模态模型（LMMs）医疗洞察发现能力的基准，包含 332 个精选医疗案例，每个案例均标注了设计好的洞察。
正文 · 原文
Abstract:In medical data analysis, extracting deep insights from complex, multi-modal datasets is essential for improving patient care, increasing diagnostic accuracy, and optimizing healthcare operations. However, there is currently a lack of high-quality datasets specifically designed to evaluate the ability of large multi-modal models (LMMs) to discover medical insights. In this paper, we introduce MedInsightBench, the first benchmark that comprises 332 carefully curated medical cases, each annotated with thoughtfully designed insights. This benchmark is intended to evaluate the ability of LMMs and agent frameworks to analyze multi-modal medical image data, including posing relevant questions, interpreting complex findings, and synthesizing actionable insights and recommendations. Our analysis indicates that existing LMMs exhibit limited performance on MedInsightBench, which is primarily attributed to their challenges in extracting multi-step, deep insights and the absence of medical expertise. Therefore, we propose MedInsightAgent, an automated agent framework for medical data analysis, composed of three modules: Visual Root Finder, Analytical Insight Agent, and Follow-up Question Composer. Experiments on MedInsightBench highlight pervasive challenges and demonstrate that MedInsightAgent can improve the performance of general LMMs in medical data insight discovery.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2512.13297 [cs.AI] |
| (or arXiv:2512.13297v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2512.13297 arXiv-issued DOI via DataCite |
Submission history
From: Zhenghao Zhu [view email]
[v1]
Mon, 15 Dec 2025 13:10:42 UTC (5,306 KB)
[v2]
Fri, 23 Oct 2026 08:56:22 UTC (4,924 KB)
来源：arXiv cs.AI · arxiv.org