AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

MedLesionVQA: A Multimodal Benchmark Emulating Clinical Visual Diagnosis for Body Surface Health

Deli Yu, Shengzhi Wang, Kai WU, Xiaozhong Ji, Bo Cui, Jieqiong Cao, Huichao Wang, Boyuan Jiang, Xu Wang, Qian Xu, ChaoGao, Yi Zhao and 5 more

ICLR 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Benchmarks and analysis of MLLMs

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Introduces MedLesionVQA, a multimodal benchmark for body-surface health conditions that emulates stepwise clinical visual diagnosis, addressing limitations of publicly sourced or disease-classification-only medical benchmarks.

Problem

Existing medical multimodal benchmarks are either built from public sources with limited expert curation or focus narrowly on disease classification, failing to capture the stepwise recognition and reasoning process of clinical visual diagnosis for body-surface conditions.

Approach

A benchmark for body-surface health conditions across clinical departments that is designed to mirror stepwise clinical visual diagnosis (lesion recognition and reasoning) rather than disease classification alone, with expert curation.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add image input to text (ablation)GPT-4V · MedLesionVQA validationaverage accuracy: 0.3089 -> 0.4938Improved
add image input to text (ablation)Qwen2.5-VL-72B · MedLesionVQA validationaverage accuracy: 0.3478 -> 0.4904Improved
add image input to text (ablation)GPT-4V · Suggestion & Treatmentaccuracy: 0.8140 -> 0.8216No clear effect
medical-domain fine-tuning (LLaVA-med vs LLaVA-v1.5, 7B)LLaVA-v1.5-7B · MedLesionVQA validationaverage accuracy: 0.2648 -> 0.0885Hurt
medical-domain model (MedGemma-4b vs InternVL2.5-4B)4B MLLM · MedLesionVQA validationaverage accuracy: 0.2632 -> 0.195Hurt
increase model size beyond 10Bopen-source MLLMs · MedLesionVQAaverage accuracy: GLM-4V-9B 0.4654 vs InternVL2.5-38B 0.3096Mixed
increase model size from under 1B to 1B-10Bopen-source MLLMs · MedLesionVQAaverage accuracy: below 0.2 -> 0.3 to 0.4Improved
compare best MLLM with primary physiciansGemini-2.5-pro · MedLesionVQA validationaverage accuracy: 0.5624 vs 0.6144Hurt
add color and size rubric to judge prompt (ablation)GPT-4 judge · open-ended scoringjudge-human consistencyImproved

Experimental setup

Benchmarks: MedLesionVQA

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy