EuropeMedQA Study Protocol: A Multilingual, Multimodal Medical Examination Dataset for Language Model Evaluation

TL;DR

EuropeMedQA is a multilingual, multimodal medical exam dataset from European countries designed to evaluate and improve the cross-lingual and multimodal capabilities of large language models in medical AI.

Contribution

It introduces the first comprehensive European multilingual, multimodal medical exam dataset with a rigorous curation process and automated translation pipeline.

Findings

Evaluated multimodal LLMs on EuropeMedQA using zero-shot prompting.

Assessed cross-lingual transfer and visual reasoning capabilities.

Provided a contamination-resistant benchmark reflecting European clinical practices.

Abstract

While Large Language Models (LLMs) have demonstrated high proficiency on English-centric medical examinations, their performance often declines when faced with non-English languages and multimodal diagnostic tasks. This study protocol describes the development of EuropeMedQA, the first comprehensive, multilingual, and multimodal medical examination dataset sourced from official regulatory exams in Italy, France, Spain, and Portugal. Following FAIR data principles and SPIRIT-AI guidelines, we describe a rigorous curation process and an automated translation pipeline for comparative analysis. We evaluate contemporary multimodal LLMs using a zero-shot, strictly constrained prompting strategy to assess cross-lingual transfer and visual reasoning. EuropeMedQA aims to provide a contamination-resistant benchmark that reflects the complexity of European clinical practices and fosters the development of more generalizable medical AI.