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.
@article{arxiv.2604.14306,
title = {EuropeMedQA Study Protocol: A Multilingual, Multimodal Medical Examination Dataset for Language Model Evaluation},
author = {Francesco Andrea Causio and Vittorio De Vita and Olivia Riccomi and Michele Ferramola and Federico Felizzi and Alessandro Tosi and Antonio Cristiano and Lorenzo De Mori and Chiara Battipaglia and Melissa Sawaya and Luigi De Angelis and Marcello Di Pumpo and Alessandra Piscitelli and Pietro Eric Risuleo and Alessia Longo and Giulia Vojvodic and Mariapia Vassalli and Bianca Destro Castaniti and Nicolò Scarsi and Manuel Del Medico},
journal= {arXiv preprint arXiv:2604.14306},
year = {2026}
}