Explore high-quality datasets for your AI and machine learning projects.
MedPix 2.0 is a comprehensive multimodal biomedical dataset developed by the Department of Engineering, University of Palermo, derived from the MedPix® database and primarily intended for continuing medical education and clinical research. It contains over 12 000 patient cases, each with at least one medical image and a detailed clinical report. The dataset was extracted via a semi‑automatic pipeline with manual correction, stored in MongoDB and navigable through a GUI. MedPix 2.0 is suitable for training multimodal large language models, especially for medical image classification and diagnostic support systems.
