MIRAI-MDT

MIRAI-MDT enhances oncology decision-making by embedding explainable artificial intelligence (AI) into Multidisciplinary Team (MDT) workflows. Across Europe, hospitals face growing diagnostic workloads and fragmented data flows between radiology, pathology, and clinical systems. This leads to incomplete case preparation and inconsistent treatment decisions. MIRAI-MDT addresses these challenges by developing a secure, interoperable platform that automatically structures and verifies clinical information, combining AI-driven text and imaging analysis with transparent governance and full compliance with European health data regulations. The project integrates a multilingual clinical language model that converts free-text reports into standardized FHIR data, a video-based ultrasound AI engine that supports real-time lesion assessment, and a privacy-preserving federated learning infrastructure allowing hospitals to collaborate without sharing patient data. Clinical validation in the Netherlands, Slovenia, the Czech Republic, and Türkiye ensures multilingual robustness and cross-border scalability. MIRAI-MDT ultimately aims to improve diagnostic quality, reduce clinician workload, and strengthen Europe’s leadership in trustworthy, standards-based healthcare AI.