Building TrustworthyDigital Twins for Patient-Centred Medicine
University of Crete
The University of Crete (UoC) is one of Greece’s leading research-intensive universities, with a strong international profile in biomedical research, medical imaging, artificial intelligence, and translational medicine. Within the School of Medicine, the Artificial Intelligence and Translational Imaging (ATI) Lab brings together expertise in radiology, medical image analysis, radiomics, deep learning, generative AI, and multimodal data integration. The lab focuses on developing and validating trustworthy AI methods that can support clinically meaningful research and future translation into healthcare. Its work spans quantitative imaging biomarkers, multimodal learning, synthetic data and agentic AI on projects related to oncologic and musculoskeletal imaging. Through close links with clinical departments, European research networks, and interdisciplinary collaborators, ATI Lab combines methodological development with access to real-world imaging expertise and clinically grounded validation, supporting the development of reproducible and clinically relevant AI solutions.
Role within TWIN-X
Within TWIN-X, the University of Crete leads Work Package 3, which focuses on medical imaging and histopathology embeddings. ATI Lab will lead the development of segmentation-guided radiology encoders for MRI, CT, radiography, and ultrasound, with particular emphasis on robustness across scanners, vendors, and acquisition protocols. The team will coordinate the radiology-related activities of WP3, contribute to the definition of common imaging pipelines, quality-control procedures, metadata requirements, and interoperable embedding interfaces, and support integration with the multimodal digital twin framework. UoC will also contribute to cross-modality calibration, uncertainty estimation, and external validation, and will lead imaging interpretability activities through feature-attribution and attention-visualisation
