Computer Vision News 34 Medical Imaging Lab The Albarqouni Lab focuses on three main areas of study: computational medical imaging, federated learning, and affordable AI . In computational medical imaging, it develops better, faster, and more accurate deep learning algorithms to make the lives of clinicians and physicians easier, regardless of disease or modality. An important part is working with data from different hospitals without requiring extensive radiologist annotations. Shadi explains that the lab’swork includes developing algorithms that can learn to recognize a limited amount of data; adapt to different domains, scanners, and demographic factors; and generalize to unseen classes or domains. It has also explored using non-imaging data, such as electronic health records, through geometric deep learning. Additionally, it has worked with colleagues fromStanford and ETH Zurich on making models more interpretable and trustworthy. Federated learning involves training algorithms in a distributed manner while preserving patient privacy . This approach allows the lab to leverage data from diverse medical settings without compromising the confidentiality of sensitive information. Shadi Albarqouni is Professor of Computational Medical Imaging Research at the University of Bonn and the University Hospital Bonn. He is also Young Investigator Group Leader at Helmholtz AI, Helmholtz Munich in Germany.
RkJQdWJsaXNoZXIy NTc3NzU=