35 Albarqouni Lab Computer Vision News The third theme, which Shadi says is closest to his heart but one for which they have not seen as much progress, is affordable AI. “ The devices themselves should be affordable, butwe’re focusing more on computational algorithms that could work in low-resource settings, ” he tells us. “ This is something I’mhighly interested in. Developing these algorithms and deploying them in low middle-income countries or even remote villages in developed countries! ” One of thelab’s recent notable achievements is a study published inNature Machine Intelligence , where an algorithm was trained to detect and segment lesions or tumors using MR brain imaging data without annotations. By training the algorithm on healthy patient data from multiple hospitals, it learned the characteristics of a healthy brain. When provided with data from patients with lesions or tumors, the algorithm generated a pseudo-healthy version of the patient, allowing for accurate quantification and analysis of the aggressiveness of the condition. The Albarqouni Lab aims to develop the next generation of AI in medicine to improve the clinical workflow for patients and professionals . If you are a postdoc or a PhD, you may be pleased to hear they are hiring for two new positions on a DFG-funded radiotherapy project. If that piques your interest, do not hesitate to contact them! To learn more about the work of the Albarqouni Lab, including their other federated learning projects; the challenges they’ve overcome; and their open-source, cross-silo healthcare dataset suite, check out our video interview with Shadi above.
RkJQdWJsaXNoZXIy NTc3NzU=