Computer Vision News - April 2023

38 Computational Medicine group, Usher, Edinburgh the way from gathering data and creating models to implementing them in practice and evaluating them, ” Miguel tells us. “ I don’t think many labs cover this whole pipeline of activity. ” The group’s research includes using retinal and brain imaging to identify biomarkers for disease diagnosis and prediction, generating novel insights and approaches that will hopefully be translated into clinical practice further down the line. “ To support that, we do a lot of work on gathering large datasets, ” he continues. “ Before any model can be trained, you need data. One of the big challenges in computer vision for medical imaging is that we don’t have the humongous datasets that people use to train models in other areas of computer vision. From very early on, we’ve paid special attention to that. ” There are ongoing efforts to release open- source datasets, including at MICCAI , where challenges regularly provide data Miguel Bernabeu is the Deputy Director and Director of Research at The Bayes Centre, the University of Edinburgh’s innovation hub for data science and artificial intelligence. He is also Professor of Computational Medicine at the Usher Institute within the Edinburgh Medical School, where he established his first research group in the Centre for Medical Informatics. He speaks to us about its work in the field of computer vision for medical imaging. Artificial intelligence has made significant advancements in recent years, and its potential to transform healthcare has become increasingly apparent. With a mix of largely PhD students and postdocs, the Computational Medicine Group at the University of Edinburgh works across every aspect of developing AI systems for medical applications , from early-stage exploratory work to real-world evaluation. “ We work on interesting projects that go all