Ghassan Hamarneh

Our team's research focuses on developing novel computer vision and machine learning methods capable of automatically interpreting images to mimic and complement human vision while being faster, more reproducible, and more accurate. The primary applications of our research are focused on advancing medical technologies and improving healthcare systems through computational analysis of ubiquitous high-dimensional biomedical imaging data. More specifically, we develop computational methods that emulate and augment heavily-trained domain experts under stringent performance guidelines to overcome problems in computer aided diagnostics, robotic and minimally invasive intervention, precision medicine, big data analytics, etc. Our computer vision and machine learning methods are designed to primarily solve image segmentation, registration, and classification problems. This in turn entails constructing, optimizing, and validating novel mathematical and computational models of shape, appearance and deformation of complex and dynamic structural and functional data. These models combine explicit encoding of expert knowledge with machine learning from big data.