Current Research and Scholarly Interests
My group works on technological developments in diagnostic imaging. Specifically, we have developed computer graphics methods for evaluation of imaging data such as CT colonography which is now widely used to screen for colonic polyps. As part of that work, we developed a number of supervised learning methods that automatically detect polyps (computer aided detection). More recently, my attention has focused on medical informatics and machine learning. The overall aim is to make medical imaging data much more computationally accessible so that prior instances of imaging diagnoses can help inform and improve diagnosis in new clinical cases. An example of this is the use of Bayesian modeling of bone tumors and automatic generation of differential diagnosis for focal bone lesions. We have also worked on liver lesions and more recently on deep learning methods for diagnosis in knee MRI.