My research focuses on the cognitive effects of chronic medical conditions, especially cancer. I apply connectomics and machine learning to neuroimaging data to predict cognitive outcomes as well as treatment response and overall survival. My work involves both human and animal neuroimaging to diagnose and predict cognitive deficits associated with malignant tumors, cancer therapies (chemotherapy, radiation) and spaceflight radiation exposure.
Left: The connectome is a mathematical model of the brain network shown here as mapped from resting state fMRI scans of 50 healthy adults. The connectome shows a “small-world” organization with high local connectivity and sparser long-range connectivity. Right: Research from my lab shows that pre-treatment connectome organization accurately predicts 1-year post-treatment cognitive impairment in patients with primary breast cancer (COG), diffuse glioma molecular status (IDH) and diffuse glioma overall survival (OS) with areas under the curve of 0.89 or higher.