Brain Functional Connectivity
I’m working on developing a statistical framework for relating static and dynamic functional and effective brain connectivity maps with clinical and more complex variables of interest such as genetics or integrated imaging in other modalities (such as EEG or structural imaging) for population inference.
We are working on developing clinical and imaging based biomarkers of Early-onset Alzheimer’s Disease as a part of the LEADS study. We develop sophisticated data analytic tools for longitudinal analysis of the imaging biomarkers, development of psychometric test scores, and investigate the association of the genotype on longitudinal imaging biomarkers.
Machine Learning for Tumor Heterogeneity Estimation
I’m interested in developing statistical models for estimation of cancer tumor heterogeneity using noninvasive imaging data for prediction of disease outcomes and survival. My group is working on developing dimension reduction methods based on high dimensional manifolds, developing deep learning algorithms for prediction using imaging data, and implementation of machine learning algorithms for estimation of tumor heterogeneity.