Rohan N. Dhamdhere

I’m Rohan Dhamdhere, a PhD Candidate in Biomedical Engineering in the Georgia Tech–Emory program. What motivates me most is closing the gap between what AI can do in a paper and what it can do for a patient — building tools that are not just accurate but interpretable, deployable, and worth a clinician’s trust. I love building translational tools, collaborating with clinicians and engineers across disciplines, and turning complex modeling problems into clear, testable products that can actually work in a hospital.

Outside of research, you’ll find me running, debating whether the Packers’ secondary will ever be elite again, following Indian cricket, or hiking through US national parks.

My research focuses on building AI systems that turn routine clinical data — retinal photographs, 12-lead ECGs, echocardiogram videos, and low-dose CT — into validated cardiovascular risk tools. Models I have developed have been tested in cohorts exceeding 1,500 patients across cardiology, nephrology, hepatology, and transplant medicine, with results published in Kidney360, Lancet eBioMedicine, MICCAI, and JACC Advances. Highlights include a retinal vascular AI model that outperforms the AHA-PREVENT risk calculator in CKD patients, a CT-based liver fat pipeline that underpins an allowed US patent, and an ECG foundation model approach that more than doubles biopsy reduction rates for heart transplant rejection monitoring.

Technically, I work across vision transformers, ECG and image foundation models, graph neural networks, radiomics pipelines, survival analysis, and vision-language models. My particular focus is designing multimodal fusion architectures that combine imaging features, structured clinical variables, and longitudinal physiological signals to produce interpretable, clinically actionable risk scores.