Puneet Kumar UT_MDA_BM

I develop AI-driven computational imaging pipelines that extract quantitative biomarkers from multimodal MRI — bridging the gap between raw imaging data and clinically actionable insights in oncology and neuroscience. As a Postdoctoral Fellow at MD Anderson Cancer Center, I design and validate end-to-end deep learning systems (CNNs, U-Net, transformer architectures, self-supervised models) for tumor characterization, tissue segmentation, and perfusion MRI analysis — deployed across large, heterogeneous, multi-site imaging datasets. My research spans the full biomarker development stack — from algorithm design and model validation to multimodal data integration (MRI, pathology, clinical) and translational study design. I collaborate closely with clinicians, radiologists, and computational biologists to ensure that model outputs are interpretable and relevant to clinical decision-making. Previously, I completed my Ph.D. in Computer Science at Jawaharlal Nehru University, where I developed novel fuzzy clustering algorithms for robust brain MRI segmentation, publishing in IEEE Transactions on Fuzzy Systems, Applied Soft Computing, and Neural Computing and Applications. Current focus areas: multimodal neuroimaging biomarker modeling · perfusion MRI (DSC/DCE) · self-supervised denoising · disease staging & longitudinal progression · precision oncology and neuroscience.

HOUSTON, USA

PostDoc Fellow

Medical Center

The University of Texas MD Anderson Cancer Center