resume
last updated: April 2026
overview
research focuses on using medical imaging in tandem with genomics to:
- better predict endpoints of interest in cancer (survival, drug response, recurrence, and subtype)
- better understand the underlying biology
modalities include MRIs, whole slide images (WSI), multiplex immunofluorescence (miF), spatial transcriptomics and other omics datasets.
Contracted to support cancer research at the National Cancer Institute's Center for Biomedical Informatics and Information Technology (CBIIT), within the Computational Genomics and Bioinformatics (CGBB) branch.
pathology & methods
- Applies vision foundation models (i.e., UNI, CONCH, TITAN, PRISM) to digital pathology tasks including tumor subtyping, survival prediction and unsupervised clustering.
- Implements explainable AI methods (GradCAM, SHAP, and custom methodology) to extract biologically interpretable insights from multimodal models.
- Works with pipelines for multiplex immunofluorescence (miF) and spatial transcriptomics to study tumor microenvironment and immune response.
radiology & collaboration
- Develops MRI-based deep learning pipelines for biomarker predictions, resulting in a manuscript submission to NPJ Precision Oncology (In Rebuttal).
- Technical lead in privacy preserving, multi-institutional collaboration using federated learning for survival prediction in Glioblastoma.
- Communicates complex technical knowledge and findings to diverse, multidisciplinary research teams.
research areas
computational pathology · radiology · computer vision · cancer genomics · machine learning · survival analysis