Research Interests

Machine learning and AI in Oncology

• Integration of machine learning and optimal transport 

• Machine learning-based outcomes modeling using genomic, radiomic, and radiation dose-volume data

• Prediction of Oncotype-DX score in breast cancer using optimal transport and deep learning 

Bioinformatics

• Differential sex effects of pregnancy-specific glycoproteins (PSGs) on survival in lung cancer

• Breast cancer subtyping using multi-omic data including DNA mutation, RNA-Seq, CNA, and histology images

Radiogenomics 

• Machine learning-based modeling to predict radiation-induced toxicity using GWAS

  • GI and GU toxicity in prostate cancer

  • Risk of developing secondary breast cancer

Radiomics

• Network-based unsupervised modeling using radiomic data

  • Survival prediction in lung and head & neck cancers

• Correlation between radiomic features and immune cell abundance (CIBERSORT)

• Outcomes modeling using deep features derived from CNN

Radiation biology

• Understanding of the difference in biological features between photon and proton irradiation using scRNA-Seq data