Abhijna Shridhara Hebbar

The thesis titled “Explore and enhance quality control in digital pathology workflow using deep learning solutions” addresses the challenge of limited generalization of QC tools across staining modalities through novel TL approaches that can learn modality-invariant representations to optimize a model on both HE and IHC staining.

• Performed quality control of histopathological tissue section scans. • Designed and integrated a custom AutoML framework for small tabular datasets, incorporating AutoGluon and TabPFN benchmark solutions alongside novel feature reduction modules. • Engineered Python scripts for clinical trial data extraction, folder/file structure validation, discrepancy reporting, and automated SharePoint tag/folder creation for Clinical Data Management.