Data Scientist, Nference
- Worked primarily on developing deep learning based solutions to solve challenges in the field of medical imaging, specifically histopathology.
- Designed a segmentation network to localise varied‑stained tissue images (IHC, H&E) at 1x resolution with AUC value of 0.95.
- Engineered a deep learning based solution to segment kidney whole slide images into constituent entities, namely glomeruli, tubules and vessel, with mean average precision value of 0.7, as a part of kidney diagnosis pipeline