Sandeep Singh

Co-developed software for polynomial non-rigid co-registration across multiple imaging modalities with ROI window feature and annotation transfer system. Successfully implemented alignment for FAF, FA, Color fundus, and OCTA enface image combinations.

Developed tool for pre/post-treatment comparison of sarcoid lesions with interactive GUI and real-time threshold adjustment for lesion segmentation. Implemented red/green tinting for direct visualization of lesion changes.

Developed measurement tool for quantifying premacular bursa dimensions on OCT slices with scaling solutions for accurate measurements. Implemented user-friendly interface for research use.

Created custom software to generate enface images from B-scans with alignment system to register enface images with fluorescein angiography. Developed pathology annotation system that projects markings from enface view to corresponding b-scans.

Contributed to the development of a novel risk assessment tool for cardiovascular events using ophthalmic imaging. Created models to generate binary vessel masks and artery/vein segmentation from fundus images.

Developed three-tier prediction system from fundus images: exact calcium score prediction, categorical classification (Low/Medium/High), and binary classification (Zero/Non-zero).

Integrated SAM2 model into custom GUI for automatic surgical instrument segmentation. Implemented X-Y plotting system to evaluate surgical movement efficiency with visualization tools for analysis.

Developing software to predict O2 concentration in blood vessels using color channel analysis. Created specialized algorithms for vascular oxygen level estimation with visualization tools for research presentation.

Created versatile model for accurate binary mask generation across various ophthalmic imaging modalities. Implemented adaptive algorithms to handle different image types and qualities.