QC tool Development
Development of Quality Control tool for pathologists
DDP Thesis | Prof. Amit Sethi, Dept. of Electrical Engineering, IIT Bombay
- Created a robust pipeline for the detection and classification of artifacts within Whole Slide Images (WSIs), employing four specialized models for identifying blur level, tissue fold, pen marker, and tissue segmentation.
- Developed a comprehensive WSI profiling system by seamlessly integrating the results from the aforementioned four models, generating a refined and usable mask for analysis
- Demonstrated exceptional performance by achieving a dice score exceeding 0.7 on 74% of the 11,529 WSIs from the TCGA dataset when comparing our profiler’s results to the standardized HistoQC’s usable masks