Developing a Remote Search Engine for Histopathology Scans

The wide-spread adoption of digital pathology has opened new horizons for histopathology. Artificial intelligence (AI) algorithms are able to operate on digitized slides to assist pathologists with different tasks. Whereas AI involving classification and segmentation methods have obvious benefits for image analysis, image search represents a fundamental shift in computational pathology. Image-based search for digitized pathology slides can provide pathologists with unprecedented access to the evidence embodied in already diagnosed and treated cases from the past. Huron Digital Pathology has developed and designed an image search system for histopathology images, called Yottixel. The proposed project is the enhancement of Yoittixel search algorithm by incorporating a novel approximate k-NN technique. The proposed method could enable up-to 10-fold speedup in searching and offer more efficient utilization of computational resources. These enhancements provide competitive advantage to Yottixel as a product, especially in hospitals and labs where computing infrastructure are not sufficient for hosting a sophisticated image search engine on-site. The ultimate goal of the research is to develop Yottixel into a robust and efficient search engine for digital histopathology archives. It has potential to be a screening tool to both speed up and improve the accuracy of cancer diagnoses by pathologists.

Faculty Supervisor:

Hamid R Tizhoosh

Student:

Shivam Kalra

Partner:

Huron Digital Pathology

Discipline:

Engineering

Sector:

University:

University of Waterloo

Program: