Automated Breast Cancer Digital Pathology Image Scoring using AI

Ki-67 is a promising biomarker in breast cancer and the Ki-67 labeling index, or the percentage of Ki-67-positive cells, has great prognostic potential particularly in carcinomas of the breast. To overcome the challenges of manual calculation of ki67 proliferation index, in this project we will design and develop computational pathology and artificial intelligence (AI) algorithms to automate Ki67 quantification for tissue microarrays (TMA) and wholeslide images (WSI). AI can provide objective and efficient Ki67 scoring which would allow for more accurate diagnosis and turn-around times, which improves quality of care. Further, automation can be used to develop guidelines and standards for Ki67 scoring which will accelerate wide-scale clinical adoption.

Faculty Supervisor:

April Khademi

Student:

Partner:

Pathcore Inc

Discipline:

Engineering

Sector:

Health and Related Sciences & Technology; Information and cultural industries

University:

Toronto Metropolitan University

Program: