Machine Learning for Digital Pathology

Histopathology is the study and examination of tissue slides under magnification and is the definitive diagnosis for many diseases including cancer. With the advent of whole-slide scanners and image management software systems, computational pathology tools can be created to measure disease in an efficient and objective manner. This is in contrast to the labourious and subjective manual analysis approaches. In Canada, breast cancer is the 2nd cause of cancer death. With digital pathology, image analysis solutions can be developed to efficiently summarize a large number of breast cancer pathology images with quantitative, reliable measures of disease which may be correlated with other clinical variables to further understanding of progression, etiology and therapeutic response. To score and grade the images, it is important to detect the nuclei and measure various nuclear properties. TO BE CONT'D

Ahlad Kumar
Faculty Supervisor: 
April Khademi
Partner University: