Electronic Thesis and Dissertation Repository


Master of Science




Dr. Fenster, Aaron

2nd Supervisor

Dr. Moussa, Madeleine


Prostate Cancer (PCa) is the most common non-cutaneous cancer in North American men. Multi-parametric magnatic resonance imaging (mpMRI) has the potential to be used as a non-invasive procedure to predict locations and prognosis of PCa. This study aims to examine non-cancerous pathology lesions and normal histology that could mimic cancer in mpMRI signals. This study includes 19 radical prostatectomy specimens from the London Health Science Centre (LHSC) that were marked with 10 strand-shaped fiducials per specimen which were used as landmarks in histology processing and ex vivo MRI. Initial registration between fiducials on histology and MR images was performed followed by the development of an interactive digital technique for deformable registration of in vivo to ex vivo MRI with digital histopathology images. The relationship between MRI signals and non-cancerous abnormalities that could mimic PCa has not been tested previously in correlation with digital histopathology imaging. The unregistered mp-MRI images are contoured by 4 individual radiology observers according to the Prostate Imaging Reporting and Data System (PI-RADS). Analysis of the radiology data showed prostatic intraepithelial neoplasia (PIN), atrophy and benign prostatic hyperplasia (BPH) as main non-cancerous abnormalities responsible for cancer like signals on mpMRI. This study will help increase the accuracy of detecting PCa and play a role in the diagnosis and classification of confounders that mimic cancer in MR images.

Included in

Pathology Commons