Bridging the Gap Between Prostate Radiology and Pathology Through Machine Learning

About this Class

Join the June NCI Imaging and Informatics Community Webinar for a discussion on the recent contributions from Dr. Mirabela Rusu’s Personalized Integrative Medicine Laboratory (PIMed) at Stanford University. Recent laboratory contributions include:

• registering whole-mount pathology images with an MRI,
• training deep learning models to extract pathomic MRI biomarkers,
• using biomarkers in training to detect and distinguish indolent and aggressive prostate cancers, and
• showing the benefits of using labels from pathology in training deep learning models to distinguish idle vs. aggressive prostate cancer.

The PIMed Laboratory focuses on improving the interpretation of prostate MRI by developing deep learning models that automatically localize prostate cancers on MRI scans. The novelty of these methods comes from using whole-mount pathology images to label MRI images and create pathomic MRI biomarkers of cancer. This approach achieved an area under the receiver operator characteristics curve of 0.93 evaluated on a per-lesion basis, outperforming existing deep learning models. In patients outside the training cohorts, such predictive models will outline the extent of cancer on radiology images in the absence of pathology images, thus helping guide the prostate biopsy and local treatment. Speaker: Dr. Mirabela Rusu, PH.D. Dr. Rusu is an assistant professor for the department of radiology at Stanford University. She is director of the PIMed Laboratory, which has a multi-disciplinary direction focused on developing analytic methods for biomedical data integration, with a particular interest in radiology-pathology fusion to facilitate radiology image labeling. Dr. Mirabela Rusu’s laboratory focuses on improving the interpretation of prostate MRI by developing deep learning models that automatically localize indolent and aggressive prostate cancers on MRI scans. The subtle difference in MRI appearance of prostate cancer and benign prostate tissue renders the interpretation of prostate MRI challenging, causing many false positives, false negatives, and wide variations in interpretation. The talk will focus on discussing recent advances by the lab through registering whole-mount pathology images with MRI, training deep learning models to extract pathomic MRI biomarkers, and using them to detect and distinguish prostate cancers.