Show abstract - Show schedule - Proceedings - PDF - Reviews

Gleason grading is a risk stratification procedure for prostate cancer that is subjective and based on the reporting pathologist's experience and skill. Deep Learning (DL) algorithms have showed potential in improving Gleason grading objectivity and efficiency. On Whole Slide Images (WSI) from a source other than training data, however, DL networks show domain shift and poor performance. Using a novel training process that learns domain agnostic features, we propose a DL approach for segmenting and grading epithelial tissue. When utilised as an aid for core needle biopsy (CNB) evaluation, our DL approach has the potential to increase grading consistency and accuracy, leading in better patient outcomes.
Hide abstract

Wednesday 6th July
Poster Session 1.1 - onsite 15:20 - 16:20, virtual 11:00 - 12:00 (UTC+2)
Hide schedule