Gleason Pattern Scoring
Cognex Deep Learning classifies cell morphologies
Graphical programming environment for deep learning-based industrial image analysis
The Gleason pattern grading scale classifies prostate cancer cells by their glandular structure on a scale from 1–5. Gleason patterns range in their degree of differentiation, from small and uniform to highly irregular and distinct. Scoring a Gleason pattern can be challenging when there are multiple scenes – in this case cells – in a single slide.
The Cognex Deep Learning classification tool categorizes different cell morphologies like the Gleason pattern based on a collection of labeled images. To train the location tool, a technician or engineer provides all images labeled in accordance with the different classes. Cognex Deep Learning develops a model of Gleason patterns 1–5, then classifies cell tissues by their degree of differentiation and sorts them by grade.