eSSENCE@LU 5:9 - Automatic classification of microscopy images in digital pathology using deep learning: Automatic Gleason scoring of prostate images

Projekt: Forskning


The project aims at developing automatic diagnosis methods for Gleason scoring of microscopic images of prostate cancer based on deep convolutional neural networks (CNN). We propose to improve the performance of these networks by developing rotational invariant CNN:s, clever usage of autoencoders, introducing intermediate segmentation steps (also based on CNN) and by incorporating not only gold standard as training data but also ground truth obtained from other biomarkers. We will also investigate new stain-separation methods that is necessary in order to produce a classification that is independent of variations in the staining process.
Kort titeleSSENCE 5:9
Gällande start-/slutdatum2018/01/012020/12/31