Multi-feature prostate cancer diagnosis of histological images using advanced image segmentation

摘要

We present a study of image features for cancer diagnosis of the histological images of prostate. In diagnosis, the tissue image is classified into the tumour and non-tumour classes. In Gleason grading, which characterises tumor aggressiveness, the image is classified as containing a low- or high-grade tumour. The primary contribution of this paper is to aggregate colour and texture properties at histological object levels for classification. Features representing different visual cues were combined in a supervised learning framework. We also compare the performance of Gaussian, k-nearest neighbour, and Bayesian classifier.