CLASSIFICATION METHOD OF PROSTATE CANCER USING SUPPORT VECTOR MACHINE

摘要

The present invention is a method for classifying prostate cancer performed by a medical image processing system having a processor and a memory. The histopathological image is stained using hematoxylin and eosin to extract a region of interest for a given histopathological image data. H&E staining step, segmenting the region of interest to segment the image to identify the stroma, lumen, and cell nucleus (Nuclei) in the stained image, Fisher coefficient (Fisher coefficient) for classification of prostate cancer coefficient) and one-way analysis of variance (SVM) for extracting morphological characteristics from cell nucleus and intraspace segmentation images, and using a support vector machine (SVM) for the extracted morphological characteristics to predict and classify the Gleason grade of prostate cancer.