Hyperchromatic Nucleus Segmentation on Breast Histopathological Images for Mitosis Detection

Abstract

Breast cancer grading is the standard clinical practice for the prognosis and diagnosis of breast cancer development. The Nottingham Histological Grading (NHG) system is widely used in the breast cancer grading. In NHG system, the mitotic count based on histopathological images (i.e. microscopic slide examination) is one of the three criteria that define the overall grade. Image processing techniques such as segmentation could be utilised to detect mitotic cells. This study proposed a new approach to segment hyperchromatic nucleus on the histopathological images based on RGB and HSI colour spaces. The results show that the proposed segmentation technique could provide a promising result in segmenting hyperchromatic nucleus and preserving the ground truth (i.e. true mitotic cells).