A Breast Tumors Segmentation Algorithm for Digital Mammograms

Breast cancer is frequently occurred in women to cause death. It is a malignant tumor caused by the abnormal division and reproduction of breast duct or acini cells. Many researches have proven that an early detection of the breast cancer can lead to successful treatments to reduce the death rate of breast cancer. For detecting the early-stage breast cancer, mammography is the most popular and effective technique. Extracting breast tumors accurately from a mammogram is a kernel stage for mammography, due to it significantly influences the overall analysis accuracy and processing speed of the whole breast tumor analysis. For this reason, tumors have to be identified and segmented from the breast region in a mammogram before further analysis. This paper aims to presented an accurate and efficient algorithm of breast tumor extraction on medio-lateral oblique (MLO) mammograms. The presented algorithm utilizes modified gradient vector flow (MGVF) snake to determine the breast region from a mammogram image, then uses Otsu thresholding and multiple regression analysis to remove the pectoral muscle from the breast region, and finally applies upper outlier detection and texture complexity analysis to extract the breast tumors. The presented algorithm was employed on the digital mammograms from the Mammogram Image Analysis Society (MIAS) database. The experimental results show that the tumor extracted by the presented algorithm approximately follows that extracted by the biopsy of MIAS. Keywords— mammography, cancer, medio-lateral oblique (MLO), Mammogram Image Analysis Society (MIAS), random