EARLY STAGE DETECTION OF BREAST CANCER USING FEATURE EXTRACTION TECHNIQUES

Abstract

This paper presents an effective and efficient approach for early detection of breast cancer in digital mammogram images, in which masses and microcalcifications appear in the form of clusters with high intensity compared to their neighborhood pixels. The presence of these clusters is considered as a prominant indicator of malignant and benign types of breast cancer and the detection of this is used to treat/prevent the disease at an early stage. In order to implement this, we propose a new algorithm that uses image enhancement by wavelet transform and adaptive histogram equalization technique followed by Segmentation with Border extraction. We compared the proposed algorithm with algorithms that used conventional watershed segmentation for detecting the masses and microcalcification efficiently.