Feature Analysis And Extraction For Detecting The Breast Abnormalities In Digital Mammograms

Abstract

This research article introduces a completely unique approach for accomplishing digital mammography feature analysis and extraction through detection of abnormal masses regarding their size, density, shape, texture, color, topology with experimental work for early breast abnormality detection. The objective is to detect the abnormal masses or tissue within breast tissues using three essential stages: Preprocessing, Segmentation and post processing stage. In preprocessing stage unwanted noise is removed and then segmentation is applied to notice the abnormal mass, subsequently post process is applied to find out the normal and abnormal tissue with the affected space within the digital mammogram breast images. In this paper, we are calculating texture, statistical and structural features. The statistical performance measures namely Sensitivity, Specificity, and F-Score are measured against the native data set to figure out the performance of the system.