Nowadays, the breast cancer can be detected early with automated Computer Aided Diagnosis (CAD) system the best available technique to assist radiologist. For developing such an efficient computer-aided diagnosis system it is necessary to pre-process the mammogram images. Hence, this paper proposes a method for effective pre-processing of mammogram images. This method consists of two phases such as (i) Breast Region Extraction and (ii) Pectoral removal. In first phase, Adaptive Local Thresholding is used to binaries the image followed by morphological operations for removing labels and artifacts. Then the breast region is extracted by identifying and retaining the largest connected component of mammogram. The pectoral muscle which is the predominant density region of mammogram that should not carry any useful information and also affects the diagnosis is to be removed in phase two. A new method called Pixel Constancy Constraint at multi-resolution approach is introduced for pectoral removal. The proposed method is experimented with Mini-MIAS database (Mammographic Image Analysis Society, London, U.K.) and yields a promising result when compared with existing approaches.
CITATION STYLE
Vidivelli, S., & Sathiya Devi, S. (2016). Breast region extraction and pectoral removal by pixel constancy constraint approach in mammograms. In Advances in Intelligent Systems and Computing (Vol. 412, pp. 195–206). Springer Verlag. https://doi.org/10.1007/978-981-10-0251-9_20
Mendeley helps you to discover research relevant for your work.