Region Based Approach for Pectoral Muscle Isolation

The pectoral muscle, a non-breast region in mammograms, acts like an additional complexity in automated analysis using Computer Aided Diagnosis (CAD) systems. The pectoral muscle represents a predominant density region in most medio-lateral oblique (MLO) views of mammograms and can affect the results of image processing of detection of breast abnormalities. Intensity-based methods can present poor results when applied to differentiate dense structures such as suspicious masses or fibro-glandular discs because the pectoral muscle has a similar opacity as tumours. Another important reason for detecting the pectoral muscle in mammograms is that CAD system can detect the possible presence of abnormal axillary lymph nodes which is a major sign of breast carcinoma. Two distinct approaches in pectoral muscle isolation are used. These are Region Based and another is edge based. In Region based approach consecutive steps are used to isolate and suppress the pectoral muscle. The initial step involves defining a rectangle to isolate the pectoral muscle from the region of interest (ROI) and the next is to suppress the pectoral muscle using proposed modified seeded region growing (SRG) algorithm. The quantitative and qualitative measures on breast images are calculated for showing better performance than existing methods. This paper actually proposed Region based approach for pectoral muscle isolation.