Possibilistic Fuzzy C Means Algorithm For Mass classificaion In Digital Mammogram

Mammography is an effective imaging modality of breast cancer abnormalities detection. Survival rate of breast cancer treatment can be increased via early detection of mammography. However detecting the mass in the early stage is a tough task for radiologist. Detection of suspicious abnormalities is a continual task. Out of thousand cases only 3 to 4 are analyzed as cancerous by a radiologist and thus abnormality may be left out. 10-30% of cancers are failed to detect by radiologist. Computer Aided Diagnosis helps the radiologists to detect abnormalities earlier than traditional procedures. Because of some negligence in capturing device, the image may be affected by noise this leads to fault diagnosis. Preprocessing can remove this unwanted noise. In this paper features such as entropy, circularity, edge detection, and correlation are extracted from the image to distinguish normal and abnormal regions of a mammogram. Classification and detection of mammogram can be done by Possibilistic Fuzzy C Means algorithm and Support Vector Machine using extracted features. KeywordsMammogram, Mass, Preprocessing, Feature extraction, Possibilistic Fuzzy C Means, Support