Breast cancer detection without removal pectoral muscle by extraction turn counts feature

Abstract

During late decade breast cancer is recognized as major cause of death among women and the number of breast cancer patients is increasing. There is more evidence that women in 15–54 years old are died by breast cancer. Breast cancer cannot be prevented because its major factors have not been identified. Therefore earlier diagnosis can increase the possibility of improvement. The aim of this study was to extract the feature without removing pectoral muscle in preprocessing stage using a new and efficient method. Database of MIAS mammography images was used to classify normal/ abnormal individuals and benign/ malignant cancer patients and the results of support vector machine classifier showed accuracy of 95.80 and 86.50 respectively.