Computer assisted surgery and therapy (CAST) for breast masses based on mammography

Breast cancer represents the most frequently diagnosed cancer in women. Mammography is the most commonly method for early detection of masses related to breast cancer. Correlation of information from multiple-view mammograms improves the performance of diagnosis by radiologists or by computer assisted systems. Detecting the location of masses accurately is highly important for radiologist to classify masses, for the surgeon to help in ease the procedure of an accurate surgery and in radio therapy process for less and efficient dose. In this paper, CAST is developed to accurately locate abnormal masses by quarter and clock segments. CAST will replace the manual localization method. The nipple is used as a reference point in localization process. A new simple nipple detection method is also proposed. The proposed CAST is examined on a new local database which was tested and classified by experts. The methodology achieved a sensitivity of 96% and specificity 73.33%.