The graph cuts technique, breast density and abnormality detection

The presence of density part of breast tissues is one of the signs considered by radiologists to determine whether a suspicious area is a tumor or cancer. The focus of this research is for the segmentation of dense area with regards to the breast anatomical structure for visualization, such as location of nipple if possible, skin-air interface, fatty tissue, glandular tissue and pectoral muscle. In this study, graph cuts are proposed for segmentation of density area and other anatomical regions. Usually, mass is represented by hyperdense structure. Overlapped fibroglandular tissue also has similar intensity with mass. By focusing on the glandular area and highlighting the hyperdense region of the glandular area, the graph cuts algorithm can automatically detect the abnormal area.