Classification of circular and lobulated masses in mammograms using fractal analysis

Fractal analysis can be used for the detection of masses exhibiting circular and lobulated shape margins in screening mammograms. Currently, mammograms have been one of the most versatile methods for the early detection of breast cancer. However, one of the characteristic lesions of breast cancer appears as dense regions of breast sizes and properties known as breast masses. To classify these breast masses into benign and malignant, fractal analysis can be used. The most versatile method for fractal analysis is based on Power Spectral Analysis (PSA) using the signatures of contours. The experimental results using fractal analysis provides an accuracy of 82% for circular and lobulated margin breast masses.