Image features extraction for masses classification in mammograms

Computer aided diagnosis of breast cancer is becoming increasingly a necessity given the exponential growth of performed mammograms. In particular, the breast mass diagnosis and classification arouse nowadays a great interest. Indeed, the complexity of processed mass shapes and the difficulty to distinguish between them require the use of appropriate descriptors. In this paper, suitable characterization methods for breast pathologies are proposed and the study of different classification methods is addressed. In order to analyze the mass shapes, a segmentation is performed manually. Once the images are segmented, a study of various descriptors proposed in the literature is conducted. In order to compare different approaches of characterization, a comparative study is performed. The descriptors commonly used in the breast cancer field are compared to test their ability to characterize the breast lesions. Obtained results show that statistical approaches of texture provides the best classification result.