An Unsupervised Method for Suspicious Regions Detection in Mammogram Images

摘要

Over the past years many researchers proposed biomedical imaging methods for computer-aided detectionudand classification of suspicious regions in mammograms. Mammogram interpretation is performed byudradiologists by visual inspection. The large volume of mammograms to be analyzed makes such readingsudlabour intensive and often inaccurate. For this purpose, in this paper we propose a new unsupervised methodudto automatically detect suspicious regions in mammogram images. The method consists mainly of twoudsteps: preprocessing; feature extraction and selection. Preprocessing steps allow to separate backgroundudregion from the breast profile region. In greater detail, gray levels mapping transform and histogramudspecifications are used to enhance the visual representation of mammogram details. Then, local keypointsudand descriptors such as SURF have been extracted in breast profile region. The extracted keypoints areudfiltered by proper parameters tuning to detect suspicious regions. The results, in terms of sensitivity andudconfidence interval are very encouraging.