A fully automated scheme for breast density estimation and asymmetry detection of mammograms

This paper presents a fully automated scheme for breast density estimation and asymmetry detection on mammographic images. Image preprocessing and segmentation techniques are first applied to the image, in order to extract the features for the breast density categorization. Also a new fractal-related feature is proposed for the classification. The classification to 3 classes is realized according to classification and regression trees (CARTs). The same segmentation result is used to extract a set of new statistical features for each breast; the difference of these feature values, between the two images of each pair of mammograms, are estimated and the asymmetric pairs are detected according to a modified version of k-nearest neighbor classifier. This composite method has been implemented and applied to miniMIAS database, consisting of 322 mediolateral oblique (MLO) view mammograms, obtained via a digitization procedure. The results are very promising, showing equal or higher success rates compared to other related algorithms in the literature, despite the fact that some of them use only small portions of the specific database. In contrast our methodology is applied to the complete datatabase.