Combining multiresolution local binary pattern texture analysis and variable selection strategy applied to computer-aided detection of breast masses on mammograms

Jae Young Choi, Dae Hoe Kim, Yong Man Ro

Abstract: In this paper, we propose new texture features, so-called multiresolution local binary pattern (LBP), for classifying breast masses and normal tissue on mammograms. The proposed texture features are able to well characterize the texture patterns of mass core region and its margin. To this end, two individual LBP patterns are extracted from the core region and the ribbon region of pixels of a given mass region, respectively. Further, in order to improve classification accuracy, SVM-RFE based variable selection solution is applied for selecting an optimal subset of variables of multiresolution LBP texture features. Extensive and comparative experiments have been conducted to evaluate our multiresolution LBP features in conjunction with SVM-RFE based variable selection on public benchmark mammogram data set. Our results demonstrate the feasibility of combining our multiresolution LBP features with variable selection strategy for classification of masses and normal tissue on mammograms.