%0 Journal Article
%T Dual Modality: Mammogram and Ultrasound Feature Level Fusion for Characterization of Breast Mass
%A Minavathi
%A Dr. Murali. S.
%A Dr. Dinesh M.S.
%J International Journal of Innovative Technology and Exploring Engineering
%D 2013
%I IJITEE
%X detection of abnormalities in breast is done in different phases using different modalities and different biomedical techniques. These techniques and modalities are able to furnish morphological, metabolic and functional information of breast. Integrating these information assists in clinical decision making. But it is difficult to retrieve all these information from single modality. Multimodal techniques supply complementary information for improved therapy planning. This work concentrates on early detection of breast cancer which characterises the breast mass as malignant or benign by investigating the features retrieved from dual modalities: mammograms and Ultrasound. Architectural distortion (AD) with Spiculated mass is an important finding for the early detection of breast cancer. Such distortions can be classified as spiculation, retraction, and distortion which can be detected in mammograms. Spiculated masses carry a much higher risk of malignancy than calcifications or other types of masses. The proposed approach is based on the fusion of two modalities at feature extraction level with Z-Score Normalization technique to improve the performance of dual modality. Gabor filters are used to retrieve texture features from region of interest (ROI) of mammograms. Shape and structural features are retrieved from ROI¡¯s of Ultrasound. In addition to that some other discriminative features like denseness texture feature, standard deviation, entropy and homogeneity are also extracted from ROI¡¯s of both modalities. Feature level fusion is then achieved by using a simple concatenation rule. Finally classification is done using Support vector machine (SVM) classifiers to classify breast mass as malignant or benign. Receiver operating characteristic curves (ROC) are used to evaluate the performance. SVM classifiers achieved 95.6% sensitivity in characterising the breast masses using the features retrieved from two modalities.
%K Mammogram
%K Ultrasound
%K Spiculated mass
%K Architectural distortion
%K Dual modality
%K SVM
%K Feature level fusion
%K Z-score Normalization.
%U http://www.ijitee.org/attachments/File/v2i6/E0632032413.pdf