Breast tumors extraction and features detection in breast magnetic resonance images using clustering and image processing algorithms

This research aims to developing new method for breast tumors extraction and features detection in breast magnetic resonance images by depending on clusteringand image processing algorithms. At the beginning, one of clustering algorithms was used for image segmentation and grouping pixels by their gray scale values. Then morphological operations were implemented in order to remove noise and undesired regions, after that suspected areas were extracted. Finally some shape features for extracted area were detected, this features could be very useful for tumors diagnosis. A database consisted of 96breast magnetic resonance images were used and proposed approach was appliedby MATLAB program, and we obtainedbreast tumors extraction and its features and compared them with the doctor's opinion .