Classification of Breast Lesions Using Feature Extraction Techniques

Abstract

The most common form of cancer being diagnosed in women worldwide is breast cancer. A region endures from damage through any disease then region is known as lesion. Therefore, characterization of breast lesions is clinically significant. Therefore, there is a significant impetus among the research community to develop computer aided diagnostic (CAD) systems for differential diagnosis between different cases of breast lesions. Thus, in order to provide the radiologists with a second opinion tool for validating their diagnosis, various CAD systems have been developed in the present work for two-class breast lesions classification. For the design of this CAD system, ultrasound images are taken and for each ultrasound image, ROI is marked according to the shape of abnormality. The CAD system consists of input ultrasound images, ROI extraction module, feature extraction module and the classification module. In the feature extraction module, four methods for extracting the features are employed, (a) Morphological methods (b) Signal processing based method (c) Transform domain based methods and (d) Statistical based methods.