The computer-aided diagnosis of soft-tissue tumors using sonographic morphology and texture features

Computer-aided diagnosis (CAD) is based on shape and echogenicity of tumors and is used to extract relevant information. The various steps of the general CAD algorithm for tumor diagnosis are discussed in this chapter. A CAD system that integrates these features has been proved to be capable of assisting radiologists in the diagnosis of soft-tissue tumors on sonography. Thus, in the characterization of benign and malignant soft-tissue tumors, the CAD system successfully implements the classification using two types of feature sets, including morphological and texture features. Notably, the performance of the morphological features seems to be more relevant in the determination of the tumor status. Nevertheless, both morphological and texture features remain the most important properties while visually interpreting soft-tissue tumors.