Computer Aided Diagnosis of Tumors Masses in Mammograms Using Texture Features

X-rays imaging is a computerized system that enhances the detail resolution of the reconstructed image of any object. Mammography (breast imaging by X-rays) is the most effective technique that is used in the diagnosis of breast cancer, where the detection of early malignant tumors is essential. Tumors in breast may appear as masses or as micro-calcification clusters. Investigators had applied various mathematical methods to detect these tumors. In this paper, fifty breast images as data base are considered. The diagnosis of masses is proposed as an application of Radon transform of the original breast images, and then the texture features of the projected images are calculated. Following a classification of the extracted features by an artificial neural network (ANN) is applied. Finally, using the receiving operating characteristics (ROC) for analyzing the obtained results, the overall sensitivity becomes better