Abstract
Breast cancer is a critical disease that affects a large number of females around the world. Early detection of breast cancer is an effective step for increasing the rate of survival. There are several computerized systems used for breast cancer classification and diagnosis. However, these systems are still required a considerable improvement to be more effective and accurate tools. In this paper, we develop an intelligent decision support system (IDSS) for breast cancer diagnosis. The developed IDSS consists of four main stages are preprocessing, segmentation, feature extraction, and classification. In the preprocessing stage, we processed the breast images to eliminate the noise and artefacts. In the segmentation stage, the region of interests (ROIs) are segmented from the mammogram images using K-means algorithm. After that, the discriminative features are extracted from the ROIs through the feature extraction stage using the discrete wavelet transform (DWT) and gray level co-occurrence matrix (GLCM) method. In the classification stage, the extracted features of breast tumor are classified into three classes, which are normal, malignant, or benign, by using the artificial neural network (ANN). The public dataset collected by the Mammographic Image Analysis Society (MIAS) is used for evaluation. The experimental results demonstrated that the IDSS is able to achieve 96.563% of average accuracy using 10-folds cross-validation technique.