Diagnosing of Malignant and Benign Breast Cancers: Using a Metasearch Neural Propagation Classifier

Abstract

One of the world’s most significant health issues is breast cancer. There is a maximal probability of survival if anomalies in breast cancer are diagnosed early. For such an early prediction, we could choose mammography. Mammography is among the more effective approaches to screening and identification of breast cancer. The classification of medical data is an easy way to diagnose the disease using mammography images. Many authors have experimented with using different classifiers in the area of Mammography classification. But the machine learning classifiers cannot yield satisfactory accuracy. We have proposed a Meta Search Neural Propagation Classifier (MNPC) to detect breast cancer. The mammogram image is first segmented using the Otsu method, and then the features are extracted by utilising Discrete Wavelet Transform (DWT). The extracted features are sent to the Meta Search Neural Propagation Classifier, which categorises the mammogram’s image malignant or benign. The analysis of the images gives that for detecting the malignant, and benign; the proposed method is accurate and reliable and the specificity, sensitivity, F-score, accuracy, precision, and the classifier’s efficiency are enhanced.