Detection and classification of normal and abnormal patterns in mammograms using deep neural network

摘要

Breast cancer detection is the most challenging aspect in the field of health monitoring system.In this paper, breast cancer detection was assessed by employing Mammographic ImageAnalysis Society (MIAS) dataset. The proposed approach contains four major steps, namely,image-preprocessing, segmentation, feature extraction, and classification. Initially, Laplacianfiltering was utilized to identify the area of edges in mammogram images and, also, it was verysensitive to noise. Then, segmentation was carried-out using modified-Adaptively RegularizedKernel-based Fuzzy-C-Means (ARKFCM); it was a flexible high level machine learning techniqueto localize the object in complex template. In conventional ARKFCM, it was hard to segmentthe ill-defined masses in mammogram images. To address this concern, the Euclidean distancein ARKFCM was replaced by correlation function in order to improve the segmentationefficiency. The hybrid feature extraction (Histogram of Oriented Gradients (HOG), homogeneity,and energy) was performed on the segmented cancer region to extract feature subsets. Therespective feature values were given as the input for a multi-objective classifier: Deep NeuralNetwork (DNN) for classifying the normal and abnormal regions in mammogram images. Theexperimental outcome shows that the proposed methodology improved accuracy in breastcancer classification up to 3% to 9% compared to other existing methods.