Survey on deep convolutional neural networks in mammography

Abstract

The limitations of traditional Computer Aided Detection (CAD) systems for mammography, the extreme importance of early detection of breast cancer and the high impact of the false diagnosis of patients lead to investigating Deep Learning methods (DL) for mammograms. Deep Learning, in particular, Convolutional Neural Networks (CNNs) have been recently used for object localization and detection, risk assessment, and classification tasks in mammogram images. CNNs help radiologists provide more accurate diagnosis by delivering precise quantitative analysis of suspicious lesions in short time. This survey reviews the strengths and limitations of major CNNs applications in analyzing mammogram images. It summarizes 51 contributions on applying CNNs on various tasks in mammography. Moreover, it discusses the best practices done using CNN methods and suggesting directions for further improvement in medical images and in particular mammogram images.