Mammogram Analysis for Breast Cancer Detection Using Deep Learning: A Review

Abstract

Breast cancer remains one of the most widespread and fastest growing diseases worldwide, particularly among women. Early detection is critical for effective management and improved patient outcomes. This review provides a comprehensive examination of deep learning techniques applied to breast cancer diagnosis such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and hybrid architectures that leverage transfer learning. The performance of these models is analyzed across multiple datasets and imaging modalities, along with a critical assessment of their strengths and limitations. The study broadly explores imaging modalities such as mammography, ultrasound, magnetic resonance imaging, and histology, emphasizing their roles in breast cancer detection. Furthermore, the paper discusses the importance of large-scale and diverse datasets in training robust deep learning models, underscoring their importance in achieving generalizable results. Finally, it highlights the transformative potential of deep learning in improving diagnostic accuracy and identifies future research directions to advance this rapidly evolving field.