Breast Cancer Diagnosis by Images Based on Deep Learning: Current State and Advancements

Breast cancer is the most common type of cancer that affects women, and no complete cure method has been discovered yet. With the rapid development of Artificial Intelligence (AI), Deep Learning (DL) techniques have proved to be effectively applicable in breast cancer detection. Instead of the risk of misdiagnosis, DL offers an effective solution for standardized diagnosis. This study presents a systematic literature review on DL based methods for breast cancer diagnosis from different images, which includes Mammography (MG), Magnetic Resonance Imaging (MRI), Ultrasonography (US), and Digital Breast Tomosynthesis (DBT). Furthermore, a multimodal approach may play a role in breast cancer screening efficiency. Meanwhile, comparative analysis and discussion of the advantages and restrictions of the abovementioned image types for breast cancer detection and classification are also investigated. By classifying and examining the DL techniques applied to different medical images and synthesizing the recent advances and trends, this narrative review aims to provide comprehensive and up-to-date views for researchers seeking to apply DL to breast cancer diagnosis.