V3I7P75

Artificial Intelligence-Driven Breast Cancer Detection: Current Progress, Challenges, and Future Directions

Tarun Kumar Das^1*, Sreshtha Karmakar²,  Sanchayan Khajanchi³, Aniket Dutta⁴

Abstract

Breast cancer remains one of the leading cancer-related causes of fatalities among women globally. Consequently, the improvement of survival rates relies heavily on the advancement of early and precise detection. Studies suggest that breast cancer detection rates will rise sharply in the coming decades (1990-2026). For instance, it is anticipated that by 2050 there will be 3.2-3.5 million new cases per annum, up from 2.3 million reported in 2022. As part of induced cancer research, new deep learning and machine learning algorithms enabled Artificial Intelligence (AI) to enhance breast cancer diagnosis. AI can improve detection in imaging techniques such as mammography, ultrasound, and MRI. The authors of the review provide a comprehensive insight into the AI-based detection systems of today. Advances in convolutional neural networks, computer-aided diagnosis, and predictions are discussed. The authors also consider the problems that have been said to have previously restricted the adoption of such technologies in the clinic. These include algorithmic inequity, data heterogeneity, the interpretability challenge, and regulatory issues. The authors explore explainable AI, the fusion of multimodal data, and trustworthy and fair validation for future breast cancer detection tools that will augment radiologists’ practice as an additional tool.

Keywords:

Breast cancer; artificial intelligence; mammography; deep learning, medical imaging, biopsy, machine learning.