Artificial Intelligence (AI) for Breast Cancer Detection: Trends, Challenges, and Future Directions

Abstract

In 2020, breast cancer accounted for nearly 700,000 deaths globally, establishing its primacy among malignancies. AI has become a powerful tool for enhancing the early detection and diagnosis of breast cancer, improving accuracy, efficiency, and clinical accessibility. This chapter elucidates the transformative potential of AI in breast cancer detection, employing advanced algorithms to enhance diagnostic accuracy across mammography, tomosynthesis, ultrasound, and magnetic resonance imaging. Given mammography’s established role in reducing mortality for women aged 40 to 74, AI amplifies this impact, prompting a critical inquiry: might it redefine survival outcomes. Prospective trials affirm their efficacy, demonstrating elevated cancer detection rates and substantial reductions in radiologist workload. AI-assisted detection, including mammography, ultrasound, magnetic resonance imaging (MRI), digital breast tomosynthesis (DBT), elastography, liquid biopsy, and genomic profiling, falters in dense breast tissue, data scarcity impedes progress, and the opacity of AI decisions challenges clinical adoption. Innovations such as transparent AI, privacy-preserving federated learning, and the integration of imaging with genomic data herald a paradigm shift toward precision oncology. Overall, this chapter examines the confluence of technological advancement, ethical considerations, and regulatory frameworks, charting a course toward a future where AI mitigates breast cancer’s toll. By synthesizing current research and exploring future directions, this chapter seeks to inform and support researchers, clinicians, and policy makers in the development and implementation of AI-based approaches for breast cancer detection, particularly those driven by deep learning (DL) and convolutional neural networks (CNN), in pursuit of improved diagnostic accuracy.

Ähnliche Arbeiten