Artificial intelligence in breast cancer detection: Advances across imaging modalities and clinical integration

Abstract

Breast cancer remains a major global health challenge, where timely and accurate detection is critical for improving survival outcomes. Recent advances in artificial intelligence (AI), encompassing machine learning and deep learning techniques, have transformed breast cancer screening and diagnosis by enhancing image interpretation across multiple imaging modalities, including mammography, ultrasound, magnetic resonance imaging (MRI), and thermography. AI-driven models particularly convolutional neural networks, hybrid radiomics-deep learning frameworks, and transfer learning architectures enable automated feature extraction, precise lesion detection, segmentation, and classification, often achieving diagnostic performance comparable to or exceeding expert radiologists. The incorporation of explainable AI approaches, such as saliency maps and attention mechanisms, further improves transparency, interpretability, and clinical trust. Beyond detection, AI supports risk stratification, workflow prioritization, biomarker prediction, and personalized clinical decision-making, addressing challenges of inter-observer variability and limited expert availability, especially in resource-constrained settings. Although issues related to data heterogeneity, standardization, and regulatory validation persist, growing clinical integration and robust evidence highlight AI’s potential to complement conventional imaging, optimize screening programs, and advance precision medicine in breast cancer care.