Breast cancer remains a leading global health concern, with early, accurate diagnosis through mammography being critical for effective treatment. The emergence of artificial intelligence (AI) has revolutionized breast cancer screening, yet the opacity of “black box” models in clinical applications has sparked pressing calls for greater transparency. Explainable AI (XAI) offers essential solutions by making model decisions interpretable, enabling clinicians to trust and adopt advanced algorithms more confidently. This review synthesizes the current landscape of XAI methods applied to mammographic imaging, examining cutting-edge techniques such as Grad-CAM, LIME, SHAP, attention mechanisms, and prototype-based models. We analyze how these approaches provide meaningful visual and textual explanations that bridge the gap between technical innovation and clinical utility. Unique to this survey is its focus on practical case studies, integration pathways, and challenges in real-world implementation, from balancing interpretability and diagnostic accuracy to the urgent need for robust, diverse datasets. As demand grows for ethical, transparent AI in medicine, our review highlights actionable strategies, future directions, and the collaborative role of radiologists, AI specialists, and patients. By connecting technical advances to clinical trust and patient-centered care, this work sets the foundation for safe, transparent breast cancer diagnosis and aims to inspire further progress throughout the field.
Hashim et al. (Mon,) studied this question.