INTELLIGENT APPROACHES FOR BREAST CANCER DIAGNOSIS: A SYSTEMATIC LITERATURE REVIEW

Abstract

Breast cancer remains a major global health challenge, making early and accurate diagnosis essential for improving clinical outcomes. The rapid development of artificial intelligence, particularly deep learning, has enabled increasingly effective analysis of breast medical images across multiple imaging modalities. This study presents a systematic literature review of deep learning-based approaches for breast cancer detection and classification, focusing on research published between 2020 and 2026. The review examines commonly used deep learning architectures, datasets, imaging modalities, evaluation metrics, and explainable artificial intelligence (XAI) techniques. The findings indicate that Convolutional Neural Networks (CNNs) remain the dominant approach, while Vision Transformers, hybrid architectures, and multimodal learning are emerging as important directions for improving diagnostic performance and robustness. Commonly investigated datasets include DDSM, CBIS-DDSM, MIAS, BreaKHis, and BACH across mammography, histopathology, ultrasound, and MRI. XAI methods such as Grad-CAM, SHAP, and LIME are increasingly employed to improve model transparency and clinical interpretability. Despite substantial progress, challenges remain in dataset diversity, generalization, computational complexity, external validation, and clinical translation. The review identifies these gaps and outlines future research directions toward more reliable, interpretable, and clinically applicable deep learning systems for breast cancer diagnosis.