Breast Cancer Detection and Prognosis using Mammographic and Genomic Data: A Comprehensive Review of Radiogenomics, Multimodal Deep Learning, and Explainable Artificial Intelligence

Gajendra Kumar Sharma, Dr. Anil Kumar

Breast Cancer, Mammography, Genomic Data, Radiogenomics, Radiomics, Multimodal Deep Learning, Explainable Artificial Intelligence, Medical Imaging, Molecular Profiling, Prognosis Prediction, Multimodal Data Fusion, Precision Oncology.

Breast cancer is a complex disease in which detection, molecular profiling, treatment selection, and outcome prediction are challenging. Mammography can provide structural and morphological information to identify abnormal breast lesions, while genomic information can reveal molecular alterations, gene expression profiles, tumour subtypes, and disease stages. The integration of these complementary modalities has paved the way for innovative radiogenomics and multimodal AI solutions in breast cancer (BC). Here, we review recent progress in the combined use of mammographic and genomic data for breast cancer detection and prognosis prediction, focusing on radiomic feature extraction, genomic representation, multimodal data fusion, deep learning (DL), transformer-based architectures, and explainable artificial intelligence (XAI). Fusion strategies for incorporating imaging and genomic data at the feature level, decision level, and attention-based level are presented and evaluated. The review covers traditional datasets, evaluation metrics, data heterogeneity, missing modalities, the limited number of matched imaging-genomic cohorts, model generalizability, and challenges in model interpretability. In addition, we discuss how explainable approaches, such as Grad-CAM and SHAP, can reveal clinically relevant imaging regions and genomic features. Finally, promising avenues, including cross-modal learning, multi-omics data integration, self-supervised learning, domain adaptation, and clinically interpretable multimodal models, are discussed. This review presents an overview of the integration of mammographic and genomic data within a precision breast cancer framework.

Received: 24 Dec 2025, Received in revised form: 21 Jan 2026, Accepted: 25 Jan 2026, Available online: 30 Jan 2026

10.22161/ijaers.131.6