Breast Cancer Care: From Traditional Treatments to AI-Driven Risk Prediction and Future Prevention
Year: 2026
Type: article
Abstract: Breast cancer management has moved from invasive surgeries to multidisciplinary therapies that utilize imaging, molecular pathology, and precision medicine. In this survey, we discuss traditional surgeries, radiation therapy, systemic treatments, modern diagnostics, and AI techniques for mammogram assessment and future cancer risk prediction.Specifically, we consider CNNs, Vision Transformers, the BCSC clinical model, mammography AI models like Mirai, and polygenic risk scores, namely PRS313. Identification of currently present cancer on the exam differs from the estimation of future disease probability among patients who are free of cancer at baseline. According to an analysis of a cohort from 2026 consisting of 82,957 women, the ten-year C-index for a combined clinical,imaging, and genetic model was 0.70, while BCSCv3 scored 0.62, PRS313-adjusted 0.61, andmammography AI 0.66.The review includes a discussion of publicly available data, metrics, problems, and opportunities for breast cancer prevention. Theprospective validation, calibration, population diversity, privacy, and clinical governance remain vital before the routine adoption of themodels under consideration.
Language: English
Topic: AI in cancer detection
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7225995215
