Applications of Artificial Intelligence (AI) in Breast Cancer Care Delivery and Education: A Scoping Review

Abstract

Artificial intelligence (AI) is increasingly being applied in breast cancer care, yet its use across the post-diagnosis phase remains poorly mapped. This scoping review aimed to identify and categorise AI applications in post-diagnosis breast cancer care, encompassing treatment planning, treatment delivery, follow-up and surveillance, survivorship, and palliative care. Following JBI methodology and PRISMA-ScR reporting guidelines, four databases (MEDLINE, EMBASE, CINAHL, and Web of Science) were searched, identifying 3784 records. After screening and full-text assessment, 54 studies published between 2016 and 2024 were included. Machine learning was the predominant technology (81%), followed by generative AI (7%), conversational agents (6%), traditional natural language processing (4%), and data mining (2%). Follow-up and surveillance were the most represented care stage (48%), driven primarily by recurrence prediction models. Most applications were provider-focused (83%), while patient-facing tools accounted for 17% of studies and relied on either conversational agents or generative AI. No studies addressed palliative care. The evidence base was predominantly retrospective (70%) and concentrated in high-income countries (74%). Future research should prioritise prospective evaluation in clinical workflows, address unsupervised patient use of generative AI, and ensure equitable development across diverse populations and care settings.

Ähnliche Arbeiten