Emerging applications of artificial intelligence for risk stratification in head and neck cancer: a scoping review.

Introduction: Head and neck cancer represents a major clinical challenge due to its pronounced biological, histopathological and anatomical heterogeneity, which limits the predictive accuracy of conventional staging systems. To optimize diagnostic and therapeutic decision-making, reduce overtreatment and advance towards more precise oncology, robust risk stratification is essential. In recent years, artificial intelligence (AI) has emerged as a promising tool to support these processes through advanced analysis of clinical, radiological and histopathological data. Objective: To identify and describe the available scientific evidence on emerging applications of AI in risk stratification for head and neck cancer. Methods: A scoping review was conducted in accordance with the methodological guidance of the Joanna Briggs Institute and the recommendations of the PRISMA-ScR checklist. Studies published between January 2015 and January 2026, in English or Spanish, were identified through systematic searches of PubMed, Scopus, Web of Science and IEEE Xplore, supplemented by manual reference screening. Study selection, data extraction and evidence synthesis were performed independently by two reviewers using the Population-Concept-Context (PCC) framework. Results: A total of 44 studies were included, applying AI techniques primarily to diagnostic tasks and prognostic risk stratification in head and neck cancer, including prediction of lymph node metastasis and extranodal extension. The most frequently employed approaches were machine learning models, deep learning architectures and radiomics-based methods. Commonly used data modalities included computed tomography, magnetic resonance imaging, digital histopathology and structured clinical variables. Overall, studies reported moderate to high predictive performance; however, the evidence was characterized by substantial methodological heterogeneity, a predominance of retrospective designs, limited external validation and insufficient assessment of the clinical impact of the proposed models. Conclusions: The available evidence suggests that AI has the potential to enhance risk stratification in head and neck cancer, complementing conventional clinical approaches and contributing to the development of more individualized oncology. Nevertheless, responsible clinical implementation of these technologies requires overcoming challenges related to methodological standardization, prospective multicentre validation, model interpretability, and the consideration of ethical and equity-related issues. Systematic review registration: https://osf.io/7aem4/overview.