Inteligencia artificial en kinesiología: aplicaciones, oportunidades y barreras: una revisión narrativa

Abstract

Artificial intelligence (AI) is entering routine kinesiology practice. This narrative review maps its current status across four dimensions of professional practice: functional assessment, therapeutic intervention, prognostic prediction, and support for documentation and decision-making. Evidence in emerging areas of the discipline, namely oncology, pelvic floor, and maxillofacial rehabilitation, is also examined and found to be incipient. Evidence published between 2022 and 2026 supports benefits in assessment objectivity, continuous monitoring and reduced documentation burden, with small effect sizes in therapeutic applications. Substantial gaps persist: insufficient external validation, algorithmic bias in underrepresented populations, lagging professional adoption, and limited interpretability of opaque models. The applicable governance framework is systematised, including World Health Organization guidance, the European Union AI Act, Chile's National Artificial Intelligence Policy and Law 21,719. We conclude that these tools operate as inputs to clinical reasoning rather than substitutes for professional judgement, and that Chile has the infrastructure and policy framework but still lacks the local scientific output needed to validate and adapt these models.