Revolutionizing stroke prediction: a systematic review of AI-powered wearable technologies for early detection of stroke.

Wearable technology, combined with artificial intelligence (AI) and machine learning (ML) algorithms, opens up new frontiers for continuously monitoring physiological or behavioural data, allowing the identification of stroke risk factors at an earlier stage. A systematic search was performed in PubMed, IEEE Xplore, Scopus, Google Scholar, Cochrane Library, and Web of Science, following the PRISMA guidelines. The search aimed to include studies using wearable devices incorporating AI/ML models for real-time stroke prediction. The reviewed studies were characterized according to methodology, population characteristics, device specifications, specific AI/ML models employed, outcome measures such as predictive accuracy, sensitivity, and specificity, and their main findings. In our review, we identified 5 studies that met our inclusion criteria.The review also finds that AI-enhanced wearables may offer the accurate prediction of stroke events, with most studies reporting high predictive performance. Sensor-equipped wearable devices to measure vital parameters, such as blood pressure and heart rate variability, when combined with AI models, can provide greater sensitivity and very good specificity in identifying early signs of diseases. However, differences in the accuracy of devices and a lack of transparency in AI algorithms imply some practical challenges. These findings highlight the potential of wearable technologies driven by AI-ML for non-invasive, real-time stroke monitoring and risk assessment, although further research will be required to optimize model reliability and device usability. This review consolidates current evidence supporting the clinical application of wearable AI/ML technology and advocates for its advancement in stroke prevention and patient care.