WEARABLE SENSORS FOR REAL-TIME SEIZURE DETECTION: FROM LABORATORY VALIDATION TO CLINICAL DEPLOYMENT

Abstract

Epileptic seizures are associated with injury risk, sudden unexpected death in epilepsy (SUDEP), and profound impacts on quality of life and independence. Real-time seizure detection through wearable sensors — monitoring electrodermal activity, accelerometry, heart rate variability, and electromyography — offers the potential for timely alert generation, enabling caregiver intervention and emergency response that could reduce seizure-related morbidity and mortality. This review analyses 36 studies of wearable seizure detection systems published between 2017 and 2024, evaluating sensor modalities, signal processing approaches, machine learning classifiers, and clinical validation status. Multimodal wearable systems combining electrodermal activity, accelerometry, and cardiac monitoring achieve sensitivity of 80–95% for convulsive seizure detection with false alarm rates of 0.1–1.0 per hour. Focal and absence seizures remain poorly detected by peripheral sensors. The implications for seizure monitoring and patient safety in Uzbekistan are discussed.