Wearable EEG systems based on artificial neural networks for seizure detection hold great promise for continuous and real-time epilepsy monitoring. However, clinical devices typically rely on large electrode arrays, which are impractical for portable use. This study investigates the feasibility of reducing system complexity by utilizing a patient-specific channel selection method. Comparing Single- and Dual-Channel configurations reveals that while Dual-Channel systems deliver optimal performance, Single-Channel systems strike an appealing balance between accuracy and simplicity. These findings underscore the potential for efficient, patient-specific wearable EEG devices while addressing limitations of current, more invasive approaches.
Wearable EEG systems: toward Single-Channel personalized configurations for seizure detection
Ferrara R.;Giaquinto M.;Percannella G.;Saggese A.;Vento M.
2025
Abstract
Wearable EEG systems based on artificial neural networks for seizure detection hold great promise for continuous and real-time epilepsy monitoring. However, clinical devices typically rely on large electrode arrays, which are impractical for portable use. This study investigates the feasibility of reducing system complexity by utilizing a patient-specific channel selection method. Comparing Single- and Dual-Channel configurations reveals that while Dual-Channel systems deliver optimal performance, Single-Channel systems strike an appealing balance between accuracy and simplicity. These findings underscore the potential for efficient, patient-specific wearable EEG devices while addressing limitations of current, more invasive approaches.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


