A Real-Time Wrist Worn Actigraphy System for Sleep Monitoring

Rachel Brandt, Mert Sevil1, Ali Cinar1

  • 1Illinois Institute of Technology

Details

18:15 - 20:15 | Mon 5 Mar | Caribbean ABC | MoPO.13

Session: Poster Session # 1 and BSN Innovative Health Technology Demonstrations

Abstract

A wrist worn actigraphy system with real-time data streaming has been developed. 20 nights of sleep were recorded and evaluated with the proposed fuzzy-logic based system offline. 30 second epochs were scored as sleep or wake and full night sleep efficiencies were determined for the evaluation of the proposed system. The proposed system had 85.7% per-epoch accuracy when compared to EEG derived sleep stages. The sleep efficiencies from the proposed system had an RMSE of 0.0694 when compared to the EEG-based sleep efficiencies while the Phillips Actiwatch Spectrum Plus had an RSME of 0.10065 when compared to EEG-based sleep efficiencies for the same nights.