Smart Shirt Respiratory Monitoring to Detect Car Driver Drowsiness

抄録

Many studies have reported method to detect car driver drowsiness by heartbeat interval signals, but it is hard to record stable signals while driving in a way that does not burden the drivers. We examined whether the driver’s drowsiness can be detected from respiration signal. In 9 healthy subjects (seven males and two females; age, 45 ± 9 y), respiration, electrocardiogram, and acceleration signals were recoded for a total of 2,359 min (137-468 min per subject) of driving with a smart shirt biometric sensor (Hexoskin). Minute-to-minute respiratory rate, heart rate, and their variability were analyzed by complex demodulation. The sleepiness of drivers was assessed by subjective reports and by a surrogate maker of Dip & Waves, which is known to be a characteristic R-R interval pattern associated with driver drowsiness. Although respiratory rate showed no significant changes associated with Dip & Wave, respiratory rate variability increased progressively from 4 min before, peaked at Dip & Waves, and decreased immediately thereafter. No such definite trend was observed in any time- or frequency domain indices of heart rate variability. The findings of this study not only show the possibility of smart-shirt sensor as a device to detect driver drowsiness but also suggest that respiration signals may provide useful clues to predict driver drowsiness, which is unique from those provided by heart rate variability.