A basic study for system to prevent drowse by physiologic indices

Abstract

Traffic accidents caused by drowsiness have become a large social problem. In this study, we aimed to develop a drowsiness detection system by heart rate variability. Eight subjects particiated in this study. Silent experimental room made them drowsy. The electrocardiogram, the electroencephalogram, the electro-oculogram, the respiration and the pulse wave were measured as physiological indices. The sampling frequency was 2,000 Hz. Video pictures of subjects were taken to monitor their body motion and aspect. Simultaneously, Visual Analog Scale (VAS) was used as a subjective assessment. Heart rate fluctuation was detected using the variance of threshold level that was set by the R-R intervals. The detected zone of drowsy state was assessed by the electroencephalogram and the electro-oculogram. It was only six subjects. Tow subjects could not detect the heart rate fluctuation. The variance of threshold level needs to be improved. We proposed drowsiness detection system using the simple threshold level. This system suggested the possibility to detect the drowsy state.