Utilizing HRV-Derived Respiration Measures for Driver Drowsiness Detection

Abstract

This study aims to utilize heart rate variability (HRV) signals obtained with a wearable sensor for driver drowsiness detection. To this end, we investigated respiration characteristics derived from HRV signals based on the known fact that respiratory activity can be estimated from the high frequency (HF) band of HRV signals. For drowsiness detection, many earlier works commonly used dominant respiration (DR) characteristics. However, in some situations where emphasized power in a power spectrum of HRV occurs at multi sub-frequency, the DR measures may possibly fail to capture overall respiration characteristics. To handle this problem, we propose two spectral indices, the weighted mean (WM) and the weighted standard deviation (WSD) of the HF band in the power spectrum. These indices are used to properly capture the overall shape of the respiratory activity shown through the HF band of the HRV power spectrum as an alternative to the DR measures. For experiments, we collected HRV data with an electrocardiogram device worn on the body under a virtual driving environment. The proposed indices somewhat clearly showed the tendency that respiratory frequency decreases and respiration regularity increases in drowsy states of all subjects, while existing DR measures hardly showed this. In addition, when the proposed indices are used alone or together with conventional HRV-related measures as input features for classification models, they showed the best performance in distinguishing drowsiness from wakefulness.

Non-Invasive Driver Drowsiness Detection System

“…Figure 12 comprises 80 points (40 for each label of drowsy/non-drowsy); however, some of the points overlapped, which is why Figure 12 has fewer points. The result reflects the fact that respiration per minute decreases in a drowsy state [ 26 ]. Further, in a drowsy state, the decrease in RPM is higher in younger as compared to older subjects.…”

“…A driver drowsiness detection system based on respiration rate acquired using an inductive plethysmography belt is proposed in [ 49 ]. A system using HRV derived respiration measures to detect driver drowsiness using a wearable ECG device (Polar H7) is presented in [ 26 ]. RR-intervals (RRI) data with constant time intervals at the sampling rate of 0.5 s is acquired by performing Cubic interpolation.…”

“…Breathing during sleep is affected by a reduction of muscular tone and the alteration of chemical and non-chemical responses [ 23 ]. Respiration rate usually decreases before the driver falls asleep [ 24 , 25 , 26 ].…”

Non-Contact Driver Respiration Rate Detection Technology Based on Suppression of Multipath Interference with Directional Antenna

“…The respiratory rate of drivers is an important index that reflects their physiological state. Effective detection of the driver's respiratory rate can provide important medical data support to recognize related scenarios of the driver, such as fatigue driving [1]. Valderas used heart rate variability analysis with spectral bands based on respiration to recognize human emotion [2].…”

Pre-Emption of Affliction Severity Using HRV Measurements from a Smart Wearable; Case-Study on SARS-Cov-2 Symptoms

“…According to medical research, the LF/HF ratio serves as an index to assess the relationship between the sympathetic and parasympathetic processes and therefore mark the variation in type and handling of causal stress [ 34 ]. An interesting application for drowsiness detection was developed that relied on HRV signals to discern dominant respiratory characteristics [ 35 ].…”