Abstract
An accurate method for detecting vital signs obtained from a Doppler radar sensor is proposed. A Doppler radar sensor can remotely obtain vital signs such as heartbeat and respiration rate, but the vital signs obtained by using the sensor do not show clear peaks like in electrocardiography (ECG) because of the operating characteristics of the radar. The proposed peak detection algorithm extracts the vital signs from the raw data. The algorithm shows the mean accuracy of 96.78% compared to the peak count from the reference ECG sensor and a processing time approximately two times faster than the gradient-based algorithm. To verify whether heart rate variability (HRV) analysis similar to that with an ECG sensor is possible for a radar sensor when applying the proposed method, the continuous parameter variations of the HRV in the time domain are analyzed using data processed with the proposed peak detection algorithm. Experimental results with six subjects show that the proposed method can obtain the heart rate with high accuracy but cannot obtain the information for an HRV analysis because the proposed method cannot overcome the characteristics of the radar sensor itself.
Highlights
Drowsiness has become a major social issue due to the increase in nighttime personal and social activities
Where HRECG is the mean heart rate (mHR) obtained from the number of R-peaks in the ECG waveforms and HRRadar where HRECG is the mHR obtained from the number of R-peaks in the ECG waveforms and HRRadar is is the mHR processed with the peak detection algorithm from the raw data using the radar sensor
The peak intervals located using the proposed algorithm are obtained with an mean accuracy of 96.78% when compared to the ECG signals, and the processing time is two times faster than that for the conventional algorithm
Summary
Drowsiness has become a major social issue due to the increase in nighttime personal and social activities. Drowsiness is a physiological, continuous response due to the action of the autonomic nervous system [3]. There is a time difference between the onset of cognitive drowsiness (confirmed by external features) and the physiological drowsiness (judged by biological signals). Detection of physiological drowsiness can allow time to cope with the problems posed by later-onset cognitive drowsiness [4]. Analyzing the characteristics of the physiological drowsiness that change continuously in stages can predict the occurrence of cognitive drowsiness [5]. Previous technologies used to detect the physiological drowsiness have generally been based on electrocardiography (ECG) and electroencephalography (EEG) signals [6,7,8]. Sensing technologies based on EEG signals are used to distinguish the sleep stage because signals in the alpha band (8−13 Hz) are related to sleep
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
