Title Driver cardiovascular disease detection using seismocardiogram /
Authors Uskovas, Gediminas ; Valinevicius, Algimantas ; Zilys, Mindaugas ; Navikas, Dangirutis ; Frivaldsky, Michal ; Prauzek, Michal ; Konecny, Jaromir ; Andriukaitis, Darius
DOI 10.3390/electronics11030484
Full Text Download
Is Part of Electronics.. Basel : MDPI. 2022, vol. 11, iss. 3, art. no. 484, p. 1-16.. ISSN 2079-9292
Keywords [eng] adaptive digital filter ; arrhythmia ; driving restrictions ; heart rate ; noninvasive method
Abstract [eng] This article deals with the treatment and application of cardiac biosignals, an excited accelerometer, and a gyroscope in the prevention of accidents on the road. Previously conducted studies say that the seismocardiogram is a measure of cardiac microvibration signals that allows for detecting rhythms, heart valve opening and closing disorders, and monitoring of patients' breathing. This article refers to the seismocardiogram hypothesis that the measurements of a seismocardiogram could be used to identify drivers' heart problems before they reach a critical condition and safely stop the vehicle by informing the relevant departments in a nonclinical manner. The proposed system works without an electrocardiogram, which helps to detect heart rhythms more easily. The estimation of the heart rate (HR) is calculated through automatically detected aortic valve opening (AO) peaks. The system is composed of two micro-electromechanical systems (MEMSs) to evaluate physiological parameters and eliminate the effects of external interference on the entire system. The few digital filtering methods are discussed and benchmarked to increase seismocardiogram efficiency. As a result, the fourth adaptive filter obtains the estimated HR = 65 beats per min (bmp) in a still noisy signal (SNR = −11.32 dB). In contrast with the low processing benefit (3.39 dB), 27 AO peaks were detected with a 917.56-ms peak interval mean over 1.11 s, and the calculated root mean square error (RMSE) was 0.1942 m/s2 when the adaptive filter order is 50 and the adaptation step is equal to 0.933.
Published Basel : MDPI
Type Journal article
Language English
Publication date 2022
CC license CC license description