Synchrosqueezing Index for Detecting Drowsiness Based on the Respiratory Effort Signal

Abstract

Objective: The aim of this work is to evaluate a new index to assess the alertness state of drivers based on the respiratory dynamics derived from an inductive band using the variability of the respiratory rhythms assessed by Synchrosqueezing (SSQ).

Background: Biomedical variables like abdominal effort, which is related to autonomic nervous system, provides direct information of the driver physiological state, instead of indirect indicia of the participant’s behavior.

Method: The respiration data used in this study was recorded by doing 18 simulator tests in a controlled scenario. In order to evaluate the viability of variability of the respiratory rhythms assessed by Synchrosqueezing (SSQ) Mean and the Standard Deviation of the Instantaneous Respiration Frequency (MIRF and SDIRF) obtained in the analysis for awake states and drowsy states individually was calculated.

Results: The results demonstrate the viability of drowsiness detection in simulator using abdominal effort signal analyzed by Synchrosqueezing based on the significant results of the statistical analysis comparing the Mean (MIRF) and de Standard Deviation (SDIRF) of both awake and drowsy intervals instantaneous respiration frequency (MIRF 0.28 Hz ± SDIRF 0.06 Hz (awake) vs. MIRF 0.21 Hz ± SDIRF 0.06 Hz (drowsy), N=18, p<0.001, t=4.88.