Stress level classification using statistical analysis of skin conductance signal while driving
Open Access
- 10 January 2021
- journal article
- research article
- Published by Springer Nature in SN Applied Sciences
- Vol. 3 (1) , 1-9
Abstract
Conventionally, multiple physiological signals are used in the field of stress realization. Although many studies have applied various methods in feature selection and classification, a desirable performance has not yet been achieved. This paper presents a novel method of stress level classification using physiological signals during the real-world driving task. Exploring the most reliable analysis method on a comprehensive physiological signal for stress realization has been commonly investigated in various studies. To obtain a high accuracy approach, a proper classification method should be applied to the most relevant physiological signal. In this study, we evaluate the feasibility and effectiveness of the analysis of variance (ANOVA) classifier learner on the single Galvanic Skin Response (GSR) signal. Three levels of stress are taken into account and two independent features including rising time and amplitude are extracted. These two features are extracted from foot and hand GSR signals in three different scenarios for the sake of training. The result indicates that the foot amplitude feature of the GSR signal solely is a reliable source of stress classification with an accuracy rate of 95.83% by applying the ANOVA approach. Accordingly, this methodology can substantially reduce the necessity of resorting to the high number of sensors and the corresponding computational burden associated with signal analysis. Besides, reducing the number of sensors during the measurement procedure would increase drivers’ safety by reducing the interference between human and measurement devices. In this study, the real data collected by Picard and his co-workers are used, available in the PHYSIONET database.Keywords
This publication has 22 references indexed in Scilit:
- Identifying Traffic Context Using Driving Stress: A Longitudinal Preliminary Case StudySensors, 2019
- A Review of Psychophysiological Measures to Assess Cognitive States in Real-World DrivingFrontiers in Human Neuroscience, 2019
- A real-time stress classification system based on arousal analysis of the nervous system by an F-state machineComputer Methods and Programs in Biomedicine, 2017
- Drowsiness detection using heart rate variabilityMedical & Biological Engineering & Computing, 2016
- Insightful stress detection from physiology modalities using Learning Vector QuantizationNeurocomputing, 2015
- A k-nearest-neighbor classifier with heart rate variability feature-based transformation algorithm for driving stress recognitionNeurocomputing, 2013
- Objective measures, sensors and computational techniques for stress recognition and classification: A surveyComputer Methods and Programs in Biomedicine, 2012
- A reasoning-based framework for car driver’s stress prediction2008
- Galvanic skin response (GSR) as an index of cognitive load2007
- Detecting Stress During Real-World Driving Tasks Using Physiological SensorsIEEE Transactions on Intelligent Transportation Systems, 2005