Study of Loss of Alertness and Driver Fatigue Using Visibility Graph Synchronization

Study of Loss of Alertness and Driver Fatigue Using Visibility Graph Synchronization

Anwesha Sengupta (Indian Institute of Technology Kharagpur, India), Sibsambhu Kar (Samsung India Software Operations, India) and Aurobinda Routray (Indian Institute of Technology Kharagpur, India)
DOI: 10.4018/978-1-4666-8723-3.ch007
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Electroencephalogram (EEG) is widely used to predict performance degradation of human subjects due to mental or physical fatigue. Lack of sleep or insufficient quality or quantity of sleep is one of the major reasons of fatigue. Analysis of fatigue due to sleep deprivation using EEG synchronization is a promising field of research. The present chapter analyses advancing levels of fatigue in human drivers in a sleep-deprivation experiment by studying the synchronization between EEG data. A Visibility Graph Similarity-based method has been employed to quantify the synchronization, which has been formulated in terms of a complex network. The change in the parameters of the network has been analyzed to find the variation of connectivity between brain areas and hence to trace the increase in fatigue levels of the subjects. The parameters of the brain network have been compared with those of a complex network with a random degree of connectivity to establish the small-world nature of the brain network.
Chapter Preview
Top

Introduction

Maintenance of a performance level over time, such as that required during driving calls for sustained vigilance, selective attention and complex decision-making abilities. Long hours of continuous work, a monotonous working environment, or working hours that interfere with the circadian rhythm may bring about degradation in performance of the individual. The link between changes in behavioral arousal and the EEG spectrum is strong enough for the EEG spectrum to be used as a direct indicator of arousal level. In general, EEG features in the frequency domain have been found to be more efficient and reliable than those in the time domain for prediction of the behavioral alertness level. Changes in EEG with vigilance have generally shown that distribution, amplitude and frequency of alpha waves in the EEG spectrum change with the onset of drowsiness. A change in the pattern of alpha wave distribution during driver fatigue has been reported (Lal & Craig, 2001) A positive relation between EEG power and cognitive performance in the alpha frequency range has been reported (Klimesch, 1999) and alpha band was found to be the most important component for judging alertness level in an expectancy task (Gale et al, 1971).

EEG has widely been used to judge the alertness level of an individual during monotonous tasks or tasks requiring sustained attention or response to specific stimuli. A study of fluctuation in attention of participants in a vigilance task examines the possibility of periodicity in the pattern (Smith et al, 2003). Participants were required to listen to the letters of the alphabet arranged randomly and were required to press a button when two consecutive letters appeared.

A considerable body of work has been carried out on EEG-based fatigue detection and various methods have been reported to find the changes in EEG signal characteristics during the onset of fatigue. Relative energy of different energy bands (alpha, beta, beta/alpha ratio and (alpha+theta)/beta ratio) has often been used as an indicator of fatigue (Eoh et al, 2005). The relative energy parameter (alpha+theta)/beta has been found to decrease with a decrease in alertness level (De Waard & Brookhuis, 1991).

Another significant domain of fatigue study using EEG includes entropy as the indicator of fatigue. Shannon Entropy, Renyi entropy, Tsallis entropy, Kullback–Leibler Entropy and Cross-Approximate Entropy have often been employed as indicators of fatigue (Papadelis et al., 2006; 2007). A method based on Shannon Entropy and Kullback-Leibler Entropy measures and alpha band relative energy for relative quantification of fatigue during driving has been proposed (Kar et al, 2010).

Alertness detection procedures based on the spectral analysis of EEG signal have also been proposed (Alvarez Rueda, 2006; Jung et al, 1997). In (Makeig & Jung, 1995), minute-scale fluctuations in the normalized EEG log spectrum during drowsiness have been correlated with concurrent changes in level of performance for a sustained auditory detection task. Almost identical linear relationships have been found to exist between normalized EEG log spectra and minute-scale changes in auditory detection probability during single and dual-task experiments alike. An algorithm has been developed for automatic recognition of alertness level using full-spectrum EEG recording in (Kiymik et al, 2004). Time-frequency analysis of EEG signals and independent component analysis have been employed (Huang et al, 2008) to analyze the tonic and phasic dynamics of EEG activities during a continuous compulsory tracking task. Relative spectral amplitudes in alpha and theta bands, as well as the mean frequency of the EEG spectrum, have been used to predict alertness level in an auditory response test (Huang et al, 2001). The mean frequency of the beta band was used for the visual task study.

Complete Chapter List

Search this Book:
Reset