Strengthening association between driver drowsiness and its physiological predictors by combining EEG with measures of body movement

Driver fatigue is acknowledged as a major contributing factor in motor vehicle accidents that result in serious injuries or death. As a result it is valuable to develop driver drowsiness monitoring and warning systems. Many systems that are currently being developed utilise a single source of data to evaluate drowsiness level, however it is anticipated that using hybrid data sources would increase the accuracy of such devices. The objective of this analysis was to determine if using a combination of EEG and body movement parameters would increase the ability to accurately predict the graduated driver drowsiness levels compared to EEG signals alone. Addition of the body movement data has increased the goodness of fit in modelling the average driver drowsiness using a linear regression (R² = 0.272 for EEG alone and R² = 0.308 for EEG and body movement combined).