Distraction and inattention of drivers have been identified as the main leading causes of car accidents. The U.S. National Highway Traffic Safety Administration has identified driver distraction as a high priority area about 30% [1]. Driver distraction by whatever cause is a significant contributor to road traffic accidents [2] [3]. Driving is a complex task in which several skills and abilities are involved simultaneously.
Monitoring drivers’ attention related brain resources is still a challenge for researchers and practitioners in the field of cognitive brain research and human–machine interaction.
Reasons of distractions found during driving were quite widespread, including eating, drinking, talking with passengers, use of cell phones, reading, fatigue, problem-solving, and using in-car equipment. Recently, commercial vehicle operators with complex in-car technologies (such as navigation, road traffic information, mobile telephones and in-vehicle entertainment systems) are also at increased risk since drivers may become increasingly distracted in the years to come, thus making it likely that the problem of driver inattention [4] [5]. Some literatures studied the behavioral effect of driver’s distraction in car. Tijerina’s study was based on measurement of the static completion time of an in-vehicle task [6]. Similarly, the distraction effect caused by cellular phones during driving has been a focal point of recent in-car applications [7] [8] [9]. Experimental studies have been conducted to assess the impact of specific types of driver distraction on driving performance. While these studies have generally reported significant driving impairment [10] [11], simulator studies cannot provide information about the impact of these decrements on the occurrence of crashes
resulting in hospital attendance by the driver. Therefore, in order to provide information before the occurrence of crashes we try to investigate the drivers’
physiological responses.
To the aspect of neural physiological investigation, some literatures focused on the brain activities of “divided attention” referring to attention divided between two or more sources of information, such as visual, auditory, shape, and color stimuli.
Madden et al. [12] investigated brain activation when subjects were instructed to divide their attention among the display positions within the visual modality. Regional cerebral blood flow (rCBF) activation was found in occipitotemporal, occipitoparietal, and prefrontal regions. And Positron emission tomography (PET) measurements were taken while subjects discriminated between shape, color, and speed of a visual stimulus under conditions of selective and divided attention. The divided condition activated the anterior cingulated and prefrontal cortex in the right hemisphere [13]. In another study, functional magnetic resonance imaging (fMRI) was used to investigate the brain activity during a dual-task (visual stimulus) experiment. This found activation in the posterior dorsolateral prefrontal cortex and lateral parietal cortex [14].
Similarly, the study used electroencephalogram (EEG) to investigate mental arithmetic-induced workload increasing, the finding is power increase in theta band in the region of frontal lobes [15]. And, several neuroimaging studies showed the importance of the prefrontal network in dual-task management [16] [17]. However, the above-mentioned studies just investigated the brain activity of dual-task interaction without considering the stimulus onset asynchrony (SOA) problem during driving and the effect of different temporal relationship of stimuli.
The current investigation utilized an array of methodological assessment techniques and compared the sensitivity of each to changes in attention processing requirements as a function of driving task demand. Some literatures of investigated
the traffic record the electroencephalogram (EEG) to compare the P300 amplitude [18]. During simulated traffic scenarios, resource allocation was assessed through as event-related potential (ERP) novelty oddball paradigm [19]. However, these are just to analyze in time course, we can take one step to analyze the relation between time and frequency course.
The electroencephalogram (EEG) has been used for 80 years in clinical practices as well as basic scientific studies. Nowadays, EEG measurement is widely used as a standard procedure in researches such as sleep studies [20], epileptic abnormalities, and other disorders diagnoses. Comparing to another widely used neuroimaging modality, functional Magnetic Resonance Imaging (fMRI), EEG is much less expensive and has the superior ability of temporal resolution for us to investigate the SOA problems. Furthermore, to avoid the interference and risks of operating an actual vehicle on the road, the use of driving simulation for vehicle design and studies of driver’s behavior and cognitive states is also expanding rapidly [21] [22]. The static driving simulation may be difficult to approach the realistic driving condition, such as the vibrations that would be experienced when driving an actual vehicle on the road.
The VR technique allows subjects to interact directly with a virtual environment rather than monotonic auditory and visual stimuli. Integrating realistic VR scenes with visual stimulus is easier to study the brain response to visual attention during driving.
Therefore, in recent years, the VR-based simulation combined with electroencephalogram (EEG) monitoring is an innovation in cognitive engineering research [20] [23].
The main goal of this study is to investigate the brain dynamics related to distraction by using EEG and VR-based realistic driving environment. Unlike the previous studies, our experiment has three main characteristics. First, the stimulus onset asynchrony (SOA) experimental design, the different appearance time of dual
tasks (mathematical questions and unexpected car deviation), has the benefits for us to investigate the driver’s behavioral and physiological response under multiple conditions and multiple distracted levels. Second, the ICA-based advanced signal analysis methods were used to extract the artifact-free brain responses and related cortical location related to the single/dual task. Third, compared with single task, the interaction and effect of dual-task-related brain activities was also investigated. The detailed contents are described in the following sections.
The thesis was organized in 6 chapters. Chapter 1 briefly introduced current knowledge in vestibular system and the goal of the study. Chapter 2 detailed the apparatus and materials of the study. Chapter 2 also described the details of experimental setup, including the time course of event onset asynchrony setup. In chapter 3, we explored the EEG with innovative methods by combining Independent Component Analysis (ICA), time-frequency spectral analysis, power spectrum and component clustering. Chapter 4 showed the results. Chapter 5 discussed and compared our finding with previous studies, and finally we concluded in Chapter 6.