دانلود مقاله ISI انگلیسی شماره 38799
عنوان فارسی مقاله

غرق در افکار در حال رانندگی: مطالعه EEG در حواس پرتی شناختی رانندگان

کد مقاله سال انتشار مقاله انگلیسی ترجمه فارسی تعداد کلمات
38799 2014 9 صفحه PDF سفارش دهید محاسبه نشده
خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.
عنوان انگلیسی
Deep in thought while driving: An EEG study on drivers’ cognitive distraction
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Transportation Research Part F: Traffic Psychology and Behaviour, Volume 26, Part A, September 2014, Pages 218–226

کلمات کلیدی
حواس پرتی های شناختی - تجزیه و تحلیل نیمکره غربی - قشر فرونتال - رفتار رانندگی
پیش نمایش مقاله
پیش نمایش مقاله غرق در افکار در حال رانندگی: مطالعه EEG در حواس پرتی شناختی رانندگان

چکیده انگلیسی

Abstract Our research employed the EEG to examine the effects of different cognitive tasks (math and decision making problems) on drivers’ cognitive state. Forty-two subjects participated in this study. Two simulated driving sessions, driving with distraction task and driving only, were designed to investigate the impact of a secondary task on EEG responses as well as the driving performance. We found that engaging the driver’s cognitively with a secondary task significantly affected his/her driving performance as well as the judgment capability. Moreover, we found that different features of the secondary task had different effects on EEG responses and different localizations in the frontal cortex. Our hemispheric analysis results showed that the most affected area during distracted driving was in the right frontal cortex region; thus, it is suggested that the activation in the right frontal cortex region may be considered the spatial index that indicated a driver who is in a state of cognitive distraction.

مقدمه انگلیسی

. Introduction Driving is a complex task that depends on a set of cognitive skills in association with the contributions of planning, memory and motor control and visual capabilities. These capabilities vary from one individual to another depending on the cognitive skills and level of attention (Shinar, 1993). In past decades, driving distraction is increasingly identified as one of significant causes of traffic accidents and has the same effect on driving performance as drugs and alcohol. In fact, NHTSA estimated that various drivers’ distraction sources caused about 20–80% of crashes and near-crashes (Stutts & Association, 2001). More recently, a wide naturalistic driving study of 100 cars found that inattention was a cause in 78% of all crashes and near crashes, thus considering it the largest crash causation factor in their analysis (Dingus et al., 2006). Driving distraction, generally, is defined as the deviation of driver’s attention away from operating safe driving toward a competing activity (Young, Lee, & Regan, 2008). Therefore, the cause of driving distraction could be due to any cognitive process such as daydreaming, mind wondering, mathematical problem solving or decision making issues in addition to using in-vehicle information systems (IVI’s) such as Audio systems, navigation systems and cell phones that may affect driver’s attention on driving. When drivers are cognitively distracted, visual information processing becomes lower which markedly impairs driving performance in detecting targets across the entire visual scene (Lee et al., 2009, Recarte and Nunes, 2000 and Recarte and Nunes, 2003). Many studies have investigated the impact of a secondary task on driving performance. These studies have used mobile phone related task (general usage of the mobile phone), conversation with passengers, and other tasks as a secondary task (Brookhuis et al., 1991, Chaparro et al., 2004, Crundall et al., 2005, Lamble et al., 1999 and Levy et al., 2006). The two major types of distraction are visual distraction and cognitive distraction. Visual distraction can be defined as “eyes-off-road”, and cognitive distraction as “mind-off-road” (Victor, 2005). Both types of distraction can affect driving performance such as lane variation, steering control, response to hazards, and visual perception efficiency. Moreover, visual and cognitive distraction interacts with each other and can occur in combination. The current study will focus on driver’s cognitive distraction. Cognitive distraction and inattention will be used interchangeably in our context of study. From the general definition both are considered as the decrement of mental concentration to a specific task (Anderson, 2009). To better understand driver psychological behavior and the sources of driver cognitive distraction, researchers have attempted to develop models that captured brain electrical activity (EEG) (e.g., Dong, Hu, Uchimura, & Murayama, 2011 and Lin, Ko, & Shen, 2009). Such models provide a better understanding of the effects of distraction on driver behavior through capturing changes in EEG activity. Measures of brain electrical activity (EEG) are the most valid measures used for distraction measurement (Lin et al., 2009). EEG has the advantage of high temporal resolution which allows for the ability to perform cognitive studies and instantaneously evaluate the corresponding brain activity. EEG recording is completely non-invasive and can be applied repeatedly to patients, normal adults, and children with no risk or limitation (Teplan, 2002). Galán and Beal (2012) in their study in evaluating whether the EEG could estimate the attention and the cognitive workload in predicting success or failure of math problem solving, suggested that EEG might be a valuable tool for assessing cognitive workload. Due to the rapid increase of in-vehicle technologies, the psychological changes in drivers are more complex and hard to detect. Therefore, a study by Schier (2000) has described the need of using more advanced technologies to study the rapid changes of the driver cognitive state during driving. The study has investigated the suitability in using EEG-based technologies simultaneously with a driving simulator through the activities in the alpha frequency band (8–13 Hz) between driving and driving-replay sessions. It has been agreed that the alpha band is the most dominant band for studying attention (Klimesch et al., 1998 and Schier, 2000 and Wolfgang, 1999). Dynamic changes in alpha activity corresponding to the changes in driving events have been documented (Schier, 2000). Furthermore, this study concluded a high effectiveness of the exploratory experimental work in demonstrating the practicality of such EEG recordings during simulated driving. Many studies have investigated the human factor in road crashes. Lee et al. (2009) have investigated the effect of drivers’ cognitive load on the relation between internal and external attention control; as reviewed, the cognitive load has a high influence on withdrawing driver attention and decrease the driver’s ability in detecting road hazards through a cue-based pedestrian paradigm. He has found that the cognitive load delayed the driver response and reduced his fixation to pedestrians and external cues. A good index of cognitive distraction that is widely accepted in EEG measurements consists of theta activity (4–8 Hz), alpha activity (8–14 Hz) and Beta activity (14–35 Hz) (Lin, Chen, Ko, & Wang, 2011). Theta and beta activity in brain frontal lobes are associated with cognitive processes such as judgment, problem solving, working memory, decision making and mathematical problem solving (Lin et al., 2011). The increasing amplitudes of these particular bands are often a result of brain engagement in such activities. The role of attention on EEG activity has been extensively studied. Klimesch et al. (1998) studied induced alpha band power changes in EEG signals and attention through an oddball task. After separating alpha into 3 sub-bands – lower, medium and upper – they found that only the lower alpha reflected the attentional demands. Also from his study on the reflection of cognitive and memory performance on alpha and theta EEG bands (Wolfgang, 1999), he suggested that alpha in different sub bands was highly influenced by attentional and semantic memory processes. One of the most important findings was the increasing in the upper alpha bands desynchronization during the semantic judgment task but there was no response from theta activity. The highest activities of alpha corresponded to the judgment task seen in the prefrontal left hemisphere, and this was supported by findings from a PET study (Endel, Shitu, Craik, Morris, & Sylvain, 1994). An important conclusion was that the increase in theta power and decrease in alpha power indicated poor cognitive and memory performance. On the other hand, the decrease in alpha power indicated high attention to a specific task while increase in theta power indicated distraction (low attention) to a specific task. As a matter of fact, drivers’ cognitive distraction is the most difficult to assess and evaluate among the three types of driver distraction due to the inability of directly observing what is going on in the driver’s brain. One possible solution to the problem is to capture changes in driving behavior using objective measures that will also serve as a qualitative assessment associated with cognitive distraction and visual distraction (Angell et al., 2006 and Engström et al., 2005). Such objective measures in tracking driving behavior and performance have been widely used to confirm the effects of different types of driving distraction. For example, (Horberry, Anderson, Regan, Triggs, & Brown, 2006) focused on two speed-related variables (mean speed and deviation from the posted speed limit) in measuring driving behavior changes. They reported that in-vehicle tasks have high negative impact on the studied driving behavior measures. A study by Boril and colleagues on the effects of cognitive load and driver emotions on driving speed used lane control capability as a driving performance indicator. They suggested that the secondary cognitive task and drivers’ emotion severely impacted driving performance as they found a high reduction in the steering wheel control ability (Boril, Omid Sadjadi, Kleinschmidt, & Hansen, 2010). In this study, we investigated the changes in drivers’ cognitive state through the changes in recorded EEG signals. As the subjects were placed into different driving situations, changes in their EEG responses were obtained to track changes that reflected changes in their cognitive state induced in the experimental design. Employing the EEG provided a reliable indicator of the fluctuations in drivers’ cognitive state during driving. As such, the obtained data might eventually be incorporated into real-time systems that could intervene or warn a driver if he/she is drifted to a cognitive state that may compromise his/her safety. The objectives of this study are to: 1. Investigate the effects of a secondary task that employs cognitive resources on driving behavior. 2. Determine cognitive changes as measured by EEG signals of a secondary task while driving. 3. Identify specific regions and frequency bands involved in the activity of simultaneously completing a secondary task and driving. 4. Provide high spatial resolution EEG data comparing cognitive states of driving and driving while distracted. 5. Provide hemispheric analysis of driver cognitive distraction by comparing EEG changes over the frontal left and right hemispheres during driving.

نتیجه گیری انگلیسی

3. Result 3.1. Driving performance Significant changes were observed in driving performance (lane deviation and number of accidents). The mean and standard deviation of these measures were presented in Table 2. A higher value for these measures indicated poorer performance as they measured the number of instances of deviating from the driving lane and near-crashes or accidents. Paired t-test analyses suggested significant impaired driving performance during the distracted-driving sessions – there were more instances of lane deviation (t(41) = −3.53; p < 0.01) and near-crashes or accidents (t(41) = −2.05; p < 0.05).This suggested that engaging drivers in a secondary cognitive task while driving significantly affected driving performance. Table 2. Pair-sample T-test of driving performance measures results. Driving Distraction p-value σ SD σ SD Lane keeping 3.57 1.93 4.85 2.48 .001 Accidents .52 .77 .78 1.04 .047 Table options 3.2. Task response The task was administered to participants during and after driving in the simulator in order to determine the performance of solving such problems while driving (Mean = 11.19; SD = 3.46) and not driving (Mean = 14.81; SD = 3.22). Paired-sample t-test results indicated significant differences in correctly answering the questions in the two conditions (t(41) = 10.04; p < 0.001). Participants did better on the secondary task when they were not driving in the simulator. Evaluating these results with results from the previous Section 3.1 provided support that the secondary task employed in the current study was able to induce cognitive distraction when one was driving. 3.3. EEG measures Greater differences were observed in EEG bands (theta, alpha and beta) for forty subjects (N = 40), we would like to mention that two subjects data were excluded due to the huge noise in the recordings. This observation suggested an increase in brain activity in the frontal lobe during distraction while and after giving the cognitive task. This finding is consistent with roles of the frontal lobe rule in attention, problem solving and decision-making ( Burgess, Alderman, Volle, Benoit, & Gilbert, 2009). Table 3 includes the mean and the standard deviation of both distracted-driving and non-distracted driving sessions in right and left frontal lobe while performing math task and decision making task. Table 3. The mean (μ) and standard deviation (σ) of the averaged EEG frequency bands in the right and left frontal lobe in the non-distracted driving session (Dri) and distracted driving session (Dis) when participants solved math problems (math) and decision making problems (DM). Distracted-driving μ σ Driving μ σ Right Amp 182,998.7 109,067.1 Amp 144,929.5 107,679.7 Theta 204,386 110,870.9 Theta 160,376.3 98,061.1 Alpha1 117,448.5 127,679.1 Alpha1 100,199.3 150,981.4 Alpha2 107,429.2 124,809 Alpha2 98,039.49 147,433 Beta1 180,865.1 134,128.4 Beta1 84,999.06 98,584.17 Beta2 811,997.6 533,767.2 Beta2 307,378.4 123,184.3 Amp_math 151,654.6 109,404.4 Amp_math 83,106.74 57,213.95 Theta_math 175,603.2 128,971.9 Theta_math 106,673.2 76,597.83 Alpha1_math 76,870.53 81,379.39 Alpha1_math 44,627.53 46,149.31 Alpha2_math 67,736.99 74,238.71 Alpha2_math 28,537.75 29,647.29 Beta1_math 152,926.6 183,319 Beta1_math 62,687.7 67,759.76 Beta2_math 753,896.5 547,129.7 Beta2_math 265,613.3 133,318.1 Amp_DM 92,550.3 115,104.7 Amp_DM 92,550.3 115,104.7 Theta_DM 121,214.3 126,205.7 Theta_DM 121,214.3 126,205.7 Alpha1_DM 190,198.9 321,401.9 Alpha1_DM 190,198.9 321,401.9 Alpha2_DM 190,872.9 323,115 Alpha2_DM 190,872.9 323,115 Beta1_DM 193,588.4 301,662.9 Beta1_DM 193,588.4 301,662.9 Beta2_DM 699,386.2 528,195.7 Beta2_DM 699,386.2 528,195.7 Left Amp 146,416.83 65,319.93 Amp 110,591 38,749.83 Theta 155,259.56 60,393.78 Theta 131,222.2 62,671.81 Alpha1 63,318.57 55,718.14 Alpha1 62,121.21 58,431.38 Alpha2 67,929.7 53,164.63 Alpha2 68,655.71 57,215.55 Beta1 116,735.88 51,660.44 Beta1 105,350.7 50,933.2 Beta2 600,959.33 174,923.31 Beta2 482,481.9 114,166.9 Amp_math 306,004.8 656,427.2 Amp_math 275,721.6 830,530.6 Theta_math 304,722.3 710,808.9 Theta_math 185,142.4 497,604.3 Alpha1_math 141,332.5 352,090.7 Alpha1_math 259,654.5 921,451.7 Alpha2_math 188,857.8 538,009.4 Alpha2_math 313,170.8 930,441.6 Beta1_math 273,542.1 557,105.7 Beta1_math 267,446.6 558,293 Beta2_math 894,511.8 835,757.4 Beta2_math 582,017.7 565,597.6 Amp_DM 149,488.3 142,793.5 Amp_DM 303,055 510,969.9 Theta_DM 182,036.8 194,258.6 Theta_DM 360,153.6 626,402 Alpha1_DM 89,806.08 87,280.16 Alpha1_DM 24,927.79 30,402.4 Alpha2_DM 93,743.8 167,142.7 Alpha2_DM 338,274.2 766,925.6 Beta1_DM 165,006.6 178,051.7 Beta1_DM 105,081.3 193,669.5 Beta2_DM 605,381.1 396,548.7 Beta2_DM 569,828.8 366,953.6 Table options Table 4 summarizes results of paired-sample t-test for extracted features from recorded EEG data during both driving and distracted-driving sessions and the effects of the secondary tasks. The data of left and right frontal hemispheres were analyzed separately as well as the data from each secondary task – decision making and math problems. Table 4. p-value and t-value from Pair- sample T-test of all frequency bands and the amplitude from EEG data recorded in both driving (Dri) session and distracted-driving (Dis) session corresponding to the distraction tasks (logical reasoning (DM) and real-life problems involving measurements (Math). Tested pair p-value t-value Right Dis_Amp – Dri_Amp p < 0.05 2.858 Dis_Amp _Math – Dri_Amp _Math p < 0.001 3.605 Dis_Theta – Dri_Theta p < 0.05 −2.223 Dis_Theta _Math – Dri_Theta _Math −2.995 Dis_Alpha1 – Dri_Alpha1 −2.801 Dis_Alpha1 _Math – Dri_Alpha1 _Math p < 0.001 −3.541 Dis_Alpha2 _Math – Dri_Alpha2 _Math −4.188 Dis_Beta1 _DM – Dri_Beta1 _DM −2.21 Dis_Beta2 _Math – Dri_Beta2 _Math 3.242 Dis_Beta2_ _DM – Dri_Beta2 _DM −2.902 Dis_Amp _DM – Dri_Amp_DM No significance Dis_Theta _DM – Dri_Theta_DM Dis_Alpha1 _DM – Dri_Alpha1_DM Dis_Alpha2 – Dri_Alpha2 Dis_Alpha2 _DM – Dri_Alpha2_DM Dis_Beta1 – Dri_Beta1 Dis_Beta2 – Dri_Beta2 Left Dis_Amp – Dri_Amp p < 0.05 −2.392 Dis_Theta _DM – Dri_Theta _DM −2.003 Dis_Alpha2 _DM – Dri_Alpha2 _DM −1.971 Dis_Amp _Math – Dri_Amp_Math No significance Dis_Amp _DM – Dri_Amp_DM Dis_Theta – Dri_Theta Dis_Theta _Math – Dri_Theta _Math Dis_Alpha1 – Dri_Alpha1 Dis_Alpha1 _Math – Dri_Alpha1 _Math Dis_Alpha1 _DM – Dri_Alpha1 _DM Dis_Alpha2 – Dri_Alpha2 Dis_Alpha2 _Math – Dri_Alpha2 _Math Dis_Beta1 – Dri_Beta1 Dis_Beta1 _Math – Dri_Beta1 _Math Dis_Beta1 _DM – Dri_Beta1 _DM Dis_Beta2 – Dri_Beta2 Dis_Beta2 _Math – Dri_Beta2 _Math Dis_Beta2 _DM – Dri_Beta2 _DM Table options The results in Table 4 suggested significant changes in EEG activity in both left and right frontal hemispheres, and these differences highlighted the influence of the secondary task used in the experiment. The largest changes were in EEG amplitudes in both right (t(41) = 2.858; p < 0.05) and left (t(41) = −2.392; p < 0.05) frontal lobe hemispheres, theta band (t(41) = −2.223; p < 0.001) and lower alpha (t(41) = −2.801; p < 0.05). The significant changes illustrated the increase in the level of human cognitive workload which reflects the distraction caused by the secondary tasks. The effects from each secondary task were studied separately in order to investigate their effects on the driver’s brain activity while driving. For that purpose, the data corresponding to distraction tasks (math and DM) were extracted and analyzed. Table 4 suggested that the math task significantly affected EEG amplitude (t(41) = 3.605; p < 0.001), theta band (t(41) = −2.995; p < 0.05), both lower and upper alpha bands (t(41) = −3.541; p < 0.001, t(41) = −4.188; p < 0.001) and upper beta band (t(41) = 3.242; p < 0.001) in the right frontal hemisphere while there were no significant changes related to math task in the left frontal hemisphere. On the other hand, t-test results suggested that there were significant effects produced by the DM task in both right and left frontal hemispheres in specific EEG bands influenced. In the right hemisphere, lower beta (t(41) = −2.21; p < 0.001) and upper beta (t(41) = −2.902; p < 0.001) while in the left hemisphere, there were significant changes in theta (t(41) = −2.003; p < 0.05) and upper alpha (t(41) = −1.971; p < 0.05). The results in Table 4 suggested that the right frontal hemisphere was most affected by the secondary tasks compared to the left hemisphere. Corresponding to the tasks given, solving the math task created a more localized effect in the right hemisphere only while solving decision making (DM) task engaged the frontal region in both hemispheres.

خرید مقاله
پس از پرداخت، فوراً می توانید مقاله را دانلود فرمایید.