Edited by: Daniel S. Margulies, Max Planck Institute for Human Cognitive and Brain Sciences, Germany
Reviewed by: Kimmo Alho, University of Helsinki, Finland; Lutz Jäncke, University of Zurich, Switzerland
*Correspondence: Tom A. Schweizer, Keenan Research Centre of the Li Ka Shing Knowledge Institute, St. Michael's Hospital, 30 Bond Street, Toronto, ON M5B 1W8, Canada. e-mail:
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc.
Driving is an essential daily activity for many people, providing mobility, independence, and sometimes a source of livelihood. Previous research has reported factors that increase the risk of vehicle crashes and driving errors including fatigue and sleepiness after extended driving (Sagaspe et al.,
Safe driving requires the ability to concentrate, to divide attention between multiple sensory events across visual and auditory modalities, and to make fast cognitive decisions in a complex and rapidly changing environment. The present study applied virtual reality (VR) technology in a functional magnetic resonance imaging (fMRI) system to investigate how brain responses of healthy adults change across various driving scenarios. Using this novel setup allowed us to obtain a real-time, spatiotemporal profile of brain activity while driving in a safe environment. Previous fMRI driving research has identified brain areas responsive to different aspects of driving (Carvalho et al.,
Cognitive psychologists have proposed that the posterior attention system engages brain areas such as the occipital-parietal and posterior cingulate regions that are critical to visual-spatial orientation and integration functions. In contrast, the anterior attention system serves a higher-level attention function, engaging anterior portions of the frontal lobes such as the prefrontal and anterior cingulate regions, which are responsible for executive attentional control in more complex cognitive tasks associated with problem-solving and decision-making, especially during multi-tasking (Posner and Dehaene,
The current study focused on identifying the underlying neural networks subserving different driving behaviors including distracted driving. Our immersive and novel VR setup (Kan et al.,
We hypothesized that complex simulated driving conditions (such as making turns, particularly left turns at busy intersections while encountering oncoming traffic) would involve posterior brain activations including motor and occipital-parietal regions for visual-spatial and visual-motor integration. On the other hand, prefrontal activation would be involved in a distracted condition consisting of performing a secondary cognitive task during simulated driving, related to the executive function demands when attentional resources are divided across multiple tasks.
Ethical approval for the study was obtained on July 18th 2010, by the Research Ethics Board at Baycrest Hospital in Toronto, Canada. All participants provided written informed consent prior to participating in the study.
Participants were recruited through the university network via advertisement and emails. All participants were right-handed with normal or corrected vision. Participants without a valid driver's license, with a history of psychological or neurological illness, or with fMRI contraindications (such as having claustrophobia or ferromagnetic implants) were excluded. Sixteen participants (7 females and 9 males) between the ages of 20 and 30 years (Mean = 25.8,
We applied a novel approach using an immersive VR environment in a 3.0 Tesla MRI system to capture brain activity with high ecological validity (Kan et al.,
Prior to fMRI, participants underwent an hour-long training session in an fMRI simulator to practice simulated driving. The tasks included straight driving (“Straight Driving,” Figure
Participants received motion training to operate the driving controls with minimal head motion and practiced on four training runs that introduced all driving conditions. During testing, participants were placed in the MRI system with the driving hardware in a comfortable position. A high-resolution anatomical scan was acquired first, followed by three fMRI runs with simulated driving. The driving scenarios were triggered synchronously with the fMRI time series data collection. Each run was approximately 9–9.5 min. Images were acquired using a research-dedicated whole body 3.0 Tesla MRI system (Magnetom TIM Trio software version b15, Siemens, Erlangen, Germany). The high-resolution anatomical scan was acquired with T1-weighted, 3D magnetization prepared rapid gradient echo imaging (MPRAGE; echo time (
The first 10 s of scanning for each driving run was discarded to allow for equilibration effects. Using AFNI freeware (Cox,
Significant brain activations are summarized below, reported relative to the straight driving control condition. Brain activation images for each task are shown in Figures
−24 | −60 | −62 | L superior parietal lobule, BA7 | |
−14 | −78 | 44 | L precuneus, BA7 | |
34 | −40 | 64 | R postcentral gyrus, BA5 | |
20 | −66 | 58 | R superior parietal lobule, BA7 | |
18 | −72 | −8 | R lingual gyrus, BA18 | |
−44 | −30 | 60 | L postcentral gyrus, BA1/2 | |
−6 | −62 | 64 | L precuneus, BA7 | |
−4 | −86 | 38 | L cuneus, BA7 | |
12 | −76 | 56 | R precuneus, BA7 | |
34 | −14 | 66 | R precentral gyrus, BA6 | |
38 | −34 | 64 | R postcentral gyrus, BA1/2 | |
58 | −44 | −34 | R cerebellum (Crus 1) | |
−30 | −28 | 68 | L precentral gyrus, BA4 | |
−28 | −4 | 64 | L superior frontal gyrus, BA6 | |
−8 | −74 | 56 | L precuneus, BA7 | |
−2 | −84 | −18 | L lingual gyrus, BA18 | |
−20 | −72 | −14 | L fusiform gyrus, BA19 | |
−32 | −36 | −50 | L cerebellum (VIII) | |
32 | −12 | 66 | R precentral gyrus, BA6 | |
36 | −34 | 66 | R postcentral gyrus, Ba1/2 | |
34 | −84 | 26 | R superior occipital gyrus, BA19 | |
4 | −96 | 18 | R cuneus, BA18 | |
10 | −76 | 54 | R precuneus, BA18 | |
26 | −92 | −18 | R fusiform gyrus, BA18 | |
30 | −36 | −50 | R cerebellum (VIII) | |
−2 | −98 | −2 | L cuneus, BA18 | |
−6 | −72 | 58 | L precuneus, BA7 | |
−4 | −84 | −20 | L lingual gyrus, BA18 | |
−46 | −70 | −16 | L fusiform gyrus, BA19 | |
−2 | −34 | 74 | L paracentral lobule | |
−32 | −36 | −52 | L cerebellum (VIII) | |
68 | −14 | 0 | R superior temporal gyrus, BA22 | |
34 | −12 | 66 | R precentral gyrus, BA6 | |
36 | −34 | 66 | R postcentral gyrus, BA1/2 | |
2 | −96 | 20 | R cuneus, BA18 | |
24 | −88 | −16 | R fusiform gyrus, BA18 | |
4 | −74 | 54 | R precuneus, BA7 | |
30 | −38 | −52 | R cerebellum (VIII) | |
−66 | −28 | 2 | L middle temporal gyrus, BA21 | |
−64 | −48 | −10 | L inferior temporal gyrus, BA37 | |
−64 | −8 | 6 | L superior temporal gyrus, BA22 | |
−54 | 20 | −6 | L inferior frontal gyrus, BA47 | |
−46 | −56 | 52 | L inferior parietal lobule, BA40 | |
−32 | −76 | 48 | L superior parietal lobule, BA47 | |
54 | 22 | −4 | R inferior frontal gyrus, BA47 | |
68 | −14 | 2 | R superior temporal gyrus, BA22 |
Driving performance showed an effect of speed differences among undistracted driving conditions (
The current study extends previous research by using an immersive fMRI-compatible driving simulator to examine how the human brain responds to various driving conditions, and by characterizing the effects of cognitive distraction on driving. First, we observed that the patterns of brain activation depend on the type of simulated driving task. Performing right turns, the simplest task, generated minimal activation relative to the control condition (Figure
Second, a significant shift in activation from the posterior to the anterior brain was observed when driving became distracted. Compared to straight driving, auditory distraction during straight driving significantly activated not only auditory areas but also the prefrontal cortices (mainly in the ventral lateral prefrontal cortex regions; Figure
To substantiate the observed shift from occipital to frontal brain activations, particularly in the prefrontal areas when comparing the left-turn-traffic condition to the left-turn-traffic plus cognitive distraction, we extracted mean BOLD percentage change values for each subject from the activated occipital and prefrontal regions of interest, and conducted
This anterior-vs.-posterior shift in BOLD signals reflects changing reactions of the brain, and highlights the effect of distracted driving and the role of the anterior frontal region, an area that has been associated with impulsiveness (e.g., Beeli et al.,
Supporting the findings and interpretation of the present work, an observation of decreased activation in parietal-visual areas and impaired driving performance in a dual-task driving condition involving concurrent language comprehension has also been previously reported (Just et al.,
The current finding has important implications regarding distracted driving. While changes in driving performance observed in the undistracted conditions (slowing down from right turns to left turns and traffic) were parallel to the results of brain activations in the posterior brain (increases in activated areas), brain activity shifted to the anterior network when there was no behavioral change from the undistracted to the distracted condition. Eye-tracking studies have shown that hands-free conversations using cell phones impair attention to visual inputs (Strayer et al.,
Previous studies have reported measures that may be able to predict those who passed on-road assessments from those who failed (Baldock et al.,
In addition, while regular driving (e.g., simple right/left turn without traffic) relies on more learned and automatic processes that activate a driving network in the posterior brain, the mid-cingulate cortex was only differentially activated when the driving conditions became more demanding (left turn with oncoming traffic). This brain region has a major role to play in response selection (Paus et al.,
A limitation of the present study is the use of young drivers, which may reduce the generalizability to older populations. Another limitation is that by using simulated driving we were unable to replicate the potential anxiety associated with driving under conditions of increasing complexity, given that there is no real crash risk. Although the use of simulated driving during fMRI may not perfectly generalize to real-world driving, it allows for the investigation of complex driving conditions that are not usually tested during on-road assessments (i.e., left turns during peak traffic or driving while talking on a cell phone). Indeed, previous research suggests that there is a significant correlation between on-road test performance and performance in the driving simulator (Freund et al.,
The present study provides new neuroimaging data of the complex brain activity associated with distracted driving and driving under different levels of complexity. We found that brain activations during driving rely on areas important for various cognitive functions including the posterior visual-spatial attentional system vs. the anterior, frontal-lobe functions in multitasking and divided attention. For most people, driving involves highly practiced skills that generally draw on automatic or practiced abilities relying on a posterior network, and does not heavily require the anterior frontal system for more effortful mental processing. However, there are potentially many distractions present during driving (Editorial,
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This work was supported by a Personnel Award from the Heart and Stroke Foundation of Canada and an Early Researcher Award from the Ontario Ministry of Research and Innovation to Dr. Tom A. Schweizer. Dr. Naglie is supported by the University of Toronto Mary Trimmer Chair in Geriatric Medicine. We would also like to thank Monica Maher, Megan Hird, and Meishan Yan for their assistance in the preparation of the manuscript, and Annette Weekes-Holder for performing fMRI of the volunteers.
1 | Right turn speed |
2 | Left turn speed |
3 | Left turn traffic speed |
4 | Left turn traffic distraction speed |
Right turn speed | 24.0366 | 3.32044 | 16 |
Left turn speed | 26.7939 | 5.16746 | 16 |
Left turn traffic speed | 29.3542 | 4.26414 | 16 |
Left turn traffic distraction speed | 28.9814 | 3.75817 | 16 |
Sphericity assumed | 287.234 | 3 | 95.745 | |
Greenhouse-geisser | 287.234 | 2.579 | 111.382 | |
Huynh–Feldt | 287.234 | 3.000 | 95.745 | |
Lower-bound | 287.234 | 1.000 | 287.234 | |
Sphericity assumed | 324.448 | 45 | 7.210 | |
Greenhouse-geisser | 324.448 | 38.682 | 8.387 | |
Huynh–Feldt | 324.448 | 45.000 | 7.210 | |
Lower-bound | 324.448 | 15.000 | 21.630 |
Sphericity assumed | 13.280 | 0.000 |
Greenhouse-geisser | 13.280 | 0.000 |
Huynh–Feldt | 13.280 | 0.000 |
Lower-bound | 13.280 | 0.002 |
1 | 2 | −2.757 |
1.078 | 0.022 |
3 | −5.318 |
0.820 | 0.000 | |
4 | −4.945 |
0.763 | 0.000 | |
2 | 1 | 2.757 |
1.078 | 0.022 |
3 | −2.560 |
0.997 | 0.021 | |
4 | −2.187 | 1.110 | 0.068 | |
3 | 1 | 5.318 |
0.820 | 0.000 |
2 | 2.560 |
0.997 | 0.021 | |
4 | 0.373 | 0.873 | 0.676 | |
4 | 1 | 4.945 |
0.763 | 0.000 |
2 | 2.187 | 1.110 | 0.068 | |
3 | −0.373 | 0.873 | 0.676 |
1 | 2 | −5.056 | −0.459 |
3 | −7.066 | −3.569 | |
4 | −6.571 | −3.318 | |
2 | 1 | 0.459 | 5.056 |
3 | −4.686 | −0.435 | |
4 | −4.554 | 0.179 | |
3 | 1 | 3.569 | 7.066 |
2 | 0.435 | 4.686 | |
4 | −1.489 | 2.234 | |
4 | 1 | 3.318 | 6.571 |
2 | −0.179 | 4.554 | |
3 | −2.234 | 1.489 |
Descriptive statistics | ||||||
---|---|---|---|---|---|---|
Straight speed | 16 | 54.37 | 67.78 | 58.5669 | 3.36209 | 0.84052 |
Straight distraction speed | 16 | 55.67 | 63.50 | 58.6931 | 2.33586 | 0.58397 |
Straight lane position | 16 | 1.91 | 2.95 | 2.3519 | 0.31327 | 0.07832 |
Straight distraction position | 16 | 1.82 | 3.26 | 2.5069 | 0.42180 | 0.10545 |
Pair 1 | Straight speed—straight distraction speed | −0.12625 | 2.91779 | 0.72945 |
Pair 2 | Straight lane position—straight distraction position | −0.15500 | 0.34065 | 0.08516 |
Pair 1 | Straight speed—straight distraction speed | −1.68103 | 1.42853 |
Pair 2 | Straight lane position—straight distraction position | −0.33652 | 0.02652 |
Pair 1 | Straight speed—straight distraction speed | −0.173 | 15 | 0.865 |
Pair 2 | Straight lane position—straight distraction position | −1.820 | 15 | 0.089 |