Analysis of Electroencephalograms of Children with ASD During the

Driving Game

Min Lei

, Rongrong Wang

, Yasong Du

and Yi Liu

School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

Shanghai Nuo Cheng Electric Co., Ltd, Shanghai 200245, China

Mental Health Center Affiliated to Medical School of Shanghai Jiao Tong University, Shanghai 200030, China

Keywords: Autism Spectrum Disorder (ASD); EEG; Virtual Driving; Power Spectrum; Wired and Wireless Data

Acquisition.

Abstract: Autism spectrum disorder (ASD) is a neurodevelopmental disorder. Its core symptoms are social

communication and communication disorder, narrow interest, stereotyped behavior and cognitive impairment.

The ASD people also shows atypical characteristics in non-social information processing. These defects are

serious to weaken their social adaptability. Driving is a complex and simple independent adaptive skill,

involving not only multiple task operations at the same time, but also attention, automatic motor execution,

memory and navigation of cognitive functions. At present, the researches focus more and more on the driving

behavior for individuals with autism in order to explore the feasibility of intervention treatment. The aim of

this paper is to use a virtual driving test to investigate electroencephalogram (EEG) characteristics of the

children with autism during complex environments. Since it is easy for EEG records to contain significant

motion artifacts and electrode artifacts, an acquisition method with the wired and wireless transmission is

used to collect EEG. This paper applies the power spectrum method to analyse the recorded EEG. It is found

that the recorded EEG can reflect the brain dynamical activity of subjects during driving. The results indicate

that this acquisition method can be used to record the EEG during driving for the brain analysis of autism.

This study provides insights for the further research on the mechanism of autism and its diagnosis, evaluation

and intervention.

1 INTRODUCTION

The research on driving behavior of people with

autism can be traced back to the late 1970s. In early

studies, the questionnaires for investigation found that

traffic accidents more likely increased for individuals

with attention deficit hyperactivity disorder during

driving. In recent years, the research on driving

behavior of autistic people has gradually increased.

Sheppard et al. (2010) found that young autistic

individuals lack the ability to identify obvious driving

risks when watching driving video clips, comparing

with healthy individuals (Sheppard et al., 2010). It is a

great challenge for autistic individuals to learn

driving, especially dealing with multitasking in

driving. However, autistic children can do well in

simple situations (such as speed control and keeping

driving) (Cox et al. 2012). For high functional autistic

adolescents, 12% of them received driving tickets or

involved in vehicle collisions. The percentage was

lower than that of ordinary adolescents. And there are

no differences in driving status or driving behavior in

terms of gender, type of autism, parental age or

education or access to public transport (Huang et al.

2012). Lindsay et al. (2016) pointed out that it is very

necessary to train the driving skills of autistic people

as well as develop the suitable transportation for

people with autism (Lindsay, 2017). However, it is

difficult to carry out the real driving experimental

researches for autistic individuals, due to a lot of

potential safety hazards in the real driving

environment.

With the development of science and technology,

virtual driving environments provide convenience for

people with autism (Wade et al. 2016). And many

autistic teenagers prefer to interact with machines,

such as robots and video games in virtual reality

environment, rather than people (Tanaka et al. 2010;

Zheng et al. 2014). The virtual driving games may be

more conducive to train the adaptive independence of

526

Lei, M., Wang, R., Du, Y. and Liu, Y.

Analysis of Electroencephalograms of Children with ASD During the Driving Game.

DOI: 10.5220/0011956300003612

In Proceedings of the 3rd International Symposium on Automation, Information and Computing (ISAIC 2022), pages 526-530

ISBN: 978-989-758-622-4; ISSN: 2975-9463

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

autistic individuals. Monahan et al. (2013) used the

driving simulation system with the real vehicle

operation to evaluate and analyse the behaviors

individuals with attention deficit hyperactivity

disorder (ADHD) and autism spectrum disorder

(ASD) under the guidance of occupational therapists.

It was found that ADHD / ASD adolescents made

more mistakes in keeping the direction, visual sense,

acceleration during driving than healthy subjects

(Monahan et al. 2013). Classen et al. used professional

driving rehabilitation experts to comprehensively

evaluate the simulated driving of 7 autistic individuals

and 22 healthy controls. The study found that young

autistic people performed worse in driving skills or

related driving skills (such as cognition, visual motor

integration, motor coordination, speed regulation,

road maintenance notices or signs) (Classen et al.

2013). Cox et al. reported the simulation driving

performance research of novice drivers with autism.

But on the whole, their driving ability is lower than

that of healthy subjects (Cox et al. 2016). Reimer et al.

(2013) used the driving simulation system to conduct

a comparative study between young people with high-

functioning autism and normal people. For individuals

with high functioning autism, the gaze of them was

higher in the vertical direction and a little right in the

horizontal dimension. However, they did not find a

significant difference in driving performance between

the two groups, except gaze states. The authors

thought that the gaze deflection of the young people

with high functioning autism may be a dangerous

driving behavior in actual driving (Reimer et al. 2013).

Wade et al. (2014) also found similar results using the

virtual driving simulation system (Wade et al. 2014).

Furthermore, Brooks et al. (2016) found that the

driving simulation system as a training tool can enable

autistic subjects to achieve the same good driving

ability as the control group (Brooks et al. 2016).

With the behavior researches of autistic

individuals based on the virtual driving system, some

scholars also have carried out research on the analysis

of physiological signals for autistic individuals during

virtual driving states, such as electrodermal activity-

EDA, heart rate, eye movement, EMG, RSP,

peripheral temperature, and so on. Fan et al. (2015)

used the spectral analysis of EEG signals to study

emotional states of autistic subjects during the virtual

driving test (Fan et al. 2015). For the driving

evaluation and intervention research on autism, Zhang

et al. (2015) used eye movement analysis to study the

effect of the difficulty level of cognitive load during

driving on autistic individuals (Zhang et al. 2015). Lei

Min et al. (2016) explored the characteristics of brain

activity of autistic children during driving by

analysing the sample entropy values of EEG signals

for autistic children during simulated driving (Lei et

al. 2016). There are few studies on the physiological

signals of autistic individuals during driving,

especially the research on the electrical signals of

brain activity (such as EEG, magnetoencephalogram

and so on). At present, there have been many studies

on the brain activity of autistic individuals (Menaka et

al. 2021; Abdulhay et al. 2020; Wang et al. 2014; Zhu

et al. 2014). However, due to measurement limitations

of EEG signals, especially requiring the head and

body not to move, it is difficult for autistic children to

record the EEG signals during movements in the

complex environments. Therefore, the experimental

designs of EEG research on autistic EEG are often

very simple and single, such as the close or open eyes

in the resting state, picture recognition (Hashemian et

al. 2014; Ahmadlou et al. 2012; Bosl et al. 2011;

Catarino et al. 2011). There is still a lack of research

on the brain dynamical activity of patients with autism

under complex multi task stimulation. This paper aims

to apply the virtual driving environment to study the

feasibility of electroencephalogram (EEG) of autistic

children during driving a car.

2 METHODS

2.1 Virtual Driving Environment and

Data Collection

The study was conducted in a virtual driving

environment consisting of a driving simulator, two

computer screens, a computer, and City Car Driving

software. The driving simulator is a Logitech G29

driving game device including steering wheel, pedal,

gear lever, driving software and development kit

(SDK). For two screens, one is a monitor for doctors

and technicians. The other is for subjects to watch a

virtual driving roadway. The computer is used to

collect the EEG data, display and monitor data online,

as well as subsequently deal with data and so on. In

order to reduce the interference of EEG data by noise,

the experiment adopts the combination of wired and

wireless data transmission. According to the

international standard 10-20 electrode placement

system, the 16 lead EEG signals are recorded from

EEG electrode cap for children into the EEG amplifier

equipment (Type: Nation-BTV, manufactured by

Shanghai NuoCheng Electric Co., Ltd.) through a set

of the short-distance wire lines. The sampling

frequency is 256Hz. Then, the recorded EEG data are

sent to the computer by the wireless transmission

format (see Figure 1). This wired and wireless data

Analysis of Electroencephalograms of Children with ASD During the Driving Game

527

acquisition method can effectively reduce the

interference of wire motion in data acquisition.

Figure 1: Schematic diagram of the wired and wireless

transmission for EEG signals during driving.

2.2 Participants

Three boys about 13 years old in this study were

diagnosed with the high functioning autism by a

psychiatrist in Shanghai Mental Health Center in

terms of DSM-V diagnostic criteria. They had IQs 70

or greater (WISC-IV), right-handed. The subjects did

not suffer from organic mental disorders,

schizophrenia, personality disorders and other mental

diseases, nor did they suffer from nervous system

degenerative diseases, brain trauma or

cerebrovascular diseases, and had no history of major

physical diseases such as severe heart, liver and

kidney dysfunction or drug dependence. The subjects

did not take any psychotropic drugs before this

experiment. The guardians of the children signed an

informed consent form.

2.3 Experiment

All subjects first were familiar with the driving

environment and operation under the guidance of

technicians (see Figure 2). Then, they took about 10

minutes of training to habituate to the driving

simulator. The driving period mainly included

startup, road driving, parking with no on-coming

traffic. The speed limit was gradually increased to 50

KMH.

2.4 Data Analysis

Since the reference electrodes could be contaminated

by artifacts, the average of all channel data per subject

is used as a common reference. The EEG data first

were offline re-referenced to the average reference.

Then, EEG signals were filtered into the frequency

band 0.2 to 50 Hz in order to remove the artifacts of

eye movement, eye blink and muscle activity.

Meanwhile, the spikes in EEG signals were also

removed. The length of the analysed data is 5 minutes

of driving state. EEG power can be measured by the

Welch’s power spectral density estimate method. The

analysed data were segmented into one-second

epochs with 75% overlap.

Figure 2: Experimental environment.

3 RESULTS

In this paper, the absolute power spectral density of

each channel is calculated in order to verify the

feasibility of EEG analysis in this study. Figure 3

shows the EEG signals of 16 channels in one second

period during the driving state. It can be seen that

from the time domain, the EEG data can be used.

Figure 4 shows the power spectral characters of all

channels for three subjects. As can be seen in Figure

4, spectrum ridge ranges are in the alpha band. The

results indicate that it is feasible to apply the EEG

acquisition system of this study in the virtual driving

experiment.

Moreover, we compute mean absolute power

spectral density (μV2) in alpha and beta frequency

bands, as well as the peak frequency (Hz) in alpha and

beta frequency bands. In Figure 5, we can see that the

spectral energy of the prefrontal region is higher than

those of other brain functional regions for the alpha

and beta frequency bands. However, the maximum

peak frequency can not be in the prefrontal region

(see Figure 6). In the alpha band, the power spectrum

peak frequencies in C4, T4, P4, T6, O2 locations are

greater for three ASD subjects. In the beta band, the

peak frequency in O2 location is lower for each ASD

individuals. Although these results can not meet the

statistical requirements due to the small number of

subjects, the results can still provide reference for the

sequential study of brain dynamic activities in virtual

driving environment.

ISAIC 2022 - International Symposium on Automation, Information and Computing

528

Figure 3: One second epochs of EEG signals for three

subjects.

Figure 4: The power spectrum of EEG signals for three

subjects.

(a)

(b)

Figure 5: The topographic maps of the mean absolute power

spectrum of EEG signals for three subjects: (a) the average

in the alpha frequency band; (b) the average in the beta

frequency band.

(a)

(b)

Figure 6: The topographic maps of the power spectrum

peak frequencies of EEG signals for three subjects: (a) the

peak frequency in the alpha band; (b) the peak frequency in

the beta band.

4 CONCLUSIONS

This paper applies the wired and wireless data

acquisition system to collect the

electroencephalogram (EEG) characteristics of the

children with autism during a virtual driving test. The

recorded EEG signals are analyzed by the spectral

analysis method. The results find that from the time

and frequency domains, the recorded EEG signals can

be used. Moreover, the higher spectral energies of the

prefrontal region in the alpha and beta frequency

bands are shown for three subjects during driving.

The alpha peak frequencies in C4, T4, P4, T6, O2

locations are higher than those in other locations. The

beta peak frequencies are lower in O2 locations of all

subjects. These results indicate that the acquisition

method with the wired and wireless transmission can

collect the EEG signals of the individuals with autism

during driving for the autistic brain dynamic analysis.

This study provides an experimental and analysis

basis for further study on autism.

ACKNOWLEDGEMENTS

Authors wish gratefully to acknowledge Mandy Chen

for her patience and enthusiastic help. This work was

supported by Shanghai "Science and technology

innovation action plan" bio-medicine science and

technology support project (Grant No.

19441907400).

REFERENCES

Abdulhay E, Alafeef M, Alzghoul L, Momani MA, Abdi

RA, Arunkumar N, Munoz R, Albuquerque V 2020

Computer-aided autism diagnosis via second-order

difference plot area applied to EEG empirical mode

decomposition Neural Comput Appl 32 10947–10956

Ahmadlou M, Adeli H, Adeli A 2012 Improved visibility

graph fractality with application for the diagnosis of

Autism Spectrum Disorder. Physica A 391(20) 4720-

4726

Bosl W, Tierney A, Tager-Flusberg H, Nelson C 2011 EEG

complexity as a biomarker for autism spectrum disorder

risk BMC Med 9 18

Brooks J, Kellett J, Seeanner J, Jenkins C, Buchanan C,

Kinsman A, Kelly D, Pierce S 2016 Training the motor

aspects of pre-driving skills of young adults with and

without autism spectrum disorder J Autism Dev Disord

46 2408–2426

Catarino A, Churches O, Baron-Cohen S, Andrade A, Ring

H 2011 Atypical EEG complexity in autism spectrum

Analysis of Electroencephalograms of Children with ASD During the Driving Game

529

conditions: A multiscale entropy analysis Clin

Neurophysiol 122(12) 2375-2383

Classen S, Monahan M 2013 Indicators of simulated

driving skills in adolescents with autism spectrum

disorder Open J Occupat Therap 1(4) Article 2

Cox N B., Reeve R E., Cox S M., and Cox D J 2012 Brief

Report: Driving and young adults with ASD: Parents’

experiences J. Autism Dev. Disord 42(10) 2257–2262

Cox S M, Cox D J, Kofler M J, Moncrief M A, Johnson R

J, Lambert A E, Cain S A, Reeve R E 2016 Driving

simulator performance in novice drivers with autism

spectrum disorder: the role of executive functions and

basic motor skills. J Autism Dev Disord. 46 1379-1391

Fan J, Wade J W, Bian D, Key A P, Warren Z E, Mion L C,

and Sarkar N 2015 A step towards EEG-based Brain

computer interface for autism intervention 37th Ann Int

Conf IEEE Engin in Med and Biol Soc (IEEE) pp 3767-

3770

Hashemian M, Pourghassem H 2014 Diagnosing Autism

Spectrum Disorders Based on EEG Analysis: a Survey

Neurophysiol 46(2) 183-195

Huang P, Kao T, Curry A E, Durbin D R 2012 Brief Report:

Factors associated with driving in teens with autism

spectrum disorders J Dev Behav Pediatr 33 70 –74

Lei M, Meng G, Zhang WM, Sarkar N 2016 Sample

entropy of electroencephalogram for children with

autism based on virtual driving game Acta Physica

Sinica 65(10) 108701-1-12 (In chinese)

Lindsay S 2017 Systematic review of factors affecting

driving and motor vehicle transportation among people

with autism spectrum disorder Disabil Rehabil 39(9)

837-846

Menaka R, Aarthy MT, Chavan RM, Perumal RC, Menon

MS 2021 Recurrence Quantification Analysis of EEG

signals for Children with ASD J Sci Ind Res India 80(5)

438-448

Monahan M, Classen S, Helsel P 2013 Pre-driving

evaluation of a teen with attention deficit hyperactivity

disorder and autism spectrum disorder Can J Occupat

Therap 80(1) 35-41

Reimer B, Fried R, Mehler B, Joshi G, Bolfek A, Godfrey

K, Zhao N, Goldin R, Biederman J 2013 Brief report:

examining driving behavior in young adults with high

functioning autism spectrum disorders: a pilot study

using a driving simulation paradigm J Autism Dev

Disord. 43 2211–2217

Sheppard E, Ropar D, Underwood G, Loon E 2010 Brief

report: driving hazard perception in autism J Autism

Dev Disord 40 504-508

Tanaka J W, Wolf J M, Klaiman C, Koenig K, Cockburn J,

Herlihy L, Brown C, Stahl S, Kaiser M D, and Schultz

R T 2010 Using computerized games to teach face

recognition skills to children with autism spectrum

disorder: The Let’s Face It! Program. J. Child Psychol.

Psychiat. 51(8) 944–952

Wade J, Bian D, Zhang L, Swanson A, Sarkar M, Warren

Z, Sarkar N 2014 Design of a Virtual Reality Driving

Environment to Assess Performance of Teenagers with

ASD UAHCI/HCII Vol 2 Universal Access in Human-

Computer Interaction Eds. Stephanidis C and Antona M

(Springer International Publishing Switzerland) p466–

474

Wade J, Zhang L, Bian D, Fan J, Swenson A, Weitlauf A,

Sarkar M, Warren Z, Sarkar N 2016 A gaze-contingent

adaptive virtual reality driving environment for

intervention in individuals with autism spectrum

disorders ACM Trans Interact Intell Syst 6(1) 1-23

Wang J, Barstein J, Ethridge LE, Mosconi MW, Takarae Y,

Sweeney JA 2014 Resting state EEG abnormalities in

autism spectrum disorders J Neurodev Disord 5 24

Zhang L, Wade J, Bian D, Fan J, Swanson A, Weitlauf A,

Warren Z, and Sarkar N 2015 Multimodal fusion for

cognitive load measurement in an adaptive virtual

reality driving task for autism intervention Int Conf

Univers Acc in Human-Comp Interact Vol 3 Universal

Access in Human-Computer Interaction Eds. Antona M

and Stephanidis C (Los Angeles USA) (Springer)

p709–720

Zheng Z, Das S, Young E M., Swanson A, Warren Z, and

Sarkar N 2014 Autonomous robot-mediated imitation

learning for children with autism Proc. IEEE Int. Conf.

on Robotics and Automation (Hong Kong, China) IEEE

pp 2707–2712

Zhu X, Song Y, Bi H 2014 Researches on Atypical

Lateralization of Autisms Language Development Prog

Biochem Biophys 41(8) 749-754 (In chinese)

ISAIC 2022 - International Symposium on Automation, Information and Computing

530