APJNT Journal · Areej Babiker, Ibrahima Faye, Aamir Saeed Malik 22 . Dr K. Jayasankara Reddy /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, ... number of studies with

i

© 2020 Asia Pacific Journal of Neurotherapy (APJNT)

APJNT Journal

Table of Contents

Title Authors Page#

1 EEG Neurofeedback Training for Mental

Health Issues of Prisoners

Dr K. Jayasankara Reddy, Adithya

Ramesh

1

2 Neurofeedback and Counseling as Integrative

Treatment

See Ching Mey 6

3 A brief literature survey into the effects of

coal seam gas exploration by-products on

Human EEG

Herbert F. Jelinek, Lynda Olling 12

4 Neural Network for Detection of Students’

interest using Approximate Entropy Features

of EEG data

Areej Babiker, Ibrahima Faye, Aamir

Saeed Malik

22

Dr K. Jayasankara Reddy /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, No 1, pages 001-005 1


EEG Neurofeedback Training for Mental Health Issues of Prisoners

Dr K. Jayasankara Reddya*, Adithya Rameshb

a,bDepartment of Psychology, CHRIST (Deemed to be University), Hosur Road, Bengaluru – 560029, India.

Email: [email protected]

Abstract

The EEG neurofeedback training is an upcoming intervention concerning many disorders such as Attention-Deficit/Hyperactivity

Disorder, Schizophrenia and Depression as well. The main aim of this review was to study the efficacy of this intervention on a vulnerable

population like prisoners. Papers for this review were selected from established databases like PubMed, ProQuest, and ScienceDirect among many.

The use of neurofeedback in the forensic psychiatric setting did not show a healthy level of applicability as expected. The main reasons for this

were the variations in the sample characteristics between the studies, along with the motivation of the prisoners as well. The prison setting also

delayed and interfered with the training and could also be alluded to its apparent difficulty in the application of this treatment. There are a small

number of studies with even smaller samples in a population that is difficult to engage. These conditions have led to inconclusive results in most

cases. There is scope for improvement, and this is an upcoming area of research. Future directions are also discussed.

© 2020 Published jointly by ANSA and APNA societies of Australasia and ASEAN Countries. Selected and peer-review by

editorial board of Asia Pacific Journal of Neuro therapy (APJNT).

Keywords: Neurofeedback, prisons, neurobiology, interventions, & protocols.

ARTICLE INFO

RECEIVED January 10, 2020

REVIEWED March 17, 2020

ACCEPTED March 27, 2020

* Corresponding author.


mailto:[email protected]




1. Introduction

“It is said that no one truly knows a nation until one has been inside its jails. A nation should not be judged by how it treats

its highest citizens, but its lowest ones.” Nelson Mandela (p. 115, 1994)

The Governments in the interest of public safety and upholding the laws of the land takes the responsibility of prosecuting

and meting out punishments as well. In most cases, these are in the form of prison sentences that might vary from being a few years

to a lifetime. More often than not, it is believed that by merely isolating the criminals from society, would in a way, take care of the

circumstances that lead to the crime in the first place. There is not much done in the prisons in terms of rehabilitation and this has

led to an increase in recidivism. The national average in India stands at 10.4% which is also likely to fluctuate between states and

union territories (Yadav, 2014). There are attempts at rehabilitation, but these are not effective as these are often viewed as

unnecessary by the officials themselves. Rehabilitation, at its core, must be a system to correct the wrongs committed by the

criminals and give them the skills and treatment that will facilitate a smooth transition into the general community. The provision

of passing qualifying exams in an attempt to educate them and setting up vocational training centres is a good step in the right

direction. These vocations, however, tend to involve skills that might not sufficiently interest the prisoners.

The prison system also neglects to cater to the interests and well-being of the prisoners, especially when it comes to their

mental health. Syed and Raghavan (2018) noticed a dismal trend when it came to studies in this area. They classify prisoners as a

minority and have highlighted the prevalence of chronic mental illnesses like schizophrenia, depression and adjustment issues along

with suicide risk as well. In terms of the level of intervention that was implemented within the prisons, it was found that yoga was

useful when it came to dealing with anxiety and depression. However, the need for newer techniques was also emphasised, along

with mental health clinics and trained professionals within the prison system itself. The current scenario in India usually involves a

psychiatrist from a government hospital who visits prisons once a month. This scenario is mostly for medications and treatment for

people with substance use disorders. Any emergencies in terms of mental health are usually referred directly to the corresponding

hospital. The high rates of recidivism and mental health could be tackled adequately with the introduction of techniques such as

neurofeedback training.

Neurofeedback training (NFT) is a technique which teaches clients to control their brain waves and in turn alter brain

functioning. This training operates on the principle of operant conditioning and uses electroencephalography (EEG). The client is

given feedback on their training in real-time, and this shows them the progress that they make as well. Based on the brain waves

targeted, many protocols are available. These can also be tailor-made to precisely certain brain areas as well depending on the goal

of the NFT. NFT has also been effective in the treatment of disorders such as attention-deficit/hyperactivity disorder, schizophrenia,

drug addiction, depression, anxiety and even pain management. Each of these disorders tends to affect the brain and its waves

differently, and the protocols are designed to counter these effects (Marzbani, Marateb, & Mansourian, 2016).

2. Objectives

The main objectives of this review article are to review papers on neurofeedback and their efficacy in dealing with mental

illnesses and antisocial behaviour among the prisoners. This would establish a neurobiological technique such as NFT as a useful

alternative for rehabilitation in prison systems along with the pre-existing mechanisms of vocational training and providing

education as well.

3. Methods

A traditional literature review was done using keywords such as neurofeedback training, prisoners, antisocial behaviour,

and mental health in prisons. Papers, as a result, were reviewed from PubMed, ScienceDirect, Elsevier and Taylor and Francis,

among others. It included papers from journals such as Basic Clinical Neuroscience, International Journal of Forensic Mental Health,

Indian Journal of Social Psychiatry and the Journal of Criminal Justice, among others.

4. Results

4.1 Efficacy of Neurofeedback in Forensic Psychiatric Settings

Neurofeedback is increasingly being used to tackle many types of disorders in the clinical psychiatric setting from anxiety

disorders to schizophrenia. However, there is no similar trend when it comes to its usage in forensic psychiatry. The presence of a

co-morbidity may counteract with other treatments but would not be treated as a contraindication for NFT (Holtmann, Albrecht, &

Brandeis, 2018). Moreover, it is the main reason why it is not frequently used in forensic settings. The forensic psychiatric

population includes people who are not wholly responsible for their crimes. Part of the blame lies in the mental illness as well.

Fielenbach, Donkers, Spreen and Bogaerts (2018) studied the effects of Neurofeedback on forensic psychiatric patients, about 19 of

them with substance use disorder along with any other comorbidity as well. Primarily, in her protocol was to test the ability to reduce

impulsivity and restore inhibitory controls within them through increasing SMR and reducing theta. To ascertain whether there was

any change due to the NFT, the EEG magnitudes must have changed from the baseline by about 8%. At the face of it, although there

were improvements in their impulsivity and craving due to increasing SMR, there were low respondents to the same. This was

attributed to the low number of sessions along with the continued use of their prescriptions as well.

Neurofeedback has opened up the possibility of neurobiological approaches to the field of forensic psychiatry and

correctional facilities. At present, it does not enjoy widespread attention from psychologists, primarily due to the overemphasis on

psychosocial factors of antisocial behaviour and psychopathy. However, traditional forms of psychotherapy are plagued with



problems such as low compliance rates and in multicultural settings, lack of proper communication. This leaves the stage open for

newer approaches such as neurofeedback. Neurofeedback therapy proved to be instrumental in dealing with cases of aggression and

anti-social behaviour as a result of substance abuse (van Outsem, 2011). It was effective in reducing the craving as well as abstinence

was also prolonged. In cases of domestic violence, which are precipitated by the fear of abandonment and pathological jealousy,

neurofeedback was used to enhance flexible thinking and emotional response as well (van Outsem, 2011). When it comes to prisoners, there is the aspect of criminal offending, recidivism, reoffending, aggression, violence, and

along with disorder associated with the same like ADHD, schizophrenia, psychosis, cluster B personality disorders, psychopathy

and substance use disorder. An important aspect when it comes to EEG-NFT is that there has to be actual learning for it to work.

This is a criterion that takes prime importance when it comes to studies which employ this intervention. A review paper found that

out of a total of 224 articles screened there were only ten that specifically looked at EEG and also included the details on the protocol

used along with a specific criterion for EEG learning. Impulsivity and difficulties with inhibitory control did seem to improve, along

with hostility and drug use. Recidivism could also be reduced if these symptoms are reduced as well. However, it was uncertain

whether improvements in brain waves are always correlated with the behavioural measures utilised. The possibility of randomised

control trials might help ascertain the effects of neurofeedback (Fielenbach, Donkers, Spreen, Visser & Bogaerts, 2018). The most

interesting study of these 10 was one which looked to study the aspect of brain self-regulation in criminal psychopaths. It utilised

slow cortical potentials demonstrated that the psychopaths were able to control brain excitability. There were reduced aggression,

impulsivity and behavioural tendencies. There are few psychophysiological treatments as intervention, and neurofeedback continues

to be the leading candidate for reducing psychopathic characteristics such as disinhibition, aggression and related behavioural

approach tendencies. This was also reflected in self-report measures for physical aggression, although reactive aggression and

aggression inhibition were not significant (Konicar et al., 2015).

4.2 Use of SMR and SCP training for Forensic Psychiatry and Substance Use Disorder

Fielenbach, Donkers, Spreen and Bogaerts (2017) explored the use of the sensorimotor rhythms protocol for reducing

impulsivity and by extension, the use of substances among the forensic psychiatry population. Impulsivity has been associated with

drug use and also a likely cause of it. The use of SMR protocols was expected to significantly reduce impulsivity and the level of

craving that the participants experienced. This protocol was further evaluated along with theta waves. The substance use and

psychiatric disorders in the forensic psychiatric population increase the likelihood of violence in the future. Hence, it is necessary

to address the lack of inhibitory control in the population. According to the protocol, the SMR frequency was enhanced, while

inhibiting the theta frequency. The Barratt Impulsivity Scale, along with a Go/No-Go reaction time task, was used to measure

impulsivity. The population was male and had been undergoing treatment for their illness as well. Therefore, the neurofeedback

sessions were accompanied by treatment as usual (TAU) in one grouping and just TAU in the other. However, this combination did

not seem to be better in any sense (Fielenbach, Donkers, Spreen & Bogaerts, 2018).

In an attempt to isolate the effects of the neurofeedback, a clinical case series was done. The benefit of this is that individual

differences can be strictly observed while still retaining the validity and stringent standards of an RCT. The control was done with

a sham neurofeedback protocol. The theta/SMR protocol was used once again. While there was evidence that there was a decrease

in the self-reported scores of impulsivity and craving, it could still not be attributed exclusively to the neurofeedback training. All

connections to the intervention were purely speculative. However, there were no significant differences found in the study. This was

especially concerning the two patients that underwent the real neurofeedback protocol. While one of them reported a reduction in

their impulsivity and level of craving, this was present at baseline as well. The other patient, despite having undergone the same

training, did not appear to have a similar reaction to the intervention. This was the result after a total of 8 sessions (Fielenbach,

Donkers, Spreen, Smit, & Bogaerts, 2019). This raises several questions concerning whether the intervention can adequately be

applied to this population. The type of externalising disorder, its severity, along with the treatment in progress, seems to be a

confounding variable when it comes to testing the efficacy of this intervention.

5. Summary of Studies: Pros and Cons

Table 1. Pros and cons of studies

Authors Pros Cons

Caria, Sitaram, Veit,

Begliomini, &

Birbaumer (2010)

It used a new approach – real-time functional magnetic resonance

imaging to assess the role of the insula in feedback.

Small sample used

Fielenbach, Donkers,

Spreen, Smit, &

Bogaerts (2019)

Utilised EEG neurofeedback in a single case experimental design to

control impulsivity and craving

Improvements on behavioural reports were not

attributable to neurofeedback intervention

Fielenbach, Donkers, Spreen, & Bogaerts

(2017)

Development of an SMR neurofeedback protocol that is designed for forensic patients

The N-of-1 approach may help explain effects encountered to the protocol but loses generalisability

Fielenbach, Donkers, Spreen, & Bogaerts

(2019)

Use of SMR and Theta protocols for craving and impulsivity. There were improvements in the self-reported measures for the same.

Only the craving improvements could be attributed to the neurofeedback training. N=19 means that there is a low

scope for generalisability.


Spreen, & Bogaerts (2018)

Use of SMR/Theta Protocol to reduce craving and impulsivity in

substance abuse disorders. A randomised control trial was used with the two groups being neurofeedback and treatment as usual.

There were changes in the level of craving with an

increase in SMR activity on an individual level, but this was not found on the group level.



Shepard (2015) Review paper showed that there were specific protocols like the

Peniston Protocol, its Scott-Kaiser modification, and the QEEG that are more helpful. Of these, Scott-Kaiser modification has the highest

rates of success.

This review affirms the use of neurofeedback therapy as a

means of treatment, especially for substance use disorders, but does not clarify its use with forensic

patients.

Konicar et al. (2015) Psychopathic offenders can gain control of brain excitability in the Fronto-central region after intense brain regulation intervention using

the SCP protocol.

Possibility of practice effects on the Flanker task that was used in the study. A small sample was used and there

were not adequate control conditions.


Spreen, Visser, & Bogaerts (2017)

There is a possibility that Neurofeedback techniques can be applied to

criminal offender populations.

Difficulty in assessing the modality of the neurofeedback

to be used in most cases. Single case experimental designs more beneficial than double-blind, randomised controls

6. Conclusion and Future Directions

The effects of the neurofeedback intervention do not seem to lend itself to the natural adaptation to the forensic psychiatric

population as expected despite having a high level of success otherwise in terms of higher rates of abstinence and lower levels of

depression. This comparison is when treated with traditional forms of intervention, such as psychotherapy. This could also be

attributed to the protocols used as they differ between researchers as well. The Peniston Protocol uses a form of alpha/theta training,

while the articles reviewed focused on SMR instead (Shepard, 2015). There has been much evidence for the use of neurofeedback

as an excellent alternative to traditional practices (Marzbani, Marateb, & Mansourian, 2016; van Outsem, 2011). However, this is

not easily visible when it comes to the prison system and the forensic psychiatric population. Some of the significant issues in this

population is the prevalence of impulsivity and substance use disorders. While many designs and protocols of the Neurofeedback

therapy deal with similar issues such as aggression and substance use disorders, they did not show any significant difference pre-

and post-intervention. More importantly, these were not restricted to the effects of the neurofeedback itself (Fielenbach, Donkers,

Spreen, Smit, & Bogaerts, 2019; Fielenbach, Donkers, Spreen, Visser & Bogaerts, 2018). Other issues which came were the fact

that the intervention did not seem exciting enough to keep the prisoners engaged. The number of sessions was also variable between

the studies. Some of these were as short as eight and rarely went over 20. This was also attributed to the low level of compliance

and motivation to participate in the study.

Along with this, the treatment that the participants were already undergoing did not stop either. This became a confound

within the experiments. The criterion of EEG must be used; otherwise, any significant positive changes would be not attributed to

the use of neurofeedback therapy. The use of behavioural measures meant that the results were susceptible to participant’s bias. The

future course of action can be directed towards increasing the number of sessions and utilising games or animations that are tailored

to the population. This would also contribute to keeping the participants engaged in the intervention.

Furthermore, the use of neurofeedback means that the results are not readily visible, hence the improvements must be

communicated between sessions to retain the participants. Finally, it has to be considered that the composition of the Forensic

Psychiatric Population is varied and it might also lack the necessary training to learn neurofeedback itself (see Fielenbach, Donkers,

Spreen, Smit, & Bogaerts, 2019; Caria, Sitaram, Veit, Gaber, Ruiz & Birbaumer, 2010). There were instances where the participants

had to be directed to participate rather than stare blankly at the screen actively. The use of a sample that can be standardised to the

extent that even the TAU is constant would help emphasise the intervention capability of the neurofeedback training.

References

Caria, A., Sitaram, R., Veit, R., Begliomini, C., & Birbaumer, N. (2010). Volitional control of anterior insula activity modulates the response to aversive stimuli. A real-time functional magnetic resonance imaging study. Biological psychiatry, 68(5), 425-432.

Fielenbach, S., Donkers, F. C. L., Spreen, M., Smit, A., & Bogaerts, S. (2019). Theta/SMR Neurofeedback Training Works Well for Some Forensic

Psychiatric.Patients, But Not for Others: A Sham-Controlled Clinical Case Series. International journal of offender therapy and comparative

criminology,0306624X19849562.

Fielenbach, S., Donkers, F. C., Spreen, M., & Bogaerts, S. (2017). Neurofeedback as a Treatment for Impulsivity in a Forensic Psychiatric Population with Substance

Use Disorder: Study Protocol of a Randomized Controlled Trial Combined With an N-of-1 Clinical Trial. JMIR research protocols, 6(1), e13.

doi:10.2196/resprot.6907

Fielenbach, S., Donkers, F. C., Spreen, M., & Bogaerts, S. (2018). The ability of forensic psychiatric patients with substance use disorder to learn

neurofeedback. International Journal of Forensic Mental Health, 1-13.

Fielenbach, S., Donkers, F. C., Spreen, M., & Bogaerts, S. (2018). Effects of a Theta/Sensorimotor Rhythm Neurofeedback Training Protocol on Measures of

Impulsivity, Drug Craving, and Substance Abuse in Forensic Psychiatric Patients with Substance Abuse: Randomized Controlled Trial. JMIR mental

health, 5(4),e10845. doi:10.2196/10845

Fielenbach, S., Donkers, F.C., Spreen, M., Visser, H. A., & Bogaerts, S. (2018). Neurofeedback Training for Psychiatric Disorders Associated with Criminal

Offending: A Review. Frontiers in psychiatry, 8, 313. doi:10.3389/fpsyt.2017.00313

Holtmann, M., Albrecht, B., & Brandeis, D. (2018). Neurofeedback. In T. Banaschewski, D. Coghill, & A. Zuddas (Eds.), Oxford Textbook of Attention Deficit

Hyperactivity Disorder (pp. 366-372). Oxford.

James, C. (2015). Neurofeedback Training for Substance Use Disorders: A Review of the Applicability in Treatment. Ideas and Research You Can Use VISTAS 2015.

American Counseling Association

Konicar, L., Veit, R., Eisenbarth, H., Barth, B., Tonin, P., Strehl, U., & Birbaumer, N. (2015). Brain self-regulation in criminal psychopaths. Scientific reports, 5,

9426. doi:10.1038/srep09426



Marzbani, H., Marateb, H. R., & Mansourian, M. (2016). Neurofeedback: A Comprehensive Review on System Design, Methodology and Clinical

Applications. Basic and clinical neuroscience, 7(2), 143–158. doi: 10.15412/J.BCN.03070208

Rabiya, S., & Raghavan, V. (2018). Prison mental health in India. Indian Journal of Social Psychiatry, 34(3), 193.

Van Outsem, R. (2014). Novel Neuropsychological Opportunities in the Treatment of Antisocial Behaviour. In J. H. Galloway (Ed.), Antisocial Behaviour (pp. 15-

42). Nova Science Publishers.

Yadav, R. (2014). Recidivism in India and Statutory Provisions Relating to the Habitual Offenders. Correction & Sentencing Law & Policy doi:

http://dx.doi.org/10.2139/ssrn.2392802

http://dx.doi.org/10.2139/ssrn.2392802

See Ching Mey /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, No 1, pages 006-011 6


Neurofeedback and Counseling as Integrative Treatment

See Ching Meya*

aLoh Guan Lye Specialists Centre, Malaysia

Abstract

This paper describes neurofeedback as a tool that is suitable for use along with counselling approaches. The benefits of integrating

neurofeedback with counselling are discussed. Neurofeedback training offers opportunities for rehabilitation through directly retraining brain

activity. For clients with severe symptoms and resistance to many other treatments, neurofeedback has provided a new beginning and has offered

hope. In this paper, four cases of patients with presenting problems including insomnia, anxiety, obsessive compulsive disorder (OCD) and depression using this integrated treatment approach are presented. All participants in this case series experienced significant benefits.



ARTICLE INFO

RECEIVED November 2019

REVIEWED February 2020

ACCEPTED March 2020






1. Introduction

Counselling is often an important part of treatment in helping patients with mental disorders. It has also been shown that

counselling can be even more effective when combined with alternative forms of treatment such as neurofeedback (Goodwin and

Montgomery 2006). Neurofeedback relies on principles of neuroplasticity and enables the patient to alter his/her brain wave activity

through training for the purpose of improving health, wellbeing and performance. Neurofeedback may be thought of as a three-step

process. First becoming aware of a brainwaves response, then learning to control the response, and finally transferring control of

the response to everyday life.

It is our view that Neurofeedback offers the mental health professionals a powerful complementary tool that may be

integrated with counseling to treat a variety of mental disorders. Furthermore, neurofeedback has the potential to provide researchers

a unique opportunity to investigate clinical interventions with biological evidence of their efficacy.

Research is highly supportive of counselling approaches which include therapeutic characteristics of listening, empathic

understanding, building relationship of trust, gaining insight, and building on strengths and wellness. The findings also support the

notion that awareness of environment, culture and self shapes the individual. The bridge between biological and psychological

processes is erasing the old distinction between mind and body, between mind and brain — the mind is the brain.

It is our thesis that the integration of counselling and neurofeedback treatments will enhance the treatment process given

to patients resulting in speedy recovery and improved mental health and wellbeing.

2. Role of Neurofeedback In Helping People With Mental Disorders

Research shows that neurofeedback training can be helpful in cases such as anxiety (Kerson, Sherman & Kozlowski, 2009;

Scheinost et al., 2013; depression (Hammond, 2013; Wang at al., 2016; Cheon, Koo & Choi, 2016); eating disorder (Bartholdy et

al., 2013; Schmidt & Martin, 2016); insomnia (Hammer et al., 2011; Buckelew, Degood & Taylor, 2013; Arns et al., 2014);

obsessive compulsive disorder (Sürmeli & Ertem, 2011); post traumatic stress (Gapen et al., 2016) and schizophrenia (Bolea, 2010;

Sürmeli et al., 2012).

A common complaint among the adult and ageing population is insomnia, generally defined as the subjective sense that

sleep is difficult to inititate or maintain, or that sleep itself is non-refreshing. Neurofeedback has been shown to positively impact

sleep. With neurofeedback, most people can train their brain to improve sleep and this improvement can be surprisingly quick

for patients who have attempted many other forms of intervention and struggled with sleep for years (Cortoos, 2010; Hammer,

2011).

OCD is characterised by recurrent and persistent thoughts, impulses, images (obsessions) and repetitive behaviours

(compulsions) or mental acts that the person is driven to in order to attempt to control the obsessions. Neurofeedback can help the

brain regulate itself better, and reduce the symptoms of brain dysregulation that occur with OCD (AboutNeurofeedback 2012).

Research evidences suggests a neurophysiological basis for depression, particularly in people with a family history of

depression (Hammond, 2013). Research indicates that the left frontal area of the brain is associated with positive emotions and

approach motivation, which is a desire to be involved with other people. The right frontal area of the brain is more associated with

depression and fear, accompanied by motivation to withdraw from and avoid other people. Neurofeedback treatments for depression

appears very promising not only in bringing relief from depression, but in modifying the underlying biological predisposition for

becoming depressed (Hammond 2013).

3. Role of Counselling In Helping People With Mental Disorders

This paper focuses on four mental disorders - namely, Anxiety, Insomnia, Obsessive Compulsion Disorder (OCD) and

Depression. In addition to neurotherapy a number of different counselling approaches were utilised in the treatment of subjects

included in this series.

There are many models, approaches and techniques in counselling. Some of the most commonly used counselling

approaches for patients with mental disorders are Cognitive Behavioural Therapy (CBT/CEBT), Behavioural Therapy,

Psychoanalysis and Reality Therapy.

Patients who are treated using CBT can expect their therapist to be problem-focused, and goal directed when addressing

and challenging problems.

Behavioural therapy focuses on human behaviour and looks to eradicate unwanted or maladaptive behaviour. Typically,

this type of therapy is used for those with mental disorders that involve unwanted behaviour. Examples of this can include Anxiety,

Insomnia, Obsessive Compulsion Disorder (OCD) and Depression. Practitioners of behavioural therapy believe that behaviour is

learned and can therefore be un-learned via therapy. As well as the behaviour itself, the therapists will look at the thoughts and

feelings that lead to the behaviour or occur as a result of the behaviour to understand it at a deeper level (Hayes 2004). Behavioural

therapy is an action-based therapy that looks to foster positive behaviour change (Cherney 2013).

Psychoanalysis is another approach based upon the theory that our present is shaped by our past (Channel 2008). Past

experiences that are hurtful can remain in the unconscious and subconscious mind, influencing present mood as well as behaviour.

This can lead to problems with self-esteem, personality, relationships and work. Psychoanalysis helps a patient take control of these

influences by tracing them back to their origins and understanding how they have developed over time. This awareness offers the

patient the opportunity to deal constructively with the way these influences have affected their present life.



Reality Theory maintains that all human behaviour is internally motivated. Behavioural choices represent the best attempt

at any one moment to meet one or more of the basic needs which are built into the generic structure (Burke 2012). This theory holds

that all human beings have the same five basic needs and they are the drivers of all human behaviour. The theory suggests that all

human beings have the same behaviour system and that behaviour is driven by a person’s attempt to meet their basic needs. (Glasser

1998). With this approach, patients are guided to make choices and control their own behaviour, not someone else's - human

behaviour is seen as determined by what goes on inside of the person due to external forces. Reality therapy focuses on the future

by helping individuals to take ownership of and responsibility for their actions, which allow them to direct their own lives (Burke

2012).

4. Case Presentation

Four patients with presenting problems including Insomnia, Anxiety, OCD and Depression are presented in this paper.

These patients were diagnosed by the psychologist and given integrative treatment using neurofeedback and counselling. Their age

ranges from 40 to 70 years old. Below is a summary of treatment and outcomes.

In all cases neurofeedback protocols have been determined by clinical experience of the author and also a number of other

neurotherapists based on the principle of “what works”. While rewarding activity in the Delta range might be seen as controversial

by some, we have had successful outcomes using the protocols outlined in these case studies.

These protocols used are described below along with the goal that the training hopes to achieve.

Neurofeedback Protocols

Case 1: Patient diagnosed with anxiety

Patient A Detail: Gender Female

Age 42

Diagnosis Generalised Anxiety Disorder (GAD)

Patient A was diagnosed with GAD. She had symptoms of anxiety such as constant worrying, sense of fear, restlessness,

and negative thinking about events/things around her resulting in her responding emotionally over small matters. She appeared calm

on the outside, but she said her brain never stopped thinking. She also had problem falling asleep, muscle tension, and shortness of

breath. She had extensive fast brainwave activity in the right frontal lobe of the brain, which would seem to corresponded to her

complaint of being unable to shut her mind off from her worries and thus, I also hypothesise that it is linked to her sleeping problems.

The patient did not want to be put on medication so the integrative treatment was recommended. Neurofeedback aimed to

decrease the level of arousal, and thus, training aimed to activate middle or low frequency brain wave activity instead of high

frequency activity (Jacobs 2013). After about 6 sessions of neurofeedback, the patient said that she was able to fall asleep easily

and slept for 6-8 hours without waking up, managed her worries and fear, and had less muscle tension.

The protocols used for the patient and their associated goals were T6 Beta:Improve social behaviour; F7 – F8 Delta:Reduce

running thoughts; P4 Delta:Improve sleep; and T3 – T4 Delta: Improve calmness.

The patient also underwent counselling sessions using CBT and Behaviour Therapy forthnightly for the months of June

and July, and once a month in the months of August and September.

Neurofeedback Sessions: Month Session Treatment Protocol

June Session 1 – Session 5

T6 Beta, F7 – F8 Delta,

P4 Delta, T3 – T4 Delta

July Session 6 – Session 10

August Session 11 – Session 12

September Session 13 – Session 16

Total sessions 16

*** P4 & T3 – T4 Delta treatment is alternated based on the symptoms she was displaying at the time.

Findings: Presenting Problem Outcome

• Constant worrying

• Negative thinking

• Emotional when she was dealing with people at home and at her

workplace • Running thoughts resulting in her inability to fall asleep

• Was able to challenge her irrational thinking

• Was able to manage her worries

• Able to manage her emotion while dealing with people

• Able to control outburst of emotion

• Can fall asleep and stay asleep for 6-8 hours

Case 2: Patient with insomnia

Patient B Detail: Gender Female

Age 48

Diagnosis Insomnia

Patient B had difficulty falling asleep or staying asleep. She complained that she felt tired after waking up. She experienced

fatigue and could not focus or concentrate on her work. She lacked motivation and experienced mood swings.



The patient had difficulty initiating sleep and staying asleep. There is some evidence that theta training is suitable for

patients who have difficulty initiating sleep while delta training benefits patients who have problems maintaining sleep (Diaz, Sloot

et al. 2012).

The protocols and associated goals used for the patient was T3 – T4 Delta: Improve calmness; P4 Delta : Improve sleep;

F2 Beta : Improve motivation; and F3 Beta: Improve Mood.

The patient underwent counselling weekly for the month of July and forthnightly for the month of August. The psychologist

used Behaviour Therapy and CBT in the counselling sessions.


July

August

Session 1 – Session 7

Session 8-16 T3 – T4 Delta, P4 Delta, F2 Beta, F3 Beta

Total sessions 16

*** F2 Beta is alternated with F3 Beta starting from Session 6-16.


• Tired after waking up

• Fatigue all day

• Could not focus nor concentrate

• Mood disturbances

• Using the techniques trained in the counselling session, she is

able to fall into sleep and sleep through the night without waking up

• Felt rested in the morning

• Able to focus and go about her daily activities

• Cheerful

Case 3: Patient with Obsessive Compulsive Disorder

Patient C Detail: Gender Male

Age 72

Diagnosis Obsessive Compulsive Disorder (OCD)

Patient C held a high position in a corporate company and had always realised that he had some sypmtoms of OCD, but

was never diagnosed nor had he sought help. When he retired, he realised that his problems affected his homelife and this led to

further stress.

The patient expressed obsessive thoughts and displayed compulsive behaviour.

Frontal lobe overarousal has been assosciated with repetitive thoughts, worries and urges (Jacobs 2013). Delta training at frontal

lobe sites aims to lower beta and hi-beta activity in order to reduce anxiety and obsessive thoughts.

The treatment protocols and associated goals were F3 Beta:Reduce depression; T4 Delta: Emotional calming; and T3 – Fz

Delta: Reduce obsessiveness.

The patient was rational and had insight. He was determined to change. He attended weekly counselling for two months

and completed all the assignments given by the psychologist. Techniques from CBT and Reality Therapy were used.


November Session 1 – Session 5 F3 Beta, T4 Delta, T3 – Fz Delta

Dec ember Session 6 – Session 11

Total sessions 11


• Fear of contamination or dirt

• Having things orderly and the way he wants it to be

• Unwanted repetitive thoughts

• Aggressive or horrific thoughts about harming himself or

others

• Reduced fear of contamination or dirt

• Able to walk away if things are not orderly and as he wants it to

be

• Noted unwanted repetitive thoughts and able to challenge the

irrational thoughts

• Able to pause and review aggressive or horrific thoughts about

harming himself or others

Case 4: Patient diagnosed with depression

Patient D Detail: Gender Female

Age 41

Diagnosis Depression



Patient D was diagnosed with early on-set of depression. She was feeling low in spirit, and sometimes felt that life was no

longer worth living. She complained of persistent feelings of sadness and loss of interest. This was accompanied with feelings of

hopelessness, worthlessness, helplessness and restlessness, along with suicidal ideation and a suicide attempt.

In neurofeedback therapy with patients who have depression, brain training typically focuses more on the patient's frontal

and temporal lobes. Hammond (2005) concluded that neurofeedback not only moderates depression but also mitigates anxiety and

rumination.

During neurofeedback, the treatment protocol and associated goals were F7 – F8 Delta : Reduce running thoughts; C4

Delta : Calm unconcious mind; F2 Beta : Improve motivation; F3 Beta: Reduce depression; and T3 – T4 Delta : Improve calmness.

Patient D came for counselling every fortnight. Using psychoanalysis techniques, she was able to identify the events and

people who made her sad. She explored her past experiences with the goal that understanding these experiences would help her to

manage her anger and the thoughts of ending her life.


August Session 1 – Session 2 F3 Beta, T4 Delta, T3 – F2 Delta

September Session 3 – Session 9

October Session 10 – Session 16

Total sessions 16


• Less motivated to do tasks

• Running thoughts

• Annoyed by past and present family issues

• Sad feelings with suicidal thoughts

• Able to occupy herself with beneficial activities

• Able to control her running thoughts

• Resolved inner conflict and became calmer when she delt

with family members. Does not get annoyed easily

• No more suicidal thoughts

5. Implications and Recommendations

This case series is limited to just four cases, each with very different presentations. The interventions are personalised and

thus vary between cases. This represents an exploratory study of outcomes where neurofeedback is integrated with various

counselling approaches as an intervention for different mental health disorders.

Future research might focus in more detail on one particular mental health disorder and investigate different protocols for

treatment. Follow up studies should ideally include objective measures to assess outcomes and use standardised treatment protocols

and procedures.

6. Conclusions

This exploratory case series provides some support for an integrative treatment model incorporating neurofeedback along

with psychotherapy/counseling approaches. There are many questions and issues outstanding, but this case series indicates some

promising directions for clinical interventions and for future research.

References

AboutNeurofeedback (2012). "Anxiety." Retrieved 10 December, 2014, from http://www.aboutneurofeedback.com/conditions/anxiety/.

AboutNeurofeedback (2012). "Obsessive Compulsive Disorder." Retrieved 10 December,2014, from http://www.aboutneurofeedback.com/conditions/obsessive-

compulsive-disorder/.

Arns M, Feddema I, Kenemans J, L. (2014). Differential effects of theta/beta and SMR neurofeedback in ADHD on sleep onset latency. Front Hum Neurosci. 8:1019.

Bartholdy, S., Musiat, P., Campbell, I, C., Schmidt, U. (2013). The potential of neurofeedback in the treatment of eating disorders: a review of the literature. European

Eating Disorders Review: The Journal of the Eating Disorders Association 21, 456–463.

Bolea A. S. (2010). Neurofeedback treatment of chronic inpatient schizophrenia. J. Neurother. 14, 47–54 10.1080/10874200903543971

Braintrain. (2014). "Neurofeedback for Other Conditions." Retrieved 10 December, 2014, from http://www.braintrainuk.com/other-conditions-that-neurofeedback-

supports/.

Buckelew, S., Degood, D., Taylor. J., (2013). Neuroflexibilty and sleep onset insomnia among college students: Implication for neurotherapy. Journal of

Neurotherapy.

Burke, D. (2012). "Choice Theory and Reality Therapy." Retrieved 9 December, 2014, from http://www.davidrburke.com/choice-theory-reality-therapy.html.

Channel, B. H. (2008). "Psychoanalysis." Retrieved 14 December, 2014, from http://medical-dictionary.thefreedictionary.com/PSYCHOANALYTIC+TREATMEN

Cheon, E., Koo, B., & Choi, J. (2016). The efficacy of neurofeedback in patients with major depressive disorder: An open labeled prospective study. Applied

Psychophysiology And Biofeedback, 41(1), 103. doi:10.1007/s10484-015-9315-8



Cherney, K. (2013). "Behavioral Therapy." Retrieved 10 December, 2014, from http://www.healthline.com/health/behavioral-therapy#Overview1. Clinic, M. (2009).

"Biofeedback." Retrieved 9 December, 2014, from http://www.mayoclinic.org/.

Cortoos, A, et al. (2009), " An Exploratory Study on the Effects of Tele-neurofeedback and Tele-biofeedback on Objective and Subjective Sleep in Patients with

Primary Insomnia", Applied Psychophysiology and Biofeedback, Vol. 35, pp 125-134, 2010.

Diaz, B. A., et al. (2012). "EEG-Biofeedback as a Tool to Modulate Arousal: Trends and Perspectives for Treatment of ADHD and Insomnia."

Fagan, J. and I. L. Shepherd (1970). "Gestalt therapy now: Theory, techniques, applications."

Gapen M., van der Kolk B. A., Hamlin E., Hirshberg L., Suvak M., Spinazzola J. (2016). A Pilot study of neurofeedback for chronic PTSD. Appl. Psychophysiol.

Biofeedback 41, 251–261. 10.1007/s10484-015-9326-5

Glasser, W. (1998). "Choice theory." New York: HarperPerennial.

Goodwin, E. A. and D. D. Montgomery (2006). "A cognitive-behavioral, biofeedback-assisted relaxation treatment for panic disorder with agoraphobia." Clinical

case studies 5(2): 112-125.

Hammer, B. U., Colbert, A. P., Brown, K. A., & Ilioi, E. C. (2011). Neurofeedback for insomnia: A pilot study of Z-score SMR and individualized protocols. Applied

psychophysiology and biofeedback, 36(4), 251-264. https://doi.org/10.1007/s10484-011-9165-y

Hammond, D. C. (2013). "Neurofeedback for Depression." Retrieved 10 December, 2014, from http://www.isnr.net/neurofeedback-info/depression.cfm.

Hammond, D. C. (2005). "Neurofeedback treatment of depression and anxiety." Journal of Adult Development 12(2-3): 131-137.

Hayes, S. C. (2004). "Acceptance and commitment therapy, relational frame theory, and the third wave of behavioral and cognitive therapies." Behavior therapy

35(4): 639-665.

Jacobs, M. S. (2013). "Psychological factors influencing chronic pain and the impact of litigation." Current Physical Medicine and Rehabilitation Reports 1(2): 135-

141.

Kerson, C., Sherman, R. A., & Kozlowski, G. P. (2009). Alpha suppression and symmetry training for generalized anxiety symptoms. Journal of Neurotherapy, 13,

146-155. doi:10.1080/10874200903107405

Leanza, N. (2012). "NeuroCounseling: Simple Therapeutic Interventions for Rewiring the Maladaptive Brain." Psych Central. Retrieved 10 December, 2014, from

http://psychcentral.com/lib/neurocounseling-simple-therapeutic-interventions-for-rewiring-the-maladaptive-brain/00011117.

Moore, N. (2000). "A review of EEG biofeedback treatment of anxiety disorders." Clinical EEG (electroencephalography) 31(1): 1-6.

Renaud, J. J. J. M. G. (2014). "Predicting Who Benefits Most From Cognitive-Behavioral Therapy for Anxiety and Depression." Journal of Clinical Psychology

70(10): 924-932.

Rupke, S. J., et al. (2006). "Cognitive therapy for depression." Am Fam Physician 73(1): 83-86.

Scheinost, D. et al. (2013). Orbitofrontal cortex neurofeedback produces lasting changes in contamination anxiety and resting-state connectivity. Transl. Psychiatry

3, e250. doi: 10.1038/tp.2013.24

Schmidt J., & Martin A. (2016). Neurofeedback against binge eating: A randomized controlled trial in a female subclinical threshold sample. Eur. Eat. Disord.

Rev.24:406–416. doi: 10.1002/erv.2453.

Surmeli, T., & Ertem, A. (2011). Obsessive compulsive disorder and the efficacy of qEEG-guided neurofeedback treatment: A case series. Clinical EEG and

Neuroscience. 2011;42(3):195–201.

Surmeli T., Ertem A., Eralp E., Kos I. H. (2012). Schizophrenia and the efficacy of qEEG-guided neurofeedback treatment: a clinical case series. Clinical EEG

Neurosci. 43, 133–144 10.1177/1550059411429531

Wang L., Feng Z., Yang G., Yang Y., Wang K., Dai O., et al. (2016). Depressive symptoms among children and adolescents in western China: an epidemiological

survey of prevalence and correlates. Psychiatry Res. 246 267–274. 10.1016/j.psychres.2016.09.050

https://link.springer.com/journal/10484

Herbert F. Jelinek /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, No 1, pages 012-021 12


A brief literature survey into the effects of coal seam gas exploration

by-products on Human EEG

Herbert F. Jelineka,b,c* and Lynda Ollingb

a School of Community Health, Charles Sturt University, Albury, Australia bMindSapce Australia, Sydney, Australia

cDepartment of Biomedical Engineering, Khalifa University, Abu Dhabi, UAE

Abstract

Environmental toxins are rapidly increasing as more new chemicals are synthesised and applied in agriculture, nutrition and industry

including conventional and unconventional mining exploration. In the coal seam gas (CSG) exploration industry, fracking or fracturing and venting

are common methods of extraction where several hundred known and unknown chemicals are used or released from the ground during the mining

and venting process. Chemicals associated with fracking can lead to headaches and migraine, stress and anxiety, toxic encephalopathy as well as

dizziness/balance problems and seizures. Quick and accurate as well as economical diagnostic tools are required for confirming neurological

disease or toxic encephalopathy. Quantitative electroencephalography (qEEG) is one such tool for obtaining baseline data that may be associated

with toxic encephalopathy. Characteristic EEG findings from the literature are focal or generalised slowing in the delta and theta frequencies or

fast beta rhythms indicative of diffuse or focal cerebral structural damage and dysfunction. This paper discusses some of the common chemicals

associated with CSG mining and associated EEG findings.



Keywords: Encephalopathy; Coal Seam Gas, toxins; Electroencephalogram

ARTICLE INFO

RECEIVED September 18, 2019

REVIEWED October 24, 2019

ACCEPTED February 23, 2020






1. Introduction

The Environmental Protection Agency of the United States of America reported the use of 84,000 chemicals by 2012,

which does not include chemicals produced in low volumes. The number of new chemicals added each year is approximately 1,000

to 2,000. Ensuring public health has become increasingly difficult despite improvements in large scale monitoring, better awareness

of health risks and increased human relevance associated with the increase of chemicals into the environment (Pool & Rusch, 2014).

Neurological problems such as headaches, dizziness, disorientation, depression and encephalopathy have been associated with

products listed by coal seam gas mining (CSG) companies. Characteristic, although not specific EEG changes have also been

reported following exposure to products associated with coal CSG mining such as chemicals and heavy metal particulate matter,

which may be signs of toxic encephalopathy.

2. Encephalopathy

Encephalopathy is defined by nonfocal or symmetrical brain dysfunction caused by histopathological changes, in the

cortical and subcortical areas of the brain but is in many cases asymptomatic depending on level and length of exposure (Kim &

Kim, 2012). Encephalopathy is associated with altered mental states, memory loss, cognitive decline, and personality changes, lack

of concentration and attention and lethargy. There are many causes of encephalopathy including bacteria, viruses, solvents, drugs,

radiation, paints, industrial chemicals, and some metals as well as metabolic or mitochondrial dysfunction. The most common

encephalopathy discussed in the literature is metabolic encephalopathy, which is characterised by systemic organ dysfunction

including the liver and kidney leading to increased toxic metabolites and cerebral histopathology (Perugula & Lippmann, 2016).

The National Institute of Neurological Disorders and Stroke listed 836 clinical trials related to encephalopathy in 2019

with one trial addressing carbon monoxide poisoning, 138 for hepatic encephalopathy, seven mitochondrial encephalopathy studies,

seven sepsis-related encephalopathy studies, 26 substance-related disorders of which two addressed hepatic encephalopathy and 17

listed as toxaemia, which included five associated with septic encephalopathy (Clinical Trials). No clinical trials on toxic

encephalopathy or environmental encephalopathy were listed. However continuous exposure to various environmental, industrial,

nutritional and/or agricultural toxins is increasing and can lead to peripheral or central nervous system dysfunction and hallmarks

of toxic encephalopathy. Examples of toxic encephalopathy with possible encephalogram anomalies include spastic paraparesis

caused by β-N-methylamino-Lalanine (BMAA) in lathyrus sativus peas, exachloraphene encephalopathy, mercury encephalopathy,

acute ethylene glycol poisoning and organic solvent encephalopathy (Dobbs, 2011; Valk & van der Knaap, 1992). A closer look at

signs and symptoms of possible neurotoxicity that may show EEG anomalies include severe fatigue, weakness, headaches,

spontaneous nose bleeds, numbness and paraesthesia, twitching or unusual movements, clumsiness or unsteadiness (McCarron,

2013).

2.1 Toxic encephalopathy diagnostics

Toxic encephalopathy may be an outcome of exposure to environmental toxins. Both physical and neurological

examination is required to establish the presence of toxic encephalopathy. Diagnosis of toxic encephalopathy can be divided into

acute diffuse toxic encephalopathy or chronic toxic encephalopathy. To date only a limited number of physical signs have been

identified that are associated with specific environmental toxins and include blue gums following lead intoxication (Rao,

Vengamma, Naveen, & Naveen, 2014). Acute diffuse toxic encephalopathy is relatively easy to identify as there is a distinct

correlation of symptoms with time since exposure and place of exposure to a known toxin. Acute diffuse toxic encephalopathy is

characterised by mild euphoria, stupor or seizures, which can lead to coma and death. Chronic toxic encephalopathy involves

continuing diffuse injury to the brain resulting from cumulative or repeated exposures over months or years to solvents or heavy

metals and involves varying degrees of impairment (Firestone & Longstrength, 2004). Chronic toxic encephalopathy is more

difficult to diagnose and consists of a number of subcategories depending on severity and length of exposure. Chronic toxic

encephalopathy manifests as mood disorders, deficits in attention, memory and learning as well as decreased psychomotor function.

Cerebellar syndromes include gait ataxia, dysarthria, intension tremor, gaze-evoked nystagmus, most often due to methyl mercury,

methyl bromide, dimethyl and trimethyl compounds. Manganese and carbon disulphide can lead to Parkinsonism and vascular

encephalopathy respectively (Kim & Kim, 2012). Encephalogram (EEG) changes associated with encephalopathies are similar,

whether due to septic, metabolic, toxic, or structural causes. Non-invasive and simple EEG analysis may provide some indication

of structural changes due to encephalopathy including toxic encephalopathy.

2.2 EEG findings

Assessment of EEG rhythm characteristics provides information on focal or lateralised abnormalities of the brain, which

could suggest a structural basis for an encephalopathy if the intermittent or persistent focal slowing is observed consistently and

hence corresponds to a focal or generalized cerebral dysfunction or both (Andraus & Alves-Leon, 2011). EEG records associated

with drug or toxin encephalopathy can show focal or generalised slowing in the delta and theta frequencies indicative of diffuse or

focal cerebral dysfunction, or fast beta rhythms (Smith, 2005). However, in more severe encephalopathy cases burst-suppression,

background suppression and cerebral electrical inactivity can also be present. In addition less common patterns include alpha coma,

spindle coma and triphasic waves (Altwegg-Boussac et al., 2017).



3. Chemical Use in Coal Seam Gas Exploration

The number of chemicals used in CSG mining, fracking and venting is estimated to be approximately 1,000. An

investigation by the United States Committee of Energy and Commerce in 2011 published results from a survey of 14 oil and gas

service companies about types and volumes of hydraulic fracturing products used between 2005 and 2009 (Kim & Kim, 2012).

Results indicated the use of more than 2,500 products containing 750 chemicals and other components. Although some were

harmless to health, others that are hazardous to health included benzene and lead and for many exposure to these is occurring daily

in non-CSG mining areas such as petrol stations. Twenty nine of the 750 chemicals and components are known to be possible human

carcinogens and were contained included in the list of the 2,500 products. BTEX compounds – benzene, toluene, ethylbenzene and

xylene were part of 60 hydraulic fracturing products (Waxman, Markey & DeGette, 2011). Most chemicals used in CSG mining are

either not disclosed or not been assessed for toxicity and short or long term health effects. In the USA between 2005 and 2009 the

14 companies investigated reported using 94 million gallons of 279 products that contained at least one product that was undisclosed

(Waxman, et al., 2011). Disclosure of chemicals used in mining and fracking is not required in Australia or in the United States of

America. A Bill passed by the American Government in 2005 which exempts fluids used in the natural gas extraction process of

hydraulic fracturing from protection under the Clean Air Act, Clean Water Act and Safe Drinking Water Act and from regulation

by the Environmental Protection Agency is referred to as the Haliburton Loophole (Palmer, Short & Auch, 2018). However,

approximately 40-50% of chemicals used during CSG mining could affect the brain or nervous system, of which some at

concentrations below current recommended as hazardous (Colborn , Kwiatkowski , Schultz & Bachran, 2001). An example of what

has been proclaimed as clean energy and hence not hazardous by the CSG Industry is the emissions of the Curtis liquefied natural

gas plant which released 4,800 tonnes of carbon monoxide, 4,300 tonnes of nitrous oxides, 620 tonnes of volatile organic compounds

and 190 tonnes of carcinogenic formaldehyde (Lloyd-Smith & Senjen, 2011). A report compiled by Wayne Somerville (Somerville,

2014) lists some of the chemicals in use for CSG mining in Australia (Table 1).

Table 1. Known chemical used in CSG mining operations in Australia

Drilling Fluid Hydraulic Fracturing Fracking

Viscosifiers (bentonite,

polyacrylamide)

Gelling agents (guar gum, diesel,

alkanes/alkenes)

Corrosion inhibitors (e.g., formamide,

methanol, naphthalene, naptha, nonyl phenols,

acetaldhyde)

Weighting agents (barium sulphate) Gel stabilisers (sodium thiosulphate) Scale inhibitors (e.g., ethylene glycols)

Bactericides/biocides

(glutaraldehyde)

Gel breakers (Ammonium persulfate, sodium

persulfate)

Iron control agents (e.g., citric acid,

thioglycolic acid)

Salts (KCl, NaCl, CaCl2) Friction reducers (polyacrylamide, mixtures of methanol, ethylene glycol)

pH adjusting agents (sodium or potassium carbonate)

Breakers (diammonium

peroxydisulphate, hemicellulase

enzyme)

Surfactants (isopropanol, 2-Butoxyethanol

/2-BE)

Diluted acid to dissolve minerals (e.g.,

hydrochloric acid, muriatic acid)

Scale inhibitors (anionic

polyacrylamide, acrylamide

copolymer)

Biocides (glutaraldehyde, Tetrakis

hydoxymethyl phosphonium sulfate/THPS,

2-Bromo-2-nitro-1,3-propanediol (Bronopol), 2,2-Dibromo-3-

nitrilopropionamide

Emulsifiers and deemulsifiers Clay stabilisers (tetramethyl ammonium chloride)

Polymer stabilisers (Sodium sulfite) Buffer fluids and cross-linking agents

Defoamers (glycol blends, light

aromatic and aliphatic oil, naptha)

Corrosion inhibitors (zinc carbonate, sodium polyacrylate)

Lubricants (chlorinated paraffins)

The National Toxins Network in Australia reported that of the 23 compounds it has identified as toxic, only two have been

assessed by the National Industrial Chemical Notification and Assessment Scheme (NICNAS) (Lloyd-Smith & Senjen, 2011)

despite glutaraldehyde, brominated biocides, propargyl alcohol, 2-butoxyethanol and heavy naphtha having been found to be

dangerous at concentrations near or below chemical detection limits (Somerville, 2014). Additional substances known and reported

to be associated with CSG mining are shown in Table 2.

Table 2. Additional CSG Chemicals

2-methyl phenol Chloride Silver

Acrylamide polymer Copper Styrene

Arsenic Cyanide Uranium

Barium HMX (Octogen) Diesel

Benzene, Toluene, Ethylbenzene, Xylene (BTEX)

Methane Radon

Cadmium Poly-aromatic hydrocarbons

Only a limited number of these compounds have been assessed for their neurological effect using any available modality

such as MRI, PET or EEG.



3.1 CSG chemicals and EEG findings

Organic solvents and other substances containing toluene associated with CSG exploration are known to cause multifocal

neurologic pathology and mental disorders, cerebral, cerebellar and brain stem atrophy, and diffuse focal white matter abnormalities

detectable by MRI and EEG changes in animals and humans (Takeuchi & Hisanaga, 1977; Valk & van der Knaap, 1992). A closer

look at some of these and other compounds associated with CSG exploration follows.

3.2 Carbon dioxide and radioactive emissions

Atmospheric enrichment of carbon dioxide and radon-222 gas has been reported in the Tara coal seam gas fields in

Australia (Tait, Santos, Maher, Cyronak & Davis, 2013). Increased carbon dioxide leading to blood alkalosis can interfere with

cerebral function by a relative increase in low frequency power in the delta (1-3 Hz), theta (4 to 7 Hz) and alpha (8-13 Hz) range of

the quantitative EEG (qEEG) spectra indicating reduced brain arousal (Kety & Schmidt, 1946; Posner & Plum, 1960; Xu et al.,

2011). Default mode network investigation has also indicated a significant reduction between normocapnia and hypercapnia in the

connection between the posterior cingulate cortex, bilateral inferior parietal regions, medial prefrontal cortex, and medial temporal

lobe. With similar results in the sensory motor networks of the brain using fcMRI (Xu et al., 2011). Radon-222 (radon) has been

studied extensively as an environmental toxin and has been shown to be associated with lung cancer. However, radon gas may also

affect different parts of the brain with different areas having different susceptibility to radon gas. The amygdala and hippocampus

being the most susceptible with the proviso that susceptibility differs between individuals. What increases susceptibility is not

known and hence setting threshold values may not prevent radon-caused toxic encephalopathy (Momčilović, Lykken & Cooley,

2006). However, radiation therapy, which has a different dose exposure to possible radon emissions from CSG mining has led to

leukoencephalopathy in some patients with EEGs characterised by diffuse background slowing consistent with moderate to severe

toxic encephalopathy and no focal or epileptiform waveforms (Cummings et al., 2016). No publications were found discussing EEG

findings associated with increased environmental cesium-137 or lead-210, two substances also found in areas of CSG exploration.

3.3 Lead and cadmium

Lead poisoning is one of the earliest reported occupational diseases known and already common in Ancient Greece and

the Roman Empire (Woolley, 1984). Lead and cadmium effects on sensory evoked potentials and qEEG have been reported from

the early 1980tees (Benignus, Otto, Muller, & Seiple, 1981; Otto, Benignus, Muller & Barton, 1981; Thatcher & Lester, 1985).

Thatcher and co-workers reported an increase in the amount of slow wave activity in the EEG and a decrease in the amplitude of

the EEG. To note is that lead and cadmium have different affinities for cortical and subcortical regions and susceptibility to altered

CNS function and convulsion varies with age and exposure levels (Thatcher, McAlaster, Lester, & Cantor, 1984). Examination of

qEEG z-scores reported by Cantor showed one third of cases with anomalous findings in the frontal regions with increased θ-band

relative power and abnormal hypercoherences (Cantor, 2000). Alpha and β activities of the EEG may also be more abundant in long

term exposure to lead (Kovala et al., 1997). Mani et al. (1998) however did not observe any EEG anomalies in their cohort.

Quantitative EEG findings included δ- and θ-band amplitude that decreased with age. Scalp sensor location at P3 and P4 indicated

a greater amplitude at P3, and bilateral communality increased with age in the δ-band. These EEG changes were observed with low

level lead exposure without any clinical or behavioural symptoms present (Benignus, et al., 1981). A more recent case study using

MRI reported bilateral symmetric anomalies of the thalamus, lentiform nucleus, external capsules, and sub-cortical white matter

suggesting toxic demyelination (Rao, et al., 2014). Cerebellar white matter and demyelinating peripheral neuropathy have also been

reported with lead poisoning and seizures may also be present due to inhibition of gamma-aminobutyric acid (GABA) transmission

(Landrigan & Todd, 1994; Mani, Chaudhary, Kanjalkar & Shah, 1998; Silbergeld, Miller, Kennedy & Eng, 1979)

3.4 Nitrous Oxide

Nitrous oxide is a NMDA receptor antagonist and used clinically for sedation and possibly for depression. Nitrogen oxides

are generally non-toxic at low concentrations, and are a by-product when CSG is treated for sale. Nitrous oxide emissions have risen

from 710 tonnes in 2013-14, to 1,300 tonnes in 2014-15 (McCarron, 2013). Neurological assessment has shown a dose-dependent

reduction in P-300 amplitude and prolongation of P-300 latency of evoked potentials when psychomotor impairment was present

following an acute dose of nitrous oxide. However, these changes in brain function were not apparent with EEG suggesting that

measures of evoked potential are possibly more sensitive compared to EEG (Estrin, Moore, Letz & Wasch, 1988). In another study

five increasing concentrations of nitric oxide were investigated and administered for 15 min. EEG was recorded bilaterally at the

frontal poles. At the highest concentration the EEG showed increased θ-band, β at the 40-50 Hz, and 70-110 Hz band powers

(Rampil, Kim, Lenhardt, Negishi & Sessler, 1998). Emerging slow wave rebound activity may also be present once nitrous oxide

is withdrawn.

3.4 BTEX in coal seam gas mining

BTEX which stands for benzene, toluene, ethylbenzene and xylenes is a Volatile Organic Compounds (VOR) found in tar,

crude petroleum, diesel and petrol fuels and a variety of petroleum-related products and used by the CSG industry as a solvent.

Naturally occurring BTEX compounds are typically found in seawater around areas of natural gas and petroleum deposits, coal



deposits and in gas emissions from volcanoes and bushfires. BTEX has been banned by both the NSW and Queensland Governments

for use in CSG mining however due to naturally occurring BTEX found in soil and released in the flowback water, ground water

contamination can still occur and become a health hazard (Waxman et al., 2011; Webb et al., 2018).

Solvent toxicity generally manifests as diffuse or localised θ-band activity. Reported prevalence of EEG anomalies

associated with industrial solvents including benzene ranged between 35% to 80% (Seppäläinen, 1988). No specific data was located

for effects of benzene on EEG. EEG findings associated with ethylbenzene are most often generalised with paroxysmal changes

(Indulski, Sińczuk-Walczak, Szymczak & Wesołowski, 1996).

Toluene toxicity has been reported in several studies with concomitant EEG findings. Toluene leukoencephalopathy is the

main neuropathology associated with white matter damage. Other neuro- and psychopathology includes ataxia, tremors, emotional

lability and mental changes (Filley, Halliday & Kleinschmidt-Demasters, 2004; Satran & Dodson, 1963). An animal model of

toluene exposure indicated that EEG characteristics vary with toluene concentration and with sleep-wake cycle as well as with

respect to CNS location. In a study comparing the highest exposure group to the lowest indicated that for the highest toluene

concentration δ1 diminished and β1 and β2 cortical components increased significantly. In contrast in the lowest toluene exposure

group, δ1, θ, and α bands were significantly reduced but β1 and β2 bands showed no significant change (Takeuchi & Hisanaga, 1977).

Xylene is generally considered to be more acutely toxic than benzene, but similar to toluene (Bergman, 1979). Acute effects

of solvents including xylene on the central nervous system can be observed already at very low concentrations (Gamberale, Annwall

& Hultengren, 1978). Xylene can lead to giddiness, anorexia, and vomiting and exacerbate seizures in susceptible individuals.

Occupational exposure to xylene has been associated with anaemia, thrombocytopenia, leukopenia, chest pain with ECG

abnormalities, dyspnoea and cyanosis, in addition to CNS symptoms (Langman, 1994). Exposure may increase the dominant α-

band frequency and percentage. However, the effects of short-term xylene exposure on EEG were minor, and no deleterious effects

were noted (Seppäläinen et al., 1991).

3.4 Carbon monoxide

Carbon monoxide is emitted during flaring and from machinery. Acute carbon monoxide poisoning can lead to lateralized

sharp waves and a focal electrographic seizure discharge within hours of the exposure as well as delayed toxic encephalopathy and

cognitive sequelae. MRI imaging found bilateral abnormal white matter changes within the cerebral hemispheres and abnormal

signal intensity within the basal ganglia putamen and caudate nuclei (Neufeld, Swanson & Klass, 1981; Tapeantong & Poungvarin,

2009). No EEG reports associated with carbon monoxide were found.

3.5 Sulfur dioxide and hydrogen sulfide

Hydrogen sulfide occurs naturally in some gas formations and can be released when gas is vented or flared, or via fugitive

emissions. Hydrogen sulfide, is a potent neurotoxin. Sulfur dioxide and hydrogen sulphide are components of coal seam gas but are

also produced by diesel equipment (Esen, Özer, Soylu, Rend & Fisne, 2018). These gases lead to oxidative stress and neurotoxicity

(Zhang, Vincent, Halliwell, & Wong, 2004). Exposure to sulfur dioxide and hydrogen sulphide can lead to decreased mitochondrial

DNA and loss of ATP, which may have an effect seen by EEG as a consequence of energy failure and toxic encephalopathy (Tzoulis

et al., 2010). Hence increased δ-band activity may be assumed in response to loss of ATP to conserve energy (Dworak, McCarley,

Kim & Basheer, 2011).

3.6 VOCs, formaldehyde, glutaraldehyde and styrene

Volatile organic compounds (VOC) contain benzene and ethyl-benzene, which have adverse neurologic and respiratory

effects. Formaldehyde forms when methane is exposed to sunlight and then transforms to carbon dioxide by photo-oxidation in

ambient air (Kaden, Mandin, Nielsen, & Wolkoff, 2010). Formaldehyde, naphthalene and benzyl chloride were reported to be used

in 34% of all hydraulic fracking operations in the United States between 2010 and 2011 (https://www.ecowatch.com/2013/cancer-

causing-chemicals-fracking-operations/). Increased formaldehyde around CSG fracking sites is also due to truck traffic and as a

biocide to prevent biodegradation of organic additives in CSG mining (Lloyd-Smith & Senjen, 2016; Sorg, Tschirgi, Swindell, Chen

& Fang, 2001). Formaldehyde may lead to thirst, headaches, dizziness, apathy, and inability to concentrate. The effect of

formaldehyde on EEG is unclear with studies reporting a dose-dependency but in some instances no EEG changes have been

observed (Fel'dman & Bonashevskaya, 1971; NRC, 1980).

Sarin is a nerve gas and belongs to the group of VOCs, which has been extensively studied and EEG analysis reported.

EEG analysis to sarin exposure, indicated long-term effects on EEG. Statistically significant effects of sarin were marked

desynchronised EEG with increased β-band activity, increased δ and θ band slowing, and decreased α-band activity. At higher doses,

VOCs may lead to convulsions and muscular paralysis, characterised by a general slowing of the EEG and spike wave discharges

(Duffy, Burchfiel, Bartels, Gaon & Sim, 1979). VOCs share similar structures and pharmacological action and therefore this study

may be extrapolated to VOCs associated with CSG mining and venting.

Glutaraldehyde has similar properties to formaldehyde and is also used as a biocide by the CSG industry. It is present in

produced water as part of the fracking operation (Campa et al., 2018). A suspected case of glutaraldehyde poisoning was reported

in 2002 with headache, loss of attention, dizziness, anxiety, drowsiness and alteration of homeostatic reflexes (Proietti, Longo &

Duscio, 2002). No specific reports discussing the toxic effects of glutaraldehyde on EEG were located in the current study.

Glutaraldehyde however has led to a significant increase in deaths due to leukaemia, and brain, colon and prostate cancers in a group

https://www.ecowatch.com/2013/cancer-causing-chemicals-fracking-operations/

https://www.ecowatch.com/2013/cancer-causing-chemicals-fracking-operations/



of embalmers, anatomists and pathologists compared to an age controlled group (van Birgelen et al., 2000). Glutaraldehyde may

also lead to mitochondrial dysfunction and ATP depletion and therefore may cause similar changes to the EEG as discussed above

for sulfur dioxide (Lin, Yuan, Deng, Niu & Chen, 2019; Tiffert, Garcia-Sancho & Lew, 1984).

Abnormal EEG were found in 24% of participants in a study of styrene exposure reported by Seppäläinen (Seppäläinen,

1988). EEG anomalies included were excessive diffuse θ-band activity and localized slow waves in the posterior regions. Bilateral

spike and wave (S&W) discharges may be present as well as focal slow wave activity. Continuous exposure to VOCs may also lead

to environmental chemical odour intolerance or chemical sensitivity, which has a different presentation to environmental toxin

exposure (Miller, 2001). Common signs and symptoms may include decreased REM percent and longer REM onset latency (Bell

et al., 1996).

3.7 Diesel

Diesel exhaust is a mixture of combustion derived nanoparticles (CDNP) and hydrocarbons, carbon monoxide, nitrous

oxides, polycyclic aromatic hydrocarbons (PAHs) and redox metals (Crüts, et al., 2008).

Particulate matter in the nanoparticle range such as found in exhaust diesel can translocate into the brain via the olfactory

nerves leading to extensive oxidative stress (Oberdorster, et al., 2004). Carbon monoxide, nitrous oxides and hydrocarbons may

also lead to decreased θ and α, and increased β and γ-band EEG changes via vagal reflexes from the lungs (Koo, 2001; Lewine,

Paulson, Bangera, & Simon, 2019; McQueen et al., 2007). A recent study of exposing volunteers to dilute diesel exhaust as a model

for ambient particulate matter exposure indicated a significant increase primarily in the β2 power frequency (20-32 Hz) in the frontal

cortex at Fp1 and Fp2 and F3, F4 following 30 minutes of exposure. Median power frequency changes spread to central C3 and C4

as well as to parietal sites, P3 and P4 but no significant change in β2 power was observed. β1 activity (15–20 Hz) also increased

during diesel exposure at the frontal cortex, but was not significant, whilst no significant change was observed for δ, θ and α-bands

when compared to the sham group. The findings suggested increased left frontal cortex activity (Crüts et al., 2008).

Many more chemicals and particulate matter have been investigated and found to have toxic effects. Of those in public

lists, not all are known to lead EEG anomalies. For a list of these the reader is referred to the various Government publications and

also for a comprehensive list by Geralyn McCarron and the publications by Mariann Lloyd-Smith (Lloyd-Smith & Senjen, 2011,

2016; McCarron, 2013). Further research is therefore required to establish biomarkers for substances associated with Coal Seam

Gas mining and venting and their association with neuropathology and EEG findings. Table 3 provides an overview of location and

effects of some of the main chemicals associated with fracking but is not exhaustive as more research is required for many of the

chemicals listed and those mentioned in the tables above.

Table 3. Local, EEG characteristics and neurology associated with known fracking chemicals.

Chemical Location EEG Characteristic Neuropathology

Organic solvents Multiple sites Multifocal neuropathology Mental disorders

CNS atrophy

Carbon Dioxide posterior cingulate cortex, bilateral inferior parietal regions, medial

prefrontal cortex, and medial

temporal lobe

Increase in low frequency power Reduced brain arousal

Radon-222 Amygdala, hippocampus, temporal

lobe, frontal lobe, occipital lobe,

parietal lobe, substantia nigra, locus ceruleus, nucleus basalis

Possible diffuse background slowing Alzheimer’s disease

Cesium-137 No publications Lead & Cadmium Frontal regions

P3 and P4

Thalamus, lentiform nucleus, external capsules, and sub-cortical

white matter.

Cerebellar white matter

Increase slow wave activity and

decrease in EEG amplitude

Hypercoherence Increased θ-band relative power in

posterior areas

More α and β activities

Seizures, irritability, headache,

mental dullness and attention

difficulty, memory loss, tremor, and hallucinations

Nitrous Oxide Frontal areas Increased θ-band, β at the 40-50 Hz,

and 70-110 Hz band powers

Atrophy, changes in corpus

callosum,

Extensive white matter change in both hemispheres

BTEX δ1, β1 and β2, θ, and α bands giddiness

Carbon Monoxide Lateralised sharp waves, Bilateral white matter change

No changes in EEG reported Focal seizures

Sulfur Dioxide & hydrogen Sulfide hippocampus Increased δ-band and Θ-band

activity

motor, behavioural, and cognitive

deficits dementia

Formaldehyde & Glutaraldehyde unclear Possible increased δ-band and Θ-

band activity with glutaraldehyde

headaches, dizziness, apathy, and

inability to concentrate

Sarin General slowing of EEG desynchronised EEG with increased

β-band activity, increased δ and θ band slowing, and decreased α-band

activity

convulsions and muscular paralysis



Styrene Posterior regions excessive diffuse θ-band activity and

localized slow waves; Bilateral spike and wave discharges;

Focal slow wave activity

decreased REM percent and longer

REM onset latency

Diesel Fp1 and Fp2 and F3, F4 C3 and C4

P3 and P4

decreased θ and α, and increased β and γ-band EEG changes via vagal

reflexes from the lungs

Alzheimer’s disease

4. Prevention and Treatment Options for Toxic Encephalopathy

Foremost is the removal of the toxic substance from the environment or in most instances removing the person from the

source of the toxin. Regular detoxing by workers exposed to fracking may reduce exposure caused neuropathy. Toxic

encephalopathy needs to be carefully investigated as it may be asymptomatic or signs and symptoms of toxin exposure may have a

similar presentation to psychiatric signs and symptoms, metabolic, inflammatory, carcinogenic and neuropathological signs of

diverse etiology (Schaumburg & Spencer, 1987). Therefore, it is essential to have an accurate occupational history and detailed

neurological examination result. Treatment is then based on the extent of the signs and symptoms displayed by the patient and the

toxic substance causing the neuropathology (Berisavac et al., 2017).

Prevention should be considered in all cases by reducing exposure to toxins and providing sufficient safeguards such as

protective clothing. Neurofeedback is an adjunct or alternative therapy often found efficacious for epilepsy, anxiety and psychosis

often associated with toxic encephalopathy. Neurofeedback treatment protocols focus on training alpha, beta, delta, theta, and

gamma or a combination of these. Alternative modes include amplitude training or z-score training amongst others. The observation

that seizure incidence due to inhalation of lunar landing fuel by NASA astronauts was lowered significantly by sensory-motor

rhythm neurofeedback training (Sterman & Egner, 2006) suggests that neurofeedback may be useful in normalising EEG rhythm

patterns or reducing the incidence of toxic encephalopathy in toxin exposed workers by enhancing neuronal or brain reserve by

enhancing the capacity to compensate for dysfunction. However, caution needs to be exercised when providing neurofeedback as

an incorrect protocol may result in an increased blood flow to the brain and hence increase the toxic effects of the chemicals. This

application of preventative neurofeedback requires further research to address the increasing use of industrial chemicals.

References

Altwegg-Boussac, T., Schramm, A. E., Ballestero, J., Grosselin, F., Chavez, M., Lecas, S., et al. (2017). Cortical neurons and networks are dormant but fully

responsive during isoelectric brain state. Brain, 140(9), 2381-2398. doi: 10.1093/brain/awx175

Andraus, M. E. C., & Alves-Leon, S. V. (2011). Non-epileptiform EEG abnormalities: an overview. Arq Neuro-Psiquiatr, 69, 829-835.

Bell, I. R., Bootzin, R. R., Ritenbaugh, C., Wyatt, J. K., DeGiovanni, G., Kulinovich, T., et al. (1996). A polysomnographic study of sleep disturbance in community

elderly with self-reported environmental chemical odor intolerance. Biol Psychiatry, 40(2), 123-133. doi: 10.1016/0006-3223(95)00330-4

Benignus, V. A., Otto, D. A., Muller, K. E., & Seiple, K. J. (1981). Effects of age and body lead burden on CNS function in young children. II. EEG spectra.

Electroencephalogr Clin Neurophysiol, 52(4), 240-248. doi: 10.1016/0013-4694(81)90053-5

Bergman, K. (1979). Whole-body autoradiography and allied tracer techniques in distribution and elimination studies of some organic solvents: benzene, toluene,

xylene, styrene, methylene chloride, chloroform, carbon tetrachloride and trichloroethylene. Scand J Work Environ Health, 5 Suppl 1, 1-263.

Berisavac, I. I., Jovanović, D. R., Padjen, V. V., Ercegovac, M. D., Stanarčević, P. D. J., Budimkić-Stefanović, M. S., et al. (2017). How to recognize and treat

metabolic encephalopathy in Neurology intensive care unit. [NI Feature: The Quest - Commentary]. Neurol India, 65(1), 123-128. doi: 10.4103/0028-3886.198192

Campa, M. F., Techtmann, S. M., Gibson, C. M., Zhu, X., Patterson, M., Garcia de Matos Amaral, A., et al. (2018). Impacts of glutaraldehyde on microbial

community structure and degradation potential in streams impacted by hydraulic fracturing. Environ Sci Technol, 52(10), 5989-5999. doi: 10.1021/acs.est.8b00239

Cantor, D. S. (2000). Neurofunctional status of community residents evidencing toxic lead levels. Arch Clin Neuropsychol, 15, 725-726.

Clinical Trials.gov. NIH U.S. National Library of Medicine, 2019.

Colborn, T., Kwiatkowski, C., Schultz, K., & Bachran, M. (2001). Natural gas operations from a public health perspective. Human Ecol Risk Ass, 17(5), 1039-1056.

doi: 10.1080/10807039.2011.605662

Crüts, B., van Etten, L., Törnqvist, H., Blomberg, A., Sandström, T., Mills, N. L., et al. (2008). Exposure to diesel exhaust induces changes in EEG in human

volunteers. Particle Fibre Toxicol, 5, 4-4. doi: 10.1186/1743-8977-5-4

Cummings, M., Dougherty, D. W., Mohile, N. A., Walter, K. A., Usuki, K. Y., & Milano, M. T. (2016). Severe radiation-induced leukoencephalopathy: Case report

and literature review. Adv Rad Oncol, 1(1), 17-20. doi: 10.1016/j.adro.2016.01.002

Dobbs, M. R. (2011). Toxic encephalopathy. Seminars Neurol, 31(2), 184-193. doi: 10.1055/s-0031-1277989

Duffy, F. H., Burchfiel, J. L., Bartels, P. H., Gaon, M., & Sim, V. M. (1979). Long-term effects of an organophosphate upon the human electroencephalogram.

Toxicol Appl Pharmacol, 47(1), 161-176. doi: 10.1016/0041-008x(79)90083-8

Dworak, M., McCarley, R. W., Kim, T., & Basheer, R. (2011). Delta oscillations induced by ketamine increase energy levels in sleep-wake related brain regions.

Neuroscience, 197, 72-79. doi: 10.1016/j.neuroscience.2011.09.027



Esen, O., Özer, S. C., Soylu, A., Rend, A. R., & Fisne, A. (2018). An investigation of the coal seam gas content and composition in Soma Coal Basin, Turkey.

Proceedings of the 18th Coal Operators' Conference, Mining Engineering, University of Wollongong. 284-293.

Estrin, W. J., Moore, P., Letz, R., & Wasch, H. H. (1988). The P-300 event-related potential in experimental nitrous oxide exposure. Clin Pharmacol Ther, 43(1),

86-90. doi: 10.1038/clpt.1988.15

Fel'dman, Y. G., & Bonashevskaya, T. I. (1971). On the effects of low concentrations of formaldehyde. Hyg Sanit (USSR) 36(5), 174–180.

Filley, C. M., Halliday, W., & Kleinschmidt-Demasters, B. K. (2004). The effects of toluene on the central nervous system. J Neuropath Exp Neurol, 63(1), 1-12.

doi: 10.1093/jnen/63.1.1

Firestone, J. A., & Longstrength, W. T. J. (2004). Neurologic and psychiatric disorders. In L. Rosenstock, M. Cullen, C. Brodkin & C. Redlich (Eds.), Textbook of

clinical occupational and environmental medicine (4 ed., pp. 645–660). Philadelphia: Saunders.

Gamberale, F., Annwall, G., & Hultengren, M. (1978). Exposure to xylene and ethylbenzene. III. Effects on central nervous functions. Scand J Work, Environ

Health, (3), 204-211. doi: 10.5271/sjweh.2705

Indulski, J. A., Sińczuk-Walczak, H., Szymczak, M., & Wesołowski, W. (1996). Neurological and neurophysiological examinations of workers occupationally

exposed to organic solvent mixtures used in the paint and varnish production. Int J Occ Med Environ Health, 9(3), 235-244.

Kaden, D. A., Mandin, C., Nielsen, G. D., & Wolkoff, P. (2010). Formaldehyde. WHO Guidelines for Indoor Air Quality: Selected Pollutants, Geneva.

https://www.ncbi.nlm.nih.gov/books/NBK138711/.

Kety, S. S., & Schmidt, C. F. (1946). The effects of active and passive hyperventilation on cerebral blood flow, cerebral oxygen consumption, cardiac output, and

blood pressure of normal young men. J Clin Invest, 25(1), 107-119. doi: 10.1172/JCI101680

Kim, Y., & Kim, J. W. (2012). Toxic encephalopathy. Saf Health Work, 3(4), 243-256. doi: 10.5491/SHAW.2012.3.4.243

Koo, B. (2001). EEG changes with vagus nerve stimulation. J Clin Neurophysiol, 18(5), 434-441.

Kovala, T., Matikainen, E., Mannelin, T., Erkkilä, J., Riihimäki, V., Hänninen, H., et al. (1997). Effects of low level exposure to lead on neurophysiological functions among lead battery workers. Occup Environ Med, 54(7), 487-493. doi: 10.1136/oem.54.7.487

Landrigan, P. J., & Todd, A. C. (1994). Lead poisoning. West J Med, 161(2), 153-159.

Langman, J. M. (1994). Xylene: its toxicity, measurement of exposure levels, absorption, metabolism and clearance. Pathology, 26(3), 301-309. doi:

10.1080/00313029400169711

Lewine, J. D., Paulson, K., Bangera, N., & Simon, B. J. (2019). Exploration of the impact of brief noninvasive vagal nerve stimulation on EEG and event-related

potentials. Neuromodulation, 22(5), 564-572. doi: 10.1111/ner.12864

Lin, W., Yuan, D., Deng, Z., Niu, B., & Chen, Q. (2019). The cellular and molecular mechanism of glutaraldehyde-didecyl dimethyl ammonium bromide as a

disinfectant against Candida albicans. J Appl Microbiol, 126(1), 102-112. doi: 10.1111/jam.14142

Lloyd-Smith, M., & Senjen, R. (2011). Hydraulic fracturing in coal seam gas: the risks to our health, communities, environment and climate. National Toxics

Network, Bangalow, Australia. https://ntn.org.au/wp-content/uploads/2011/04/NTN-Fracking-Briefing-Paper-April-2011.pdf.

Lloyd-Smith, M., & Senjen, R. (2016). Unconventional gas exploration and production: human health impacts and environmental legacy. National Toxics

Network,Bangalow, Australia: https://ntn.org.au/wp-content/uploads/2018/02/NTN-Unconventional-Gas-Report-April-2016-1.pdf.

Mani, J., Chaudhary, N., Kanjalkar, M., & Shah, P. U. (1998). Cerebellar ataxia due to lead encephalopathy in an adult. J Neurol Neurosurg Psychiatry, 65(5), 797. doi: 10.1136/jnnp.65.5.797

McCarron, G. (2013). Symptomatology of a gas field - An independent health survey in the Tara rural residential estates and environs. Unconventional Gas Mining Submission 12 - Attachment 2, https://www.r-cause.net/uploads/8/0/2/5/8025484/tara_health_report.pdf.

McQueen, D. S., Donaldson, K., Bond, S. M., McNeilly, J. D., Newman, S., Barton, N. J., et al. (2007). Bilateral vagotomy or atropine pre-treatment reduces

experimental diesel-soot induced lung inflammation. Toxicol Appl Pharmacol, 219(1), 62-71. doi: https://doi.org/10.1016/j.taap.2006.11.034

Miller, C. S. (2001). The compelling anomaly of chemical intolerance. Ann New York Acad Sci, 933(1), 1-23. doi: 10.1111/j.1749-6632.2001.tb05810.x

Momčilović, B., Lykken, G. I., & Cooley, M. (2006). Natural distribution of environmental radon daughters in the different brain areas of an Alzheimer Disease

victim. Mol Neurodegen, 1(1), 11. doi: 10.1186/1750-1326-1-11

Neufeld, M. Y., Swanson, J. W., & Klass, D. W. (1981). Localized EEG abnormalities in acute carbon monoxide poisoning. Arch Neurol, 38(8), 524-527. doi:

10.1001/archneur.1981.00510080086014

NRC. (1980). National Research Council. Committee on toxicology. Formaldehyde - an assessment of its health effects (pp.

https://www.ncbi.nlm.nih.gov/books/NBK217652/). Washington DC.

Oberdorster, G., Sharp, Z., Atudorei, V., Elder, A., Gelein, R., Kreyling, W., et al. (2004). Translocation of inhaled ultrafine particles to the brain. Inhal Toxicol,16(6-

7), 437-445. doi: 10.1080/08958370490439597

Otto, D. A., Benignus, V. A., Muller, K. E., & Barton, C. N. (1981). Effects of age and body lead burden on CNS function in young children. I. slow cortical

potentials. Electroencephalogr Clin Neurophysiol, 52(4), 229-239. doi: 10.1016/0013-4694(81)90052-3

https://www.ncbi.nlm.nih.gov/books/NBK138711/

https://ntn.org.au/wp-content/uploads/2011/04/NTN-Fracking-Briefing-Paper-April-2011.pdf

https://ntn.org.au/wp-content/uploads/2018/02/NTN-Unconventional-Gas-Report-April-2016-1.pdf

https://www.r-cause.net/uploads/8/0/2/5/8025484/tara_health_report.pdf

https://doi.org/10.1016/j.taap.2006.11.034



Palmer, R. C., Short, D., & Auch, W. E. T. (2018). The human right to water and unconventional energy. Int J Environ Res Public Health, 15(9),

1858.doi:10.3390/ijerph15091858

Perugula, M. L., & Lippmann, S. (2016). Encephalopathy or psychosis? Innov Clin Neurosci, 13(11-12), 41-42.

Pool, R., & Rusch, E. (2014). Roundtable on Environmental Health Sciences, Research, and Medicine; Identifying and Reducing Environmental Health Risks of

Chemicals in Our Society: Workshop Summary (Vol. https://www.ncbi.nlm.nih.gov/books/NBK268889/). Washington DC: National Academies Press.

Posner, J. B., & Plum, F. (1960). The toxic effects of carbon dioxide and acetazolamide in hepatic encephalopathy. J Clin Invest, 39(8), 1246-1258. doi:

10.1172/JCI104140

Proietti, L., Longo, B., & Duscio, D. (2002). Suspected glutaraldehyde poisoning: a case report. Med Lav, 93(1), 43-47.

Rampil, I. J., Kim, J. S., Lenhardt, R., Negishi, C., & Sessler, D. I. (1998). Bispectral EEG index during nitrous oxide administration. Anesthesiol, 89(3), 671-677.

doi: 10.1097/00000542-199809000-00017

Rao, J. V. B., Vengamma, B., Naveen, T., & Naveen, V. (2014). Lead encephalopathy in adults. J Neurosci Rural Pract, 5(2), 161-163. doi: 10.4103/0976-

3147.131665

Satran, R., & Dodson, V. N. (1963). Toluene habituation. Report of a case. N Engl J Med, 268, 719-721. doi: 10.1056/nejm196303282681308

Schaumburg, H. H., & Spencer, P. S. (1987). Recognizing neurotoxic disease. Neurology, 37(2), 276-278. doi: 10.1212/wnl.37.2.276

Seppäläinen, A. M. (1988). Neurophysiological approaches to the detection of early neurotoxicity in humans. Crit Rev Toxicol, 18(4), 245-298. doi:

10.3109/10408448809037468

Seppäläinen, A. M., Laine, A., Salmi, T., Verkkala, E., Riihimaki, V., & Luukkonen, R. (1991). Electroencephalographic findings during experimental human

exposure to m-xylene. Arch Environ Health, 46(1), 16-24. doi: 10.1080/00039896.1991.9937424

Silbergeld, E. K., Miller, L. P., Kennedy, S., & Eng, N. (1979). Lead, GABA, and seizures: effects of subencephalopathic lead exposure on seizure sensitivity and

GABAergic function. Environ Res, 19(2), 371-382. doi: 10.1016/0013-9351(79)90062-8

Smith, S. J. M. (2005). EEG in neurological conditions other than epilepsy: when does it help, what does it add? J Neurol Neurosurg Psychiatry, 76(suppl 2), ii8-

ii12. doi: 10.1136/jnnp.2005.068486

Somerville, W. (2014). CSG and your health: Understand the risks, protect your family. www.creeksbend.com/csg.

Sorg, B. A., Tschirgi, M. L., Swindell, S., Chen, L., & Fang, J. (2001). Repeated formaldehyde effects in an animal model for multiple chemical sensitivity. Ann N Y Acad Sci, 933, 57-67. doi: 10.1111/j.1749-6632.2001.tb05814.x

Sterman, M. B., & Egner, T. (2006). Foundation and practice of neurofeedback for the treatment of epilepsy. Appl Psychophysiol Biofeedback, 31(1), 21-35. doi:

10.1007/s10484-006-9002-x

Tait, D. R., Santos, I. R., Maher, D. T., Cyronak, T. J., & Davis, R. J. (2013). Enrichment of radon and carbon dioxide in the open atmosphere of an Australian coal

seam gas field. Environ Sci Technol, 47(7), 3099-3104. doi: 10.1021/es304538g

Takeuchi, Y., & Hisanaga, N. (1977). The neurotoxicity of toluene: EEG changes in rats exposed to various concentrations. Br J Ind Med, 34(4), 314-324. doi:

10.1136/oem.34.4.314

Tapeantong, T., & Poungvarin, N. (2009). Delayed encephalopathy and cognitive sequelae after acute carbon monoxide poisoning: report of a case and review of

the literature. J Med Assoc Thai, 92(10), 1374-1379.

Thatcher, R. W., & Lester, M. L. (1985). Nutrition, Environmental toxins and Computerized EEG: A Mini-Max Approach to Learning Disabilities. J Learn Dis,

18(5), 287-297. doi: 10.1177/002221948501800508

Thatcher, R. W., McAlaster, R., Lester, M. L., & Cantor, D. S. (1984). Comparisons among EEG, hair minerals and diet predictions of reading performance in

children. Ann N Y Acad Sci, 433(1), 87-96. doi: 10.1111/j.1749-6632.1984.tb14761.x

Tiffert, T., Garcia-Sancho, J., & Lew, V. L. (1984). Irreversible ATP depletion caused by low concentrations of formaldehyde and of calcium-chelator esters in

intact human red cells. Biochim Biophys Acta, 773(1), 143-156. doi: 10.1016/0005-2736(84)90559-5

Tzoulis, C., Neckelmann, G., Mørk, S. J., Engelsen, B. E., Viscomi, C., Moen, G., et al. (2010). Localized cerebral energy failure in DNA polymerase gamma-

associated encephalopathy syndromes. Brain, 133(5), 1428-1437. doi: 10.1093/brain/awq067

Valk, J., & van der Knaap, M. S. (1992). Toxic encephalopathy. AJNR Am J Neuroradiol, 13(2), 747-760.

van Birgelen, A. P. J. M., Chou, B. J., Renne, R. A., Grumbein, S. L., Roycroft, J. H., Hailey, J. R., et al. (2000). Effects of glutaraldehyde in a 2-year inhalation

study in rats and mice. Toxicol Sci, 55(1), 195-205. doi: 10.1093/toxsci/55.1.195

Waxman, H. A., Markey, E. J., & DeGette, D. (2011). Chemicals used in hydraulic fracturing. United States House of Representatives Committee on Energy and

Commerce Minority Staff: https://www.voltairenet.org/IMG/pdf/Hydraulic-Fracturing-Chemicals.pdf.

Webb, E., Moon, J., Dyrszka, L., Rodriguez, B., Cox, C., Patisaul, H., et al. (2018). Neurodevelopmental and neurological effects of chemicals associated with

unconventional oil and natural gas operations and their potential effects on infants and children. Rev Environ Health, 33(1), 3-29. doi: 10.1515/reveh-2017-0008

Woolley, D. E. (1984). A perspective of lead poisoning in antiquity and the present. Neurotoxicol, 5(3), 353-361.

http://www.creeksbend.com/csg

https://www.voltairenet.org/IMG/pdf/Hydraulic-Fracturing-Chemicals.pdf



Xu, F., Uh, J., Brier, M. R., Hart, J., Jr., Yezhuvath, U. S., Gu, H., et al. (2011). The influence of carbon dioxide on brain activity and metabolism in conscious

humans. J Cereb Blood Flow Metab, 31(1), 58-67. doi: 10.1038/jcbfm.2010.153

Zhang, X., Vincent, A. S., Halliwell, B., & Wong, K. P. (2004). A mechanism of sulfite neurotoxicity: direct inhibition of glutamate dehydrogenase. J Biol Chem,

279(41), 43035-43045. doi: 10.1074/jbc.M40275920054

Areej Babiker /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, No 1, pages 022-026 22


Neural Network for Detection of Students’ interest using Approximate

Entropy Features of EEG data

Areej Babikera*, Ibrahima Fayeb, Aamir Saeed Malikc

aDepartment of Electrical and Electronics Engineering Universiti Teknologi PETRONAS Malaysia bDepartment of Fundamental and Applied Science Universiti Teknologi PETRONAS Malaysia

cDepartment of Electrical and Electronics Engineering Universiti Teknologi PETRONAS Malaysia

Abstract

Studies have found that optimizing course contents or lecture materials according to student’s interest could improve learning outcomes.

In this work, we advocate the use of Electroencephalogram (EEG) as a modality to qualify student’s interest based on their respective brain

activities. An EEG-based detection of interest in real classroom environment using Approximate Entropy (ApEn) is proposed. The method involved

the use of Artificial Neural Networks (ANN) to discriminate the EEG data as relevant to either high or low situational interest. The qualitative

assessment was performed by Personal Interest (PI), Situational interest (SI) questionnaires and knowledge tests which were considered as ground

truth for the classifications. An accuracy of 100% and R2=0.996 was achieved in classifying 17 students as high SI or low SI. The results show that the proposed features could be used to differentiate brain activities based on student’s interest.



Keywords: Situational interest, EEG, classroom, Approximate Entropy and ANN.

ARTICLE INFO

RECEIVED December 2019

REVIEWED February 2020

ACCEPTED April 2020






1. Introduction

Researchers divided interest into personal and situational interest. while both lead to focused attention on the object of

interest, situational interest is easily triggered and controlled by the situation. Therefore, many studies were carried out to explain

the science of interest development and how to utilize it for the benefit of learners. For a field like mathematics “Proficiency in

mathematics is a major advantage in industrialized nations” (Maloney, Schaeffer, & Beilock, 2013) where many students are scared

of and not willing to do the effort, situational interest have remarkable effect. In recent study (Bernacki & Walkington, 2018), it was

reported that the increase in situational interest caused by personalizing 4 units of algebra story problems to students led to greater

interest in mathematics and higher score in class tests for those students compared to control group.

One way to study this interest is by studying the underlying brain activities. Electroencephalogram (EEG) is among the most

effective non-invasive equipment in this case due to the advanced technology that made it available with relatively low cost.

Moreover, the signals produced by EEG cannot be manipulated by participant unlike questionnaires. Therefore, its used extensively

in classroom experiments to detect student’s attention and engagement e.g. (Poulsen, Kamronn, Dmochowski, Parra, & Hansen,

2017), (Ko, Komarov, Hairston, Jung, & Lin, 2017) and (Sezer, Inel, Çağdaş Seçkin, & Uluçınar, 2015). These studies presented

the potential of EEG to indicate attention or engagement but didn’t investigate the possibility of classifying attentive and inattentive

students during stimulation or in real time.

To analyze the EEG data, Approximate Entropy (ApEn) is proposed. It is an index of signal disorder and complexity which

makes it proper for time series data that is changing overtime. In studying the cognitive load and brain activities ApEn is

recommended (LI, JIANG, HONG, DONG, & YAO, 2016), (Cavanaugh, Mercer, & Stergiou, 2007). The features extracted by

ApEn are then fed into Artificial Neural Networks (ANN) classifier to design and validate the model that can best show the

differences between high and low SI students. ApEn and ANN are popular in classification process for Brain Computer Interface

(BCI) systems. ApEn is widely used in analyzing physiological time series data (Alcaraz & Rieta, 2010) because of its robustness

on moderate length time series data while ANN has a great ability to mimic the neural activities of the brain and achieving high

accuracy in cognitive task classifications reaching 98.8% in (Zarjam, Epps, & Lovell, 2015), average of 90% in (Heger, Putze, &

Schultz, 2010), (Mazher, Aziz, Malik, & Amin, 2017). Moreover, ANN is non-parametric method which eliminates possible errors

resulted from parameter estimation. It also suites well non-linear data such as EEG.

The structure of this paper is as following: section 2 describes the materials and methods used in this study, section 3

presents the result of applying the proposed methods and discusses its implications. Section 4 concludes the paper.

2. Materials and Methods

2.1 Participants

The participants of this experiment were engineering students from Universiti Teknologi PETRONAS (UTP) first-year

undergraduate, who have no records of mental illness and not under any medication. The experimental procedure was approved by

UTP Ethical Committee. The participants were selected based on a questionnaire regarding joining mathematics club in the

university. The questionnaire was run as a pre-evaluation for the level of personal/ individual interest of students and hence a

balanced group with high, low and moderate interest participants were selected. Some recent studies showed no significant

differences between females and males in terms of academic performance e.g. (Moldovan, Ghergulescu, & Muntean, 2017) and

therefore was not considered in this study. Overall 30 students (4 females) participated in this experiment. Selected participants

were notified through emails and thoroughly briefed about the experiment and each participant was compensated for their time.

Upon the arrival of participants, researchers assisted them in wearing the 8-channels Enobio EEG cap. After that, baseline

data of 4 min eyes opened and 4 min eyes closed were acquired followed by about 22min of lecture. The presentation was delivered

through projector to a projector screen. After the EEG recording, Situational Interest (SI) questionnaire and post-knowledge test

were undertaken.

From the situational interest questionnaire result, subjects scored ≥ 77 out of 100 were considered high SI students while

subject scored ≤ 63 out of 100 were considered low SI students. Subjects scored between 76 and 63 were considered to have

moderate situational interest. This evaluation was selected based on trial and error to achieve proper separation between classes/

groups while maintaining proper number of subjects in each group. This procedure yielded 17 subjects, 10 high SI and 7 low SI

subjects. Presentation questionnaire was used to validate the interestingness of the lecture as described in the below section.

2.2 Stimuli

A lecture on Laplace transform (duration about 22 min) that followed the curriculum for Ordinary Differential Equation

(ODE) course was designed and presented in an interesting way to first-year undergraduate students. This was done by relating the

materials to everyday life through a careful selection of a Laplace applications and meaningful content throughout the lecture. Using

colors, fonts, shapes and pictures along with the animation to facilitate presenting the contents was helpful. A brief history of Laplace

transform was added to the original lecture. The presentation slideshow was sent to experts for improvement and feedback. It was

then tested with different group of students to rate the interesting points and improve the slides further. The novelty of the content

i.e. being new to participants, was ensured by selecting participants who have not taken the course or the lecture before.

To test the effect of these stimuli during the experiments, a questionnaire was run at the end of each session followed by a

verbal, non-formal interview. The questionnaire result showed that 83.34% agree that the lecture was interesting, and all participants

agree that they look forward for similar lectures, suggesting the stimuli was interesting for majority of participants.



2.3 Methods

Entropy can extract very useful information from EEG especially when changes in time is expected to occur frequently.

Entropy value is higher when signal irregularity is higher and entropy value becomes lower with regular signal (Wei et al, 2013).

High SI students are hypothesized to have high ApEn value while low SI students will have low ApEn value. Two

parameters were specified to calculate the ApEn, the tolerance: r= 20% of the standard deviation and the embedding dimension:

m=2 which are preferred values according to similar previous studies (Vega et al, 2013), (Hosseini et al, 2011). The ApEn was

calculated as following:

𝐴𝑝𝐸𝑛(𝑚, 𝑟, 𝑁) = 𝛷𝑚(𝑟) − 𝛷𝑚+1(𝑟) (1)

where 𝝓m(r) is denoted to be:

(2)

N is the number of points and ln is the natural logarithm of the correlation integral Ci m(r).

Since EEG is subjective measure and differ from one to another, the baseline data (eyes opened) were used to obtain reliable

measurement for each subject and to avoid such subjectivity. The baseline condition data were processed in similar way to the

lecture condition to obtain the ApEn features. The percentage of change was calculated by subtracting this baseline-ApEn from each

lecture-ApEn values then divide the result by the baseline-ApEn. This procedure was performed for each of the 17 subjects (10 high

SI and 7 low SI).

ANN was used to classify the features obtained by ApEn as either high or low SI. It is consisted of multiple artificial

neurons that mimic the information processing and interactions of the brain. The neurons work together to form nonlinear decision

boundaries which makes it suitable for the non-stationary and nonlinear EEG data. Therefore, it can recognize and create non-linear

relationship between the input data and/ or outputs to yield optimum pattern recognition result.

Six networks shown in Table 1 were designed and tested to obtain the optimum classification result.

Table 1. The architecture of the designed neural networks

Layers Net 1 Net 2 Net 3 Net 4 Net 5

Input 39 39 39 39 39

Hidden 5 10 15 20 25

Output 2 2 2 2 2

The input layers in all the networks contained 39 neurons corresponding to the 20 min of the lecture which were segmented

into 30 s comprising 40 points. The first 30 s was removed because it contained some noise at the start of recording, hence, the input

layer consisted of 39 neurons. The network had two classification outputs which are high and low SI. The number of neurons in the

hidden layer was determined by increasing 5 neurons each time until reaching the least mean square error (MSE) and coefficient of

determination (R2).

The data were randomly divided such that 6 subjects out of 17 subjects were used for model validation and therefore were

not part of the training data that used to design the model (remaining 11 subjects) which is about 35.3% of data used for validation.

During designing the model (training phase), the training dataset (11 subjects) was further divided randomly into 70% for training,

15% for validation and 15% for testing the model. The training process was repeated several times with different parameters each

time to obtain the best performance with the least Mean Square Error (MSE) and coefficient of determination R2. The best network

model was exported and tested with the validation dataset (6 subjects) that were not part of the training set.

MATLAB 2017 toolbox was used for the implementation of ApEn and ANN.

3. Results and Discussion

This section is divided into two divisions that study the result of employing ApEn and the result of classification using

ANN.

3.1 ApEn Result

The result of employing ApEn through about 20 min is shown in Figure 1. The figure shows the ApEn values for the

average of randomly selected 7 high and 7 low SI subjects. The first 8 min doesn’t have consistent differences between the two

classes. After that, clear differences indicating high values for high SI subjects compared to low SI as expected. These high values

that show the complexity of the signal due to the brain activities are also found at the 3rd and 7th minute for the high SI subjects.

For the low SI subjects, the values were high only at the beginning of the lecture and decreased significantly after the 8th minute

till the end of the lecture. It suggests that high SI subjects could keep their brain performance till the end of the lecture unlike the

low SI ones.



Fig 1. Averaged ApEn values between high and low SI subjects. After the 8th minute, clear difference in brain activities is observed between both

groups.

3.2 Classification Result

The MSE value was decreasing by increasing the number of neurons until reaching 30 neurons where the MSE started to

increase. Figure 2 shows the performance of each network described in Table 1.

Fig 2. The performance of the designed ANN networks to classify high and low situational interested students based on ApEn features. The horizontal

axis represents the 6 designed networks while the vertical axis shows the MSE and R2 values for each network.

Figure 2 shows the effect of the size of neural network on the system performance using two statistical standards the MSE

and R2. Network 3 and Network4 had comparable performance while the least error was achieved with 20 neurons (Network4)

MSE = 8.7E-04, R2 = 0.996. Network5 and Network6 had the highest MSE because of the generalization drop effect that could be

caused by overfitting. Increasing the number of neurons further increases the time required for processing. Therefore, 20 neurons

network was found to give the best performance with less processing time.

The obtained model was validated using new 6 subjects (division of data into training and validation sets is described in

methods section). The proposed algorithm has achieved 100% accuracy in classifying correctly the 6 subjects as high or low SI

which is higher than the nearest similar study (Zhu et al, 2017) that achieved 98.99% in classifying high and low SI students. It also

showed that ApEn is suitable for describing the changes of EEG signals that could be caused by cognitive functions occurring in

the brains. ANN was found suitable for training the model with the acquired features using the artificial intelligence to understand

the relationship between the values of ApEn and the output classes.

This study has some limitations. Even though the selected sample size was 30 subjects, only 17 of them were classified as

high and low SI based on the criteria described in participants section. Therefore, future studies are encouraged to increase the

number of samples to confirm and support the result obtained in this paper. It is also highly recommended to perform the experiment

for different subjects/ materials, for example, science or art and compare the results to gain better understanding about interest

phenomenon.

4. Conclusion

Increasing students’ situational interest during learning focuses their attention and therefore maximizes their learning and

knowledge gain. Assessing the performance of these students in terms of high SI and low SI, and feeding back this information to

lecturers and educators assist them to improve teaching strategies in a way that triggers and maintain student’s interest in the class.

ApEn features obtained for the raw EEG data and fed into ANN model achieved optimum result with R2 = 0.996. This result suggests

the possibility of a real-time monitoring of student’s situational interest in classroom.

5. Acknowledgement

The authors would like to acknowledge the support of Ministry of Higher Education Malaysia for HiCoE grant for CISIR

(0153CA-002).

-1

-0.5

0

0.5

1 2 3 4 5 6 7 8 9 10111213141516171819

Ave

rage

Ap

En

Time (min)

Hi interest Low interest



References

Alcaraz, R., & Rieta, J. J. (2010). A review on sample entropy applications for the non-invasive analysis of atrial fibrillation electrocardiograms. Biomedical Signal

Processing and Control, 5(1), 1–14. https://doi.org/https://doi.org/10.1016/j.bspc.2009.11.001

Bernacki, M. L., & Walkington, C. (2018). The role of situational interest in personalized learning. Journal of Educational Psychology, 110(6), 864–881.

https://doi.org/10.1037/edu0000250

Cavanaugh, J. T., Mercer, V. S., & Stergiou, N. (2007). Approximate entropy detects the effect of a secondary cognitive task on postural control in healthy young

adults: a methodological report. Journal of NeuroEngineering and Rehabilitation, 4, 42. https://doi.org/10.1186/1743-0003-4-42

Heger, D., Putze, F., & Schultz, T. (2010). Online Workload Recognition from EEG Data during Cognitive Tests and Human-Machine Interaction BT - KI 2010:

Advances in Artificial Intelligence. In R. Dillmann, J. Beyerer, U. D. Hanebeck, & T. Schultz (Eds.) (pp. 410–417). Berlin, Heidelberg: Springer Berlin Heidelberg.

Hosseini, S. A., & Mohammad Bagher Naghibi-Sistani. (2011). Emotion recognition method using entropy analysis of EEG signals. International Journal of Image,

Graphics and Signal Processing, 5, 30–36.

Ko, L.-W., Komarov, O., Hairston, W. D., Jung, T.-P., & Lin, C.-T. (2017). Sustained Attention in Real Classroom Settings: An EEG Study. Frontiers in Human

Neuroscience, 11, 388. https://doi.org/10.3389/fnhum.2017.00388

LI, X., JIANG, Y., HONG, J. U. N., DONG, Y., & YAO, L. E. I. (2016). ESTIMATION OF COGNITIVE WORKLOAD BY APPROXIMATE ENTROPY OF

EEG. Journal of Mechanics in Medicine and Biology, 16(06), 1650077. https://doi.org/10.1142/S0219519416500779

Maloney, E. A., Schaeffer, M. W., & Beilock, S. L. (2013). Mathematics anxiety and stereotype threat: shared mechanisms, negative consequences and promising

interventions. Research in Mathematics Education, 15(2), 115–128. https://doi.org/10.1080/14794802.2013.797744

Mazher, M., Aziz, A. A., Malik, A. S., & Amin, H. U. (2017). An EEG-Based Cognitive Load Assessment in Multimedia Learning Using Feature Extraction and

Partial Directed Coherence. IEEE Access, 5, 14819–14829. https://doi.org/10.1109/ACCESS.2017.2731784

Moldovan, A.-N., Ghergulescu, I., & Muntean, C. H. (2017). Analysis of Learner Interest, QoE and EEG-Based Affective States in Multimedia Mobile Learning,

398–402. https://doi.org/10.1109/icalt.2017.93

Poulsen, A. T., Kamronn, S., Dmochowski, J., Parra, L. C., & Hansen, L. K. (2017). EEG in the classroom: Synchronised neural recordings during video presentation.

Scientific Reports, 7, 43916. https://doi.org/10.1038/srep43916

Sezer, A., Inel, Y., Çağdaş Seçkin, A., & Uluçınar, U. (2015). An Investigation of University Students’ Attention Levels in Real Classroom Settings with NeuroSky’s

MindWave Mobile (EEG) Device. In Conference: International Educational Technology Conference–IETC 2015At: Istanbul, Turkey.

Vega, C. H. F., Noel, J., & Fernández, J. R. (2013). Cognitive task discrimination using approximate entropy (ApEn) on EEG signals. In 2013 ISSNIP Biosignals

and Biorobotics Conference: Biosignals and Robotics for Better and Safer Living (BRC) (pp. 1–4). https://doi.org/10.1109/BRC.2013.6487521

Wei, Q., Liu, Q., Fan, S.-Z., Lu, C.-W., Lin, T.-Y., Abbod, F. M., & Shieh, J.-S. (2013). Analysis of EEG via Multivariate Empirical Mode Decomposition for Depth

of Anesthesia Based on Sample Entropy. Entropy . https://doi.org/10.3390/e15093458

Zarjam, P., Epps, J., & Lovell, N. H. (2015). Beyond Subjective Self-Rating: EEG Signal Classification of Cognitive Workload. IEEE Transactions on Autonomous

Mental Development, 7(4), 301–310. https://doi.org/10.1109/TAMD.2015.2441960

Zhu, Z., Ober, S., & Jafari, R. (2017). Modeling and detecting student attention and interest level using wearable computers. In 2017 IEEE 14th International

Conference on Wearable and Implantable Body Sensor Networks (BSN) (pp. 13–18). https://doi.org/10.1109/BSN.2017.7935996

Documents

APJNT Journal · Areej Babiker, Ibrahima Faye, Aamir Saeed Malik 22 . Dr K. Jayasankara Reddy /Asia Pacific Journal of Neurotherapy (APJNT), 2020, Vol 2, ... number of studies with