Iran J Psychiatry 2023; 18: 2: 127-133
Original Article
Possible Neuropathological Mechanisms Underlying the Increased
Complexity of Brain Electrical Activity in Schizophrenia: A
Computational Study
Ali Khaleghi1*, Mohammad Reza Mohammadi1, Kian Shahi1, Ali Motie Nasrabadi2
Abstract
Objective: Schizophrenia is a complex neurodevelopmental illness that is associated with different deficits in the
cerebral cortex and neural networks, resulting in irregularity of brain waves. Various neuropathological hypotheses have
been proposed for this irregularity that we intend to examine in this computational study.
Method: We used a mathematical model of a neuronal population based on cellular automata to examine two
hypotheses about the neuropathology of schizophrenia: first, reducing neuronal stimulation thresholds to increase
neuronal excitability; and second, increasing the percentage of excitatory neurons and decreasing the percentage of
inhibitory neurons to increase the excitation to inhibition ratio in the neuronal population. Then, we compare the
complexity of the output signals produced by the model in both cases with real healthy resting-state
electroencephalogram (EEG) signals using the Lempel-Ziv complexity measure and see if these changes alter (increase
or decrease) the complexity of the neuronal population dynamics.
Results: By lowering the neuronal stimulation threshold (i.e., the first hypothesis), no significant change in the pattern
and amplitude of the network complexity was observed, and the model complexity was very similar to the complexity of
real EEG signals (P > 0.05). However, increasing the excitation to inhibition ratio (i.e., the second hypothesis) led to
significant changes in the complexity pattern of the designed network (P < 0.05). More interestingly, in this case, the
complexity of the output signals of the model increased significantly compared to real healthy EEGs (P = 0.002) and the
model output of the unchanged condition (P = 0.028) and the first hypothesis (P = 0.001).
Conclusion: Our computational model suggests that imbalances in the excitation to inhibition ratio in the neural network
are probably the source of abnormal neuronal firing patterns and thus the cause of increased complexity of brain
electrical activity in schizophrenia.
Key words: Computational Modelling; Neurophysiology; Neuropathology; Schizophrenia
1. Psychiatry and Psychology Research Center, Tehran University of Medical Sciences, Tehran, Iran.
2. Department of Biomedical Engineering, Shahed University, Tehran, Iran.
*Corresponding Author:
Address: South Kargar Avenue, Roozbeh Hospital, Psychiatry and Psychology Research Center, Tehran, Iran, Postal Code:
1333715914.
Tel: 98-21 55422002, Fax: 98-21 55421959, Email: [email protected]
Article Information:
Received Date: 2021/08/07, Revised Date: 2022/12/22, Accepted Date: 2023/01/03
Copyright © 2023 Tehran University of Medical Sciences. Published by Tehran University of Medical Sciences.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International license
(https://creativecommons.org/licenses/by-nc/4.0/). Noncommercial uses of the work are permitted, provided the original work is properly cited
Khaleghi, Mohammadi, Shahi, et al.
associated with abnormal firing patterns and increased
Schizophrenia is a severe and chronic psychiatric complexity of brain dynamics involved in schizophrenia.
In this computational study, we intended to use a
illness that disrupts patients’ life at different
neurobiological, cognitive, emotional and social levels. mathematical model of a neuronal population to
Although the onset of schizophrenia often occurs in investigate each of these impairments and their effect on
early adulthood, it has been shown that some traits and brain dynamics. This model, recently developed by our
symptoms, including cognitive and social impairments, research team (20), is based on cellular automata and
are already manifest in childhood (1, 2). Several studies uses simple laws of electrical events, called action
have pointed out genetic, neurobiological, potentials, and the inherent properties of neurons and the
environmental, social and psychological factors as main interactions between them, and is well able to simulate
contributory processes (3, 4). Meanwhile, many real EEG signals. Therefore, we use this model to
neuroscience studies have focused on neurobiological examine two hypotheses about the neuropathology of
and neurophysiological factors as abnormal underlying schizophrenia: first, reducing neuronal stimulation
processes. These studies suggest that changes in synapse thresholds to increase neuronal excitability; and second,
reorganization and abnormal patterns of neuronal firing increasing the excitation to inhibition ratio by increasing
are important neurobiological contributory factors, the percentage of excitatory neurons and decreasing the
reflecting abnormal neurophysiological activity and percentage of inhibitory neurons in the neuronal
irregular brain signals in schizophrenia (5-7). In this population. Then, we compare the complexity of the
regard, researchers in the field of computational output signals produced by the model in both cases with
neuroscience, considering the concept of time-dependent real healthy EEG signals and see if these changes alter
disorders and dynamical diseases, have utilized chaotic (increase or decrease) the complexity of the neuronal
approaches to study patterns of neural activity in population dynamics.
schizophrenia (8, 9). Electroencephalogram
(EEG)/Magnetoencephalogram (MEG) studies have Materials and Methods
used nonlinear and chaotic methods to estimate the To test the mentioned hypotheses, we used a model
complexity patterns of brain electrical activity and to recently developed and published by our research team
investigate the brain dynamics in schizophrenia due to that has been shown to be able to simulate recorded
(i) the nonlinear properties of dynamical neural systems, electrical signals from a healthy human brain. Details on
(ii) dynamical nature of schizophrenia symptoms and this model can be found in (20), but we will briefly
their severity, reflecting a disturbance in underlying describe it here. This model is based on cellular
nonlinear processes as state switches in the cortical automata and is designed using the Python programming
system, and (iii) deficits in stability, self-organization language and follows simple rules based on the intrinsic
and hierarchical processes of brain system in properties of neurons and the properties of action
schizophrenia (10-15). potentials. These rules are as follows:
Fernandez et al. (7), in a comprehensive review, (a) Define a 40 × 40 network,
concluded that young, drug-naïve patients with positive (b) Define the initial state for each neuron in the network
symptoms (i.e., paranoia and hallucinations) are randomly from the four states of resting, firing,
expected to show increased complexity in MEG and hyperpolarization and refractory, which are different
EEG signals, probably resulting from abnormal firing parts of an action potential,
patterns in neuronal activity in the critical brain areas. (c) Define the number of neighborhoods or synapses for
Neurons are supposed to be the computational parts of each neuron as the minimum number of synapses
brain function that release a neurotransmitter into the (Nmin) to the maximum number of synapses (N max),
synaptic space due to the action potentials. Any (d) If the sum of input synapses from the neighbor
disturbance in this neuronal environment can lead to neurons is greater than a resting threshold (T rest), then
abnormal firing patterns, resulting in impaired brain a resting cell in the network is activated,
activity and subsequent high-level behaviors (16). (e) After activation, the cell should go through firing,
Postmortem brain studies in schizophrenia have shown hyperpolarization and refractory states in order and
that there is an imbalance in the excitation to inhibition then go back to its resting state,
ratio due to impairments in the levels of glutamate, as an (f) In the firing and hyperpolarization states, a neuron
excitatory neurotransmitter, and N-Methyl-D-aspartate will never produce the next action potential,
receptor (NMDA-R) signaling, resulting in the raised (g) In the refractory state, the neuron may be activated if
excitability of pyramidal neurons (17, 18). On the other the sum of input synapses from its neighbors is
hand, a variety of impairments in the inhibitory greater than a relative threshold (Trelative), where
neurotransmitter Gamma-Aminobutyric Acid (GABA) Trelative > Trest,
have been observed in postmortem schizophrenia brain (h) The output time series is the sum of the effects of
researches, which has demonstrated that schizophrenia is each neuron in the network.
closely related to decreased inhibitory neurotransmission Based on the results of the previous study, we define
(19). Each of these deficits and impairments can be initial states for neurons as Nfiring = 40%, Nrefractory =
128 Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir
Neuropathological Mechanisms in Schizophreni
35%, Nrest = 20% and Nhyperpolarization = 5%. Furthermore, time series is determined according to the rate of
20 percent of the network neurons were considered as expansion of the differences between successive
inhibitory neurons and the remaining 80 percent as samples. Larger MLE values demonstrate more complex
excitatory neurons. Also, we set Nmin = 14, Nmax = 60, patterns in a dynamical system (23, 24). Moreover, the
Trest = 8 and Trelative = 12. embedding dimension is an important parameter for
As described, some important features of a neuronal reconstructing the state space of dynamic systems. In
network are embedded in this cellular automata model. short, in order to reconstruct the state space, the state
Our published results on the model proved that this vectors are replaced by delay vectors, and the number of
model can produce various dynamics, from limit-cycle components in these vectors is the embedding
to chaotic dynamics, just like the different dynamics and dimension. In this work, we used the well-known Cao
behaviors seen from a real neuron population in the algorithm to estimate the embedding dimension (25).
human brain at different conditions. Moreover, output Furthermore, Lempel-Ziv complexity was calculated to
time series from the model showed a high-dimensional quantitatively measure the complexity of the healthy
chaotic behavior similar to a healthy brain. Quantitative EEG signal and the model output time series. It uses the
and qualitative comparisons of the model output time concepts of producibility, reproducibility and exhaustive
series and the real EEG data in both nonlinear and linear history of a sequence to estimate the complexity of a
domains proved the ability of this mathematical model to time series. For more information and details on these
simulate the real EEG signals. Therefore, this methods for estimating the complexity of EEG signals,
computational model is useful and can be utilized to you can refer to (7, 22, 23).
investigate healthy and pathological neural populations Statistical analysis
and gain insight into their various functional First, the Shapiro-Wilk test was used to determine the
mechanisms. As a result, we used this model to test data distribution. This test showed that the data has a
existing hypotheses about the neuropathology of normal distribution (P > 0.05). Therefore, parametric
schizophrenia. For the first hypothesis, we reduced T rest tests should be used to determine statistically significant
and Trelative to increase the excitability of neurons; and differences between features and indices extracted from
for the second one, we changed the percentage of real and simulated signals in different conditions. As a
excitatory and inhibitory neurons in the network. result, in this study, analysis of variance (ANOVA) and
Real EEG data independent t tests were used as parametric statistical
In this work, EEG data available on the PhysioNet tests to compare means. All statistical analyzes were
website was used to qualitatively and quantitatively performed using the SPSS software (version 21) and P <
compare simulated signals in different conditions with 0.05 was considered as a significant criterion.
real signals. This database contains 72 EEG signals from
healthy adults at rest and during cognitive tasks. EEG Results
signals were recorded using the Neurocom EEG device We, first, qualitatively examined and compared the
based on the 10-20 international standard protocol and complexity of real healthy EEG signals and the model
through Ag/AgCl electrodes. Earlobes were used as output time series (simulated EEG) for the mentioned
references in signal recording. In the pre-processing parameters using MLE and the embedding dimension.
stage of these signals, a notch filter (50 Hz) and a low- As shown in Figure 1 (top row), the complexity of the
pass filter with a cutoff frequency of 30 Hz were used. In signals generated by the model is very similar in pattern
addition, independent component analysis was used to and amplitude to the complexity of real healthy EEG
reduce signal artifacts. Therefore, a clean 60-second signals. In the Cao algorithm, two functions are used to
EEG signal fragment was provided for each individual in estimate the embedding dimension of a time series: E 1(d)
this database (21). Here, the O1 channel of 20 resting- and E2(d). When d is greater than or equal to the
state signals from this database was used for further embedding dimension (i.e., close to 1), E1(d) does not
comparison purposes . change. On the other hand, E 2(d) is used to distinguish
Complexity measures deterministic signals from chaotic signals. For
There are various techniques from multiple conceptual deterministic signals, E2(d) is equal to 1 for some
approaches to measure complexity (22). In this work, we dimension, while for chaotic signals, E2(d) is equal to 1
used maximum lyapunov exponent (MLE) and the for almost all values of d. Therefore, as proved in our
embedding dimension to compare nonlinear previous article, this model is well able to simulate the
characteristics and complexity patterns of the real and dynamics of the human brain's electrical activity and real
simulated EEGs qualitatively. Lyapunov exponents are EEG signals.
considered as dynamical measures of attractor
complexity. They reveal the exponential convergence or
divergence of adjacent trajectories of the attractor
present in the phase space. MLE is interpreted as a
measure of dynamic complexity because it indicates
dependence on initial conditions. A dynamical system or
Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir 129
Khaleghi, Mohammadi, Shahi, et al.
Figure 1. Maximum Lyapunov Exponent (Top Row) and Embedding Dimension (Below Row) Diagrams
Calculated for a Real EEG Signal (Blue) and a Simulated EEG Signal Generated by the Cellular
Automata Model (Green)
Next, to test the hypotheses under consideration, we as the unchanged condition) and extracted their
made changes to the value of the model parameters and complexity feature using the Lempel-Ziv measure. Table
again compared the complexity of the time series 1 shows the complexity values of the healthy EEG
generated by the model in different conditions with the signals and model outputs with and without applied
complexity of the real EEG signals. To test the first changes (unchanged condition, first hypothesis and
hypothesis, we adjusted the T rest = 6 and Trelative = 10 second hypothesis). As you can see, by lowering the
values to lower the excitation threshold of the network neuronal stimulation threshold (i.e., the first hypothesis),
neurons and increase the excitability of the neurons. there was no significant change in the pattern and
Experiments showed that lower values of these amplitude of the network complexity, and the model
parameters lead to loss of network dynamics. To test the complexity is very similar to the complexity of a healthy
second hypothesis, we increased the percentage of EEG signal (P > 0.05). However, increasing the
excitatory neurons to 85% and reduced the percentage of excitation to inhibition ratio (i.e., the second hypothesis)
inhibitory neurons to 15% to increase the excitation to led to significant changes in the complexity pattern of
inhibition ratio in the neural network. These percentages the designed network (P < 0.05). More interestingly, in
were chosen considering maintaining optimal network this case, the complexity of the output signals of the
dynamics to generate simulated EEG signals after model significantly increased compared to real healthy
several trials. Finally, we analyzed 20 real healthy EEGs (P = 0.002) and the model output of the
EEGs, 20 simulated EEGs for the first hypothesis, 20 unchanged condition (P = 0.028) and the first hypothesis
simulated EEGs for the second hypothesis, and 20 (P = 0.001).
baseline simulated EEGs (with initial parameter values
130 Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir
Neuropathological Mechanisms in Schizophreni
Table 1. Comparison of the Complexity of Real Healthy EEG Signals, the Simulated Signals Produced
by the Cellular Automata Model with Initial Parameter Values (Unchanged Condition), with Decreased
Neuronal Stimulation Thresholds (First Hypothesis) and with Increased Excitation to Inhibition Ratio
(Second Hypothesis) Using the Lempel-Ziv Measure
95% Confidence Interval
(I) Label (J) Label Mean Difference (I-J) Std. Error Sig.
Lower Bound Upper Bound
First Hypothesis 0.00385 0.01354 1.000 -0.0296 0.0372
Real Healthy EEG Second Hypothesis -0.04794* 0.01354 0.002 -0.0813 -0.0145
Unchanged Condition 0.00309 0.01745 1.000 -0.0443 0.0505
Healthy EEG -0.00385 0.01354 1.000 -0.0372 0.0296
First Hypothesis Second Hypothesis -0.05178* 0.01354 0.001 -0.0852 -0.0184
Unchanged Condition -0.01751 0.01745 1.000 -0.0649 0.0299
Healthy EEG 0.04794* 0.01354 0.002 0.0145 0.0813
Second
First Hypothesis 0.05178* 0.01354 0.001 0.0184 0.0852
Hypothesis
Unchanged Condition 0.05103* 0.01745 0.028 0.0036 0.0984
Healthy EEG -0.00309 0.01745 1.000 -0.0505 0.0443
Unchanged
First Hypothesis 0.01751 0.01745 1.000 -0.0299 0.0649
Condition
Second Hypothesis -0.05103* 0.01745 0.028 -0.0984 -0.0036
Discussion previous studies have shown that schizophrenia is
In this work, we used a mathematical and computational associated with impairments in GABAergic markers (31,
model to investigate the neuropathological mechanisms 32). One of the main functions of GABA is the
in schizophrenia. Such a study lies within the fields of synchronization of neuronal populations, which
computational neuroscience and computational produces bursting or rhythmic neural activity, called
psychiatry and has been applied by many researchers in brain oscillations. Appropriate oscillating firing activity
recent years to investigate the relationship between in different brain regions, especially in the prefrontal
different pathological and phenomenological aspects of cortex, is thought to be a neural mechanism of working
psychiatric disorders (26-30). Our model, based on the memory (33, 34). Given that deficits in working memory
simple rules of cellular automata, was able to simulate have been suggested to underpin a variety of cognitive
the dynamics of a real neuronal population of the human impairments in patients with schizophrenia, disturbance
brain. Our experiments and the results of this model of brain oscillations can be involved in schizophrenia
showed that the reduction of neuronal stimulation (35, 36). A reduction in the expression of the GAD67
thresholds does not lead to a significant change in the (GABA synthesizing enzyme) and an impairment in the
dynamics of the neuronal population and the complexity parvalbumin-positive GABAergic interneurons (PVIs)
of neuronal function. Thus, previous evidence of deficits within the prefrontal cortex are the strongest findings in
in glutamate levels and NMDA-R signaling associated postmortem schizophrenia studies (37, 38). Given that
with internalization of receptors involved in synaptic and the expression of PVIs and GAD67 is dependent on the
extra-synaptic environments in both inhibitory and activity, GABAergic drive is decreased within the
excitatory neurons may not possibly be the prefrontal cortex in schizophrenia (39).
neuropathological mechanism underlying the dynamic
change in brain neurons and the increased complexity of Limitation
electrophysiology in schizophrenia (6). In addition, the Like many other researches and studies, this study also
experiments and the results of the model showed that suffers from some limitations. The results obtained in
increasing the ratio of excitation to inhibition in a this study are based on a mathematical model and
neuronal population can cause substantial changes in its therefore their accuracy should be checked by animal
behavioral and dynamical pattern and lead to increased models and studies. In addition, in this study, we
complexity of neuronal function in the schizophrenic examined only two neuropathological mechanisms
brain. Therefore, decreased inhibitory neurotransmission reported in schizophrenia, while other neuropathological
may be the neuropathological mechanism underlying the mechanisms such as the disconnection hypothesis of
dynamic change in brain neurons and, thus, the increased schizophrenia have been emphasized in many previous
complexity of cortical electrophysiology in studies, which cannot be examined using the current
schizophrenia. Inhibitory neurotransmitter GABA is an model.
important molecule in modulating neuronal firing, and
Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir 131
Khaleghi, Mohammadi, Shahi, et al.
Conclusion synaptic optogenetics in an animal model. Proc
Various neuropathological mechanisms may lead to Jpn Acad Ser B Phys Biol Sci. 2019;95(5):179-
97.
distortion of neuronal activity and brain function, and
7. Fernández A, Gómez C, Hornero R, López-Ibor
consequently abnormal high-level behaviors associated JJ. Complexity and schizophrenia. Prog
with the disorder in diseases such as schizophrenia. Neuropsychopharmacol Biol Psychiatry.
These mechanisms may alter the dynamics of neuronal 2013;45:267-76.
function in brain networks by affecting the regular firing 8. Ibanez-Molina AJ, Lozano V, Soriano MF,
pattern of different neurons and may be the source of Aznarte JI, Gomez-Ariza CJ, Bajo MT. EEG
abnormalities reported in schizophrenia. Our Multiscale Complexity in Schizophrenia During
computational model suggests that imbalances in the Picture Naming. Front Physiol. 2018;9:1213.
excitation to inhibition ratio in the neural network are 9. Yang AC, Hong CJ, Liou YJ, Huang KL, Huang
CC, Liu ME, et al. Decreased resting-state brain
probably the source of abnormal neuronal firing patterns
activity complexity in schizophrenia
and thus the increased complexity of brain electrical characterized by both increased regularity and
activity in schizophrenia. In addition, this study showed randomness. Hum Brain Mapp.
that such a model can be used in the future to better 2015;36(6):2174-86.
understand the neuropathological mechanisms involved 10. Breakspear M. The nonlinear theory of
in other psychiatric and neurological disorders. schizophrenia. Aust N Z J Psychiatry.
However, findings of the present study should be 2006;40(1):20-35.
interpreted with caution due to the nature of 11. Akar SA, Kara S, Latifoglu F, Bilgic V. Analysis
computational studies. Indeed, future in vivo and in vitro of the Complexity Measures in the EEG of
Schizophrenia Patients. Int J Neural Syst.
studies should confirm our theoretical findings.
2016;26(2):1650008.
12. Kutepov IE, Dobriyan VV, Zhigalov MV,
Acknowledgment Stepanov MF, Krysko AV, Yakovleva TV, et al.
This study was supported by Tehran University of EEG analysis in patients with schizophrenia
Medical Sciences through a grant from the Psychiatry based on Lyapunov exponents. Inform Med
Unlocked. 2020;18:100289.
and Psychology Research Center [grant number 37425].
13. Namazi H, Aghasian E, Ala TS. Fractal-based
classification of electroencephalography (EEG)
Conflict of Interest signals in healthy adolescents and adolescents
None. with symptoms of schizophrenia. Technol Health
Care. 2019;27(3):233-41.
14. Bai D, Yao W, Lv Z, Yan W, Wang J. Multiscale
multidimensional recurrence quantitative
analysis for analysing MEG signals in patients
with schizophrenia. Biomed Signal Process
References Control. 2021;68:102586.
1. Alavi SS, Mohammadi MR, Hooshyari Z, 15. Rojas DC. Review of schizophrenia research
Mohammadi Kalhori S, Salehi M, Salmanian M, using MEG. Magnetoencephalography: From
et al. Epidemiology of Psychotic Disorders Signals to Dynamic Cortical Networks.
Based on Demographic Variables in Iranian 2019:1121-46.
Children and Adolescents. Iran J Psychiatry. 16. Bassett DS, Sporns O. Network neuroscience.
2021;16(1):1-12. Nat Neurosci. 2017;20(3):353-64.
2. Mohammadi MR, Ahmadi N, Khaleghi A, 17. Aine CJ, Bockholt HJ, Bustillo JR, Cañive JM,
Mostafavi SA, Kamali K, Rahgozar M, et al. Caprihan A, Gasparovic C, et al. Multimodal
Prevalence and Correlates of Psychiatric Neuroimaging in Schizophrenia: Description and
Disorders in a National Survey of Iranian Dissemination. Neuroinformatics.
Children and Adolescents. Iran J Psychiatry. 2017;15(4):343-64.
2019;14(1):1-15. 18. Čiháková D, Eaton WW, Talor MV, Harkus UH,
3. Janoutova J, Janackova P, Sery O, Zeman T, Demyanovich H, Rodriguez K, et al. Gut
Ambroz P, Kovalova M, et al. Epidemiology and permeability and mimicry of the Glutamate
risk factors of schizophrenia. Neuro Endocrinol Ionotropic Receptor NMDA type Subunit
Lett. 2016;37(1):1-8. Associated with protein 1 (GRINA) as potential
4. Modai S, Shomron N. Molecular Risk Factors for mechanisms related to a subgroup of people
Schizophrenia. Trends Mol Med. with schizophrenia with elevated antigliadin
2016;22(3):242-53. antibodies (AGA IgG). Schizophr Res.
5. Duchatel RJ, Shannon Weickert C, Tooney PA. 2019;208:414-9.
White matter neuron biology and 19. Egerton A, Modinos G, Ferrera D, McGuire P.
neuropathology in schizophrenia. NPJ Neuroimaging studies of GABA in
Schizophr. 2019;5(1):10. schizophrenia: a systematic review with meta-
6. Obi-Nagata K, Temma Y, Hayashi-Takagi A. analysis. Transl Psychiatry. 2017;7(6):e1147.
Synaptic functions and their disruption in 20. Khaleghi A, Mohammadi MR, Shahi K, Motie
schizophrenia: From clinical evidence to Nasrabadi A. A neuronal population model
132 Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir
Neuropathological Mechanisms in Schizophreni
based on cellular automata to simulate the 35. Chatterjee I, Kumar V, Sharma S, Dhingra D,
electrical waves of the brain. Waves Random Rana B, Agarwal M, et al. Identification of brain
Complex Media. 2021:1-20. regions associated with working memory deficit
21. Zyma I, Tukaev S, Seleznov I, Kiyono K, Popov in schizophrenia. F1000Res. 2019;8:124.
A, Chernykh M, et al. Electroencephalograms 36. Wu G, Palaniyappan L, Zhang M, Yang J, Xi C,
during mental arithmetic task performance. Liu Z, et al. Imbalance Between Prefronto-
Data. 2019;4(1):14. Thalamic and Sensorimotor-Thalamic Circuitries
22. Zarafshan H, Khaleghi A, Mohammadi MR, Associated with Working Memory Deficit in
Moeini M, Malmir N. Electroencephalogram Schizophrenia. Schizophr Bull. 2022;48(1):251-
complexity analysis in children with attention- 61.
deficit/hyperactivity disorder during a visual 37. Lewis DA, Curley AA, Glausier JR, Volk DW.
cognitive task. J Clin Exp Neuropsychol. Cortical parvalbumin interneurons and cognitive
2016;38(3):361-9. dysfunction in schizophrenia. Trends Neurosci.
23. Mohammadi MR, Khaleghi A, Nasrabadi AM, 2012;35(1):57-67.
Rafieivand S, Begol M, Zarafshan H. EEG 38. Murray AJ, Woloszynowska-Fraser MU, Ansel-
classification of ADHD and normal children Bollepalli L, Cole KL, Foggetti A, Crouch B, et
using non-linear features and neural network. al. Parvalbumin-positive interneurons of the
Biomed Eng Lett. 2016;6(2):66-73. prefrontal cortex support working memory and
24. Barreira L. Lyapunov exponents: Springer; cognitive flexibility. Sci Rep. 2015;5:16778.
2017. 39. Chung DW, Fish KN, Lewis DA. Pathological
25. Cao L. Practical method for determining the Basis for Deficient Excitatory Drive to Cortical
minimum embedding dimension of a scalar time Parvalbumin Interneurons in Schizophrenia. Am
series. Physica D: Nonlinear Phenomena. J Psychiatry. 2016;173(11):1131-9.
1997;110(1-2):43-50.
26. Huys QJ, Maia TV, Frank MJ. Computational
psychiatry as a bridge from neuroscience to
clinical applications. Nat Neurosci.
2016;19(3):404-13.
27. Khaleghi A, Mohammadi MR, Shahi K,
Nasrabadi AM. Computational Neuroscience
Approach to Psychiatry: A Review on Theory-
driven Approaches. Clin Psychopharmacol
Neurosci. 2022;20(1):26-36.
28. Khaleghi A, Sheikhani A, Mohammadi MR,
Nasrabadi AM, Vand SR, Zarafshan H, et al.
EEG classification of adolescents with type I
and type II of bipolar disorder. Australas Phys
Eng Sci Med. 2015;38(4):551-9.
29. Moeini M, Khaleghi A, Mohammadi MR.
Characteristics of Alpha Band Frequency in
Adolescents with Bipolar II Disorder: A Resting-
State QEEG Study. Iran J Psychiatry.
2015;10(1):8-12.
30. Khaleghi A, Zarafshan H, Mohammadi MR.
Visual and auditory steady-state responses in
attention-deficit/hyperactivity disorder. Eur Arch
Psychiatry Clin Neurosci. 2019;269(6):645-55.
31. Torrey EF, Barci BM, Webster MJ, Bartko JJ,
Meador-Woodruff JH, Knable MB.
Neurochemical markers for schizophrenia,
bipolar disorder, and major depression in
postmortem brains. Biol Psychiatry.
2005;57(3):252-60.
32. Lewis DA, Hashimoto T, Volk DW. Cortical
inhibitory neurons and schizophrenia. Nat Rev
Neurosci. 2005;6(4):312-24.
33. Bartos M, Vida I, Jonas P. Synaptic
mechanisms of synchronized gamma
oscillations in inhibitory interneuron networks.
Nat Rev Neurosci. 2007;8(1):45-56.
34. Fries P. Neuronal gamma-band synchronization
as a fundamental process in cortical
computation. Annu Rev Neurosci. 2009;32:209-
24.
Iranian J Psychiatry 18: 2, April 2023 ijps.tums.ac.ir 133
Copyright of Iranian Journal of Psychiatry is the property of Tehran University of Medical
Sciences and its content may not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.
