ADHD is the foremost behaviorally- and neurologically-based syndrome in the world. There are no established etiologies, and medication and coaching are the leading first-line treatments, with arguable success. The pursuit for better approaches to reduction and/or remediation of the debilitating symptoms is essential. Dysregulations in the electroencephalogram (EEG) may hold the key towards future treatment options.
The brain is the most complicated system known. It holds many of the mysteries and beauties of all living beings. When it functions well, it regulates human behavior by integrating a multitude of specialized interconnections, known as networks, between brain structures and areas, all leading to successfully completing a task.
Specific dysregulations in the EEG of ADHD children are becoming more evident. A high theta/beta ratio (TBR), dysregulated alpha, beta issues, spindling, and excessive delta are surface power issues linked to specific ADHD presentations. Beyond those, findings from many studies (Fassbender et al., 2009; Menon, 2011; Menon & Uddin, 2010; Uddin, 2015; Vossel et al., 2014; Wojciulik & Kanwisher, 1999) indicate the default mode network (DMN), central executive network (CEN), salience (SalN), attention dorsal (AttDN), and the frontal-parietal network (FPN) to be the most dysregulated networks in this population (Table 1).
Table 1: General consensus of network and Brodmann area dysregulations in ADHD
Network, Brain region, Brodmann areas, 10 20 site(s), and functions of five most referenced dysregulated networks in ADHD. DMN=default mode network, CEN=central executive network, SalN=salience network, AttDN=attention dorsal network, FPN=fronto-parietal network.
ICAN Study Analyses of ADHD Cohort
The cohort used for these analyses is from the ICAN study (Neurofeedback Collaborative Group, 2023), which used theta/beta ratio (TBR) neurofeedback training with children aged 7- to 11-years old who were rigorously diagnosed with inattention or combined-type (impulsivity and inattention) ADHD. They underwent 38 neurofeedback training sessions and three EEG assessments (baseline, post treatment, and 13-week follow-up). The baseline EEG was used in this study, and future studies should examine whether baseline EEG features could predict level of success in this treatment modality, and attention improvement based upon the Conners Parent and Teacher Rating Scales using machine learning analyses.
The EEG contains a vast number of datapoints per second, which makes it very unwieldy to study. Since we looked at coherence and phase in five networks, all Brodmann Area connections associated with them, and in each of the main frequency bands (delta, theta, alpha1, alpha2, beta1, beta2, and high beta) each datapoint had at least four covariates and hundreds of possibilities. Using the machine learning model known as Monte Carlo, we compared these children’s baseline EEG data to the NeuroGuide age-matched normative database. We used SynapseData.org to investigate connectivity dysregulations (Kerson et al., 2023).
Monte Carlo is a machine learning algorithm that generates random samples and performs deterministic computations to aggregate non-random estimations of complex data. Simply put, it grouped the same data from the large array of information and tested it multiple times to ascertain confidence in its results.
In our 2023 study findings, we discovered that children with ADHD differed substantially in coherence efficiency in the DMN, SalN, and AttDN, specifically in hypocoherence (reduced connectivity between brain regions). This makes sense considering the ADHD presentation of disorganization. By extension, Brodmann areas (BAs) 7, 10, and 11 within these networks were the most dysregulated. BA 10 and 7 are represented in Table 1 marking the BAs within the networks noted above. Brodmann graphic © sciencepics/Shutterstock.com.
Our team also found BA 11 to be highly dysregulated. This area is found in the orbital prefrontal cortex (OFC), part of the ventromedial prefrontal cortex (vmPFC), and is closely tied to the DMN , SalN, and BA 10 in location and function. Seven-network graphic by Feirrera et al. (2022).
Caption: Brain networks: One atlas with seven networks. Seven brain networks derived from resting‐state fMRI data were adapted from Schaefer et al. (2018).
In addition to the three primary dysregulated BAs, some were secondary as they were often paired with the primary ones within connectivity metrics. These included BAs 21 (middle temporal gyrus; language, face/object recognition), 30 (part of the cingulate; space objectivity, planning), 35 (perirhinal cortex; cognition, identifying objects in space), 37 (fusiform gyrus; face and object recognition), 39 (angular gyrus; reading, verbal reasoning [Wernicke’s area], and 40 (supramarginal gyrus; executive function, language, visuo-motor planning).
Conclusion
The findings from this study contribute to the growing body of research supporting the neurobiological basis of ADHD, particularly in large-scale network dysregulations. Identifying hypocoherence in the DMN, SalN, and AttDN, alongside Brodmann areas 7, 10, and 11, underscores the structural-functional abnormalities that may drive ADHD symptomatology. The secondary dysregulated areas, such as the middle temporal gyrus, cingulate cortex, and fusiform gyrus, further refine our understanding of this population's cognitive and attentional disruptions. Importantly, these connectivity patterns may serve as neurophysiological biomarkers, providing clinicians with more targeted diagnostic tools and informing personalized neurofeedback or other intervention strategies.
The use of machine learning approaches, such as Monte Carlo modeling, represents a significant step forward in handling the vast complexity of EEG data, offering predictive insights into treatment outcomes. Future research should explore these network dysregulations' longitudinal stability and responsiveness to intervention, enhancing both assessment and remediation approaches for ADHD.
Key Takeaways
Network dysregulation in ADHD: ADHD is associated with hypocoherence in major neural networks, particularly the DMN, SalN, and AttDN, which impacts cognitive control and attention regulation.
Brodmann area abnormalities: Specific regions, such as BAs 7, 10, and 11, exhibit significant dysregulation, suggesting a structural-functional basis for ADHD symptomatology.
Secondary dysregulated regions: Additional Brodmann areas linked to language, object recognition, and executive function (e.g., BAs 21, 30, 35, 37, 39, 40) also show abnormal connectivity, influencing ADHD-related cognitive deficits.
Machine learning in EEG analysis: The Monte Carlo model enables advanced pattern recognition within large EEG datasets, improving biomarker identification and potentially predicting treatment outcomes.
Clinical implications: These findings support the development of neurophysiological biomarkers for ADHD assessment and could inform personalized interventions, such as neurofeedback, based on specific connectivity profiles.
Guest Author
We are delighted to publish this essay by Dr. Cynthia Kerson, 2 PhD, QEEGD, BCN, BCB Senior Fellow, BCB-HRV. Dr. Kerson is an Associate Professor, Saybrook University, Department of Applied Psychophysiology.
Glossary
ADHD (Attention-Deficit/Hyperactivity Disorder): a neurodevelopmental disorder characterized by inattention, hyperactivity, and impulsivity, linked to dysregulations in brain connectivity.
attention dorsal network (AttDN): a neural network involved in directing and sustaining attention, found to be dysregulated in ADHD.
Brodmann area (BA): A classification of cortical regions based on cytoarchitectonic differences, with specific areas linked to cognitive and sensory functions.
central executive network (CEN): a large-scale network involved in higher-order cognitive control and working memory, often impaired in ADHD.
coherence: a measure of functional connectivity between EEG signals from different brain regions, with hypocoherence indicating reduced synchronization.
default mode network (DMN): Aa network associated with self-referential thought and mind-wandering, often dysregulated in ADHD.
electroencephalogram (EEG): a method for recording brain electrical activity, used to assess connectivity and neural dysregulations.
fusiform gyrus: a brain region involved in object and face recognition, showing abnormal connectivity in ADHD.
frontal-parietal network (FPN): a network involved in executive function and attentional control, often implicated in ADHD.
hypocoherence: reduced connectivity between brain regions, indicative of functional network inefficiencies in ADHD.
machine learning: a computational approach using algorithms to identify patterns in large datasets, as applied to EEG analysis in this study.
Monte Carlo model: a statistical method that generates random samples to estimate the properties of complex data, utilized for EEG connectivity analysis.
normative database: a reference dataset of EEG recordings used to compare individual brain activity against age-matched norms.
orbital prefrontal cortex (OFC): a region within the ventromedial prefrontal cortex associated with decision-making and emotional regulation, found to be dysregulated in ADHD.
salience network (SalN): a network responsible for detecting and responding to relevant stimuli, often disrupted in ADHD.
theta/beta ratio (TBR): a common EEG marker in ADHD, with elevated theta power relative to beta indicating attentional deficits.
ventromedial prefrontal cortex (vmPFC): a prefrontal region linked to emotion regulation and decision-making, implicated in ADHD-related dysfunction.
References
Fassbender, C., Schweitzer, J. B., & Murphy, K. (2009). The role of the orbitofrontal cortex in ADHD. Journal of Psychiatric Research, 43(7), 1107–1113. https://doi.org/10.1016/j.jpsychires.2009.03.012
Kerson, C., Lubar, J., deBeus, R., Pan, X., Rice, R., Allen, T., Yazbeck, M., Sah, S., Dhawan, Y., Zong, W., Roley-Roberts, M. E., & Arnold, L. E. (2023). EEG connectivity in ADHD compared to a normative database: A cohort analysis of 120 subjects from the ICAN study. Applied Psychophysiology and Biofeedback, 48(2), 191–206. https://doi.org/10.1007/s10484-022-09569-9
Menon, V. (2011). Large-scale brain networks and psychopathology: A unifying triple network model. Trends in Cognitive Sciences, 15(10), 483–506. https://doi.org/10.1016/j.tics.2011.08.003
Menon, V., & Uddin, L. Q. (2010). Saliency, switching, attention, and control: A network model of insula function. Brain Structure and Function, 214(5–6), 655–667. https://doi.org/10.1007/s00429-010-0262-0
Neurofeedback Collaborative Group. (2023). Neurofeedback for Attention-Deficit/Hyperactivity Disorder: 25-month follow-up of double-blind randomized controlled trial. Journal of the American Academy of Child and Adolescent Psychiatry, 62(4), 435–446. https://doi.org/10.1016/j.jaac.2022.07.862
Uddin, L. Q. (2015). Salience network dysfunction in psychiatric disorders: A neuroimaging review. Frontiers in Neuroscience, 9, 239. https://doi.org/10.3389/fnins.2015.00239
Vossel, S., Geng, J. J., & Fink, G. R. (2014). The role of the right parietal cortex in attention shifts: Evidence from neuroimaging and neuropsychology. Neuropsychologia, 58, 50–58. https://doi.org/10.1016/j.neuropsychologia.2014.03.006
Wojciulik, E., & Kanwisher, N. (1999). The generality of parietal involvement in visual attention. Neuron, 23(4), 747–764. https://doi.org/10.1016/s0896-6273(01)80033-7