Evaluating Predictive EEG Theta/Beta Features in Adult ADHD via Machine Learning

Overview

Attention-Deficit/Hyperactivity Disorder (ADHD) is characterized by differences in attention and cognitive control, yet assessment remains largely dependent on subjective behavioral evaluations. While the Theta/Beta Ratio (TBR) has been validated as an objective Electroencephalogram (EEG) biomarker in pediatric ADHD, its relevance in adults remains unclear. This study hypothesized that adults with ADHD would show elevated relative TBR and theta power and reduced beta power across brain regions, with regional features supporting machine learning classification. EEG recordings from 51 adults aged 18–59 (26 medication-naive ADHD, 25 controls) during a Go/No-Go task were obtained from a publicly available dataset. Following rigorous preprocessing and artifact correction, regional features were extracted across five cortical areas (Frontal, Parietal, Occipital, Temporal, and Central) and evaluated using parametric tests. Cross-validated Logistic Regression models examined whether these patterns generalize to individual-level distinctions. Results revealed significant group differences with moderate effect sizes, most prominently elevated temporal TBR, alongside increased frontal theta and reduced parietal beta. Significant temporal activity across features suggests a stronger role of auditory and language-related processing in adult ADHD than traditionally emphasized. Machine learning classification performance was modest (ROC-AUC ≈ 0.55), reflecting substantial individual variability. Overall, the hypothesis was partially supported. While EEG features capture meaningful group-level distinctions in attention-related neural activity, they are insufficient for reliable individual classification. This highlights the complexity of cognitive processes in adult ADHD and the novel insight that Temporal neural dynamics may play a larger role in attention-related variability. Future research should incorporate larger datasets and subtype-specific analysis to further clarify the cognitive significance of EEG-derived neural patterns in adult ADHD.

Competition history

• CSEF 2026 Cognitive Science (Junior Division) · Entry J-06-03

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Behavioural and Social Sciences projects
• CSEF Behavioural and Social Sciences projects
• Every CSEF 2026 project