---
title: "Enhancing ADHD Risk Screening with Socioeconomic Context and Multimodal EEG Analysis"
id: "plos-one-23-integrating-socioeconomic-context-with-multimodal-eeg-data-for-improved-adhd"
canonical_url: "https://medichelpline.com/clinical-feed/plos-one-23-integrating-socioeconomic-context-with-multimodal-eeg-data-for-improved-adhd"
content_type: "clinical_feed_article"
specialty: "Neurology"
source_name: "PLOS ONE (Medicine)"
source_url: "https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0357213"
published_at: "2026-09-01T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Enhancing ADHD Risk Screening with Socioeconomic Context and Multimodal EEG Analysis
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-one-23-integrating-socioeconomic-context-with-multimodal-eeg-data-for-improved-adhd
- **Specialty:** [Neurology](https://medichelpline.com/clinical-feed/neurology.md)
- **Primary Source:** PLOS ONE (Medicine)
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0357213)
- **Published At:** 2026-09-01T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- ADHD affects millions, yet diagnosis primarily relies on subjective behavioral assessments. - Current methods for ADHD risk screening using EEG neglect important socioeconomic factors affecting neurodevelopment. - A new multimodal deep learning model integrates temporal EEG dynamics, spatial information, and socioeconomic data to enhance ADHD classification accuracy. - The model showed a notable increase in accuracy (2.1–5.9%) and sensitivity (up to 12.2%) over EEG-only baselines. - Evaluation across four cognitive tasks demonstrated the integrated approach's consistent effectiveness, achieving accuracy up to 0.96, emphasizing the necessity of incorporating socioeconomic elements. - Results highlight the potential of context-aware ADHD risk screening while indicating limitations as the model is not a diagnostic tool.
## Clinical Analysis & Structured Key Points
Integrating socioeconomic context with multimodal EEG data for improved ADHD risk screening | PLOS One Browse Subject Areas ? Click through the PLOS taxonomy to find articles in your field. For more information about PLOS Subject Areas, click here . Article Authors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract Attention-deficit/hyperactivity disorder (ADHD) affects millions globally, yet current diagnostic approaches rely on subjective behavioral assessments without objective neurophysiological markers. While machine learning on electroencephalogram (EEG) data shows promise for automated ADHD risk screening, current methods focus only on brain signals and ignore socioeconomic factors that strongly affect neurodevelopment and ADHD risk. We introduce a novel multimodal deep learning architecture integrating three complementary streams: temporal EEG dynamics via one-dimensional convolutional-recurrent networks, spectro-temporal patterns via two-dimensional convolutional networks with spatial and channel attention, and socioeconomic context via feedforward processing, combined through an attention-based fusion mechanism. Using the Cognitive Electrophysiology in Socioeconomic Context dataset, we evaluate performance across four cognitive tasks with 5-fold stratified cross-validation, ablation studies and benchmarking against a state-of-the-art EEG classification model. Under epoch-level cross-validation, the multimodal approach outperforms EEG-only baselines across all four tasks, achieving accuracy improvements of 2.1–5.9% and sensitivity gains up to 12.2%, with strong positive-class F1-scores (96.4–99.8%). Results showed higher epoch-level performance when socioeconomic context was incorporated alongside neurophysiological signals, a pattern that held across diverse cognitive paradigms. Leave-One-Subject-Out Cross-Validation across all four tasks yielded accuracy of 0.86–0.91 for the EEG-only model and 0.92–0.96 for the multimodal model, with sensitivity of 0.71–0.96 and specificity of 0.95–1.00 for the multimodal model. These subject-independent estimates are more modest than the epoch-level figures and McNemar’s test on paired predictions did not reach significance on any task. The EEG backbone, evaluated without modification on an independent paediatric dataset, also achieved 80.4% subject-independent accuracy, outperforming the prior benchmark. Labels derive from a validated self-report screening instrument rather than clinical diagnosis; this model should be understood as a proof-of-concept for ADHD risk screening, not a diagnostic tool. This work suggests the feasibility of context-aware ADHD risk screening that accounts for environmental influences on neurodevelopment alongside neurophysiological signals. Citation: Sadi SH, Hossain MA, Tawhid MNA (2026) Integrating socioeconomic context with multimodal EEG data for improved ADHD risk screening. PLoS One 21(9): e0357213. https://doi.org/10.1371/journal.pone.0357213 Editor: Assoc. Prof. Phakkharawat Sittiprapaporn, Mae Fah Luang University School of Anti Aging and Regenerative Medicine, THAILAND Received: March 23, 2026; Accepted: August 6, 2026; Published: September 1, 2026 Copyright: © 2026 Sadi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant dataset files are available from the OpenNeuro database (accession number ds005863, URL: https://openneuro.org/datasets/ds005863/versions/1.0.0 ). Funding: The author(s) received no specific funding for this work. Competing interests: The authors have declared that no competing interests exist. Introduction Attention-deficit/hyperactivity disorder (ADHD) is one of the most prevalent neurodevelopmental disorders, with global estimates suggesting it affects roughly 5–7% of children and approximately 2–5% of adults worldwide [ 1 , 2 ]. It is characterized by ongoing patterns of inattention, hyperactivity and impulsivity [ 3 ] that emerge during preschool years and intensify in school-age environments [ 4 , 5 ]. These symptoms negatively impact academic, personal and social functioning, with effects persisting into adulthood [ 6 , 7 ]. While prevalence estimates vary by region and diagnostic criteria, a 2023 umbrella review estimated the global rate in children and adolescents at 8.0% [ 8 ], reflecting both genuine cross-national differences and variation in screening and diagnostic practices. Current diagnostic practices rely heavily on subjective behavioral assessments and clinical interviews, which are susceptible to reporting biases, cultural variations and lack objective neurophysiological markers [ 9 ]. This often results in delayed or missed diagnoses, particularly where specialized expertise is limited. Electroencephalography (EEG) offers a promising, low-cost avenue for developing objective ADHD biomarkers by directly measuring neural activity with high temporal resolution [ 10 – 12 ]. Recent deep learning advances have demonstrated automated EEG-based ADHD classification [ 13 , 14 ]; however, these approaches focus exclusively on brain signals and overlook socioeconomic context that profoundly influences neurodevelopment and ADHD risk. Extensive neuroscience research establishes that socioeconomic status (SES) including parental education, household income, food security and neighborhood quality significantly shapes brain development and cognitive function throughout life [ 15 , 16 ], with lasting effects into adulthood [ 9 ]. Socioeconomic adversity associates with increased attention and executive function deficits overlapping with ADHD symptomatology [ 17 ], suggesting SES factors may modulate neural signatures captured by EEG. Despite well documented relationships between SES and both brain function and ADHD risk, machine learning approaches have not systematically integrated socioeconomic information with neurophysiological data. While neuroscience research has established measurable associations between socioeconomic indicators and electrophysiological responses [ 18 ], this knowledge has not been translated into computational models for clinical classification. Additionally, prior EEG-based studies typically rely on single data representations, temporal or spectral features, potentially missing complementary information [ 19 ]. In this study, we introduce a multimodal deep learning architecture integrating three complementary streams: (1) temporal features from raw EEG via 1D convolutional-recurrent networks, (2) spectro-temporal features from spectrograms via 2D convolutional neural networks (CNNs) with attention and (3) socioeconomic features via feedforward processing. Using the Cognitive Electrophysiology in Socioeconomic Context dataset [ 18 ], which includes EEG data from 127 young adults across four cognitive tasks along with detailed socioeconomic information, this study aims to address two key gaps in the existing research. First, although neuroscience has clearly shown that socioeconomic factors influence the neural activity captured by EEG, no previous studies have examined whether incorporating this knowledge can directly improve the accuracy of automated ADHD classification. We examine whether combining detailed socioeconomic information covering childhood family background and adult living conditions with multimodal EEG data can improve classification accuracy compared to relying on EEG signals alone. Second, different cognitive tasks measure different aspects of attention and executive control. We therefore test whether the benefits of adding socioeconomic information are limited to specific tasks or reflect a consistent, generalizable effect. We evaluate our approach across four diverse cognitive paradigms: passive attention processing (auditory oddball), active target detection (visual oddball), response inhibition (flanker task) and visual search efficiency. This multi-task evaluation allows us to assess whether socioeconomic context provides consistent discriminative value across different cognitive demands or if its utility depends on the specific attentional processes being assessed. We hypothesized that (H1) a multimodal model incorporating both EEG and SES features would achieve higher ADHD risk screening accuracy than an EEG-only baseline, and (H2) this benefit would hold across multiple cognitive paradigms rather than being limited to a single task; both hypotheses concern the SES effect and are tested exclusively on our primary dataset, since the external paediatric dataset used later in this study contains no socioeconomic variables. We separately hypothesized that (H3) the EEG-only backbone, applied without modification to that independent paediatric dataset, would achieve subject-independent accuracy at least comparable to the prior subject-independent benchmark reported for it; this hypothesis concerns architectural transfer alone and carries no SES claim. Related work This section reviews methods for EEG-based ADHD detection, the influence of socioeconomic factors on neurodevelopment and approaches to multimodal integration. EEG-based ADHD classification Traditional machine learning approaches. Early EEG-based ADHD classification relied on manually engineered features (that is, domain-specific signal properties such as power spectra and entropy measures defined by experts rather than learned from data) and conventional machine learning algorithms [ 12 , 20 ]. As computational approaches matured, researchers employed statistical measures, power spectral densities across frequency bands [ 11 ] and nonlinear dynamics features such as entropy and fractal dimensions [ 21 – 23 ]. Feature selection techniques included statistical hypothesis testing and LASSO (Least Absolute Shrinkage and Selection Operator) regularization [ 24 ]. Altinkaynak et al. [ 25 ] showed that integrating time-domain Event-Related Potential features with frequency-domain characteristics achieved 91.3% binary classification accuracy on a small pediatric dataset using a support vector machine with leave-one-out cross-validation, suggesting multimodal feature representations within EEG could enhance classification performance. However, traditional approaches faced fundamental limitations: dependence on domain expertise for feature definition, potential information loss through dimensionality reduction and inability to capture complex hierarchical patterns [ 12 ]. Reported classification accuracies varied widely (70–95%) depending on feature engineering choices, dataset characteristics and evaluation protocols [ 20 ]. It is worth noting that many of these studies relied on relatively small, demographically specific samples, which may limit how broadly their findings can be applied. More recently, Chandela et al. [ 26 ] proposed a unified approach combining successive multivariate variational mode decomposition (SMVMD) with K-nearest neighbor classification for detecting multiple neurodevelopmental disorders in children, achieving 99.17% accuracy for ADHD classification. While such methods demonstrate high performance through carefully engineered features, they remain constrained by the need for domain expertise and may not fully capture hierarchical representations that deep learning can automatically discover. Deep learning advances. Deep learning transformed EEG analysis by enabling end-to-end feature learning. Researchers have successfully applied CNNs to spectro-temporal EEG representations. Dubreuil-Vall et al. [ 19 ] applied CNNs to event-related spectral EEG for ADHD classification, while Tawhid et al. [ 27 ] demonstrated that spectrogram-based approaches could achieve high accuracy (exceeding 95–99%) for autism spectrum disorder, establishing the viability of treating EEG spectrograms as images for CNN processing. Moghaddari et al. [ 28 ] developed a 13-layer CNN achieving 99.06% accuracy using segmented 4-second epochs. Attention mechanisms have enhanced both performance and interpretability [ 29 ], with recent work proposing autoencoder-ResNet pipelines [ 13 ] and specialized architectures for EEG feature maps [ 14 ]. Tawhid et al. [ 30 ] demonstrated that generic CNN architectures can achieve robust cross-disorder classification across six neurological conditions including ADHD, suggesting shared neural signatures across neurodevelopmental disorders. Most recently, Hossain and Tawhid [ 31 ] extended the spectrogram-CNN framework to schizophrenia detection, achieving 98.31–99.82% accuracy across two public EEG datasets and demonstrating that STFT-derived mel-spectrogram representations retain high discriminative power even when evaluated through brain-lobe-specific channel subsets, further reinforcing the viability of spectro-temporal image inputs for EEG-based psychiatric classification. Methodological variations substantially impact classification outcomes [ 32 ], with reported accuracies ranging from 76–99% [ 33 , 34 ] reflecting algorithmic differences, dataset variations and evaluation protocols. Despite these advances, recent approaches share two critical limitations: (1) they process neurophysiological data isolated from socioeconomic context, despite established neuroscience showing environmental influences on brain development [ 16 ] and (2) most rely on single EEG representations (either temporal or spectral) rather than exploiting their complementary nature. Even in cases where EEG-only baselines approach high accuracy, there are reasons to expect SES integration to add value. Screening instruments such as the ASRS capture symptom burden shaped by environmental conditions, meaning that brain signals alone may not fully account for the variance introduced by socioeconomic adversity. Modeling this context explicitly may therefore improve predictive accuracy by accounting for environmental sources of variance that brain signals alone do not capture. Socioeconomic context in neurodevelopment While computational approaches have focused exclusively on brain signals, developmental neuroscience has established that socioeconomic status profoundly shapes brain structure and function [ 15 , 16 ]. Early-life SES influences cognitive development in domains central to ADHD symptomatology, particularly executive function and attention [ 15 , 16 , 35 – 37 ]. Electrophysiological research shows SES affects event-related potential components during cognitive tasks [ 18 ], with effects persisting into adulthood. Epidemiological studies consistently link socioeconomic adversity to elevated ADHD prevalence [ 17 ]. Despite robust evidence connecting SES to both brain function and ADHD risk, this knowledge has not been translated into computational models for clinical classification. It is also important to acknowledge that ADHD has a strong heritable component [ 38 ], which means low SES alone is not predictive of the disorder; rather, socioeconomic context likely interacts with genetic predisposition to modulate symptom expression and neural signatures. Furthermore, access to formal clinical diagnosis is itself SES-stratified: individuals from lower-income backgrounds are less likely to receive a diagnosis even when symptom burden is comparable [ 17 ], which means ASRS-based screening labels may themselves carry socioeconomic gradients that a model must account for rather than ignore. Multimodal learning and research gap Multimodal learning integrates heterogeneous data types through fusion strategies including early concatenation, late combination and attention-based weighting [ 39 ]. In broader healthcare applications, multimodal approaches have successfully combined medical imaging with clinical variables. Within EEG analysis specifically, recent state-of-the-art methods in related affective and neurological domains have begun demonstrating the critical value of contextual data. For instance, demographic factors have been shown to significantly impact the generalizability of EEG-based emotion recognition models [ 40 ] and demographic attention mechanisms have been successfully employed to improve EEG-based depression detection [ 41 ]. Furthermore, compact multilayer perceptrons (MLPs) have been effectively used to fuse clinical and demographic tabular data with complex modalities to predict post-stroke cognitive decline [ 42 ]. Despite these advances in related fields, EEG-based ADHD classification has primarily focused on combining different neurophysiological measurements rather than integrating neural data with contextual socioeconomic information. This reveals a critical disconnect: while neuroscience demonstrates socioeconomic shaping of neural substrates, computational models treat brain signals in contextual isolation. Our work bridges this gap by testing whether SES features improve ADHD classification when integrated with EEG data through attention-based fusion of temporal dynamics, spectro-temporal patterns and socioeconomic context. Materials and methods In this section, we outline the approach we took for ADHD risk screening, including a breakdown of the dataset, how we preprocessed the data, the architecture of our model and our experimental setup. Dataset description This study utilizes the “Cognitive Electrophysiology in Socioeconomic Context in Adulthood” dataset [ 18 ] which is publicly available on OpenNeuro at accession ds005863 ( https://openneuro.org/datasets/ds005863 ). The dataset was collected at the University of Florida (USA) and investigates relationships between cognitive electrophysiology, socioeconomic status and ADHD symptomatology in young adults. The dataset includes 127 young adults (mean age: years; 57.5% female) recruited at the University of Florida (USA) from diverse socioeconomic backgrounds [ 18 ]. We used data from four ERP CORE cognitive tasks. The auditory oddball task required participants to press a button whenever an infrequent target tone appeared among frequent standard tones, probing passive auditory attention and P3 generation. The visual oddball task followed the same logic but used visual stimuli, requiring detection of an infrequent target shape. The flanker task measured response inhibition: participants indicated the direction of a central arrow while ignoring flanking arrows that were either congruent or incongruent with the target. The visual search task assessed attentional selection by requiring participants to find a target letter among distractor letters. Full task specifications are available in the original dataset paper [ 18 ]. Available participant counts per task were: auditory oddball (114), visual oddball (99), flanker (73) and visual search (72). Not all participants completed every task, as the dataset was collected across multiple ongoing studies [ 18 ]. Our inclusion criteria required both complete EEG recordings for a given task and a valid ADHD screening label from the ASRS-v1.1. After applying these criteria, final sample sizes were: auditory oddball (n = 95), visual oddball (n = 84), flanker (n = 60) and visual search (n = 59). The reduction from the total pool of 127 therefore reflects both task non-completion and missing screening data, rather than EEG data quality failure alone. Fig 1 illustrates the participant flow and exclusion criteria for each cognitive task. Download: PNG larger image TIFF original image Fig 1. Participant flow diagram across four cognitive tasks. From an enrolled pool of N = 127 participants, task-specific samples reflect two sequential inclusion criteria: (1) availability of a complete EEG recording for the given task and (2) a valid ASRS-v1.1 screening label. Final analysed samples: Auditory Oddball ( n = 95); Visual Oddball ( n = 84); Flanker ( n = 60); Visual Search ( n = 59). https://doi.org/10.1371/journal.pone.0357213.g001 ADHD symptomatology was assess
## Related Clinical Research

- [Circadian Rhythm Disturbances in Young Adults: Clinical, Functional and Genetic Correlates](https://medichelpline.com/clinical-feed/medrxiv-19-cumulative-indicators-of-circadian-rhythm-disturbance-and-its-clinical.md)
- [Genomic SEM Reveals Functionally Distinct Genetic Factors Across the Psychosis Spectrum](https://medichelpline.com/clinical-feed/medrxiv-22-parsing-the-functional-heterogeneity-of-the-psychosis-spectrum-a-clinically.md)
- [Gut Microbial Dysbiosis Patterns in Depression, Anxiety, Schizophrenia and ASD: Comparative Analys](https://medichelpline.com/clinical-feed/biorxiv-23-deciphering-the-gut-brain-dialogue-a-survey-based-and-in-silico-comparative.md)
- [EEG-derived indices classify sporadic Creutzfeldt-Jakob disease from Alzheimer’s and healthy aging](https://medichelpline.com/clinical-feed/plos-one-17-classification-of-sporadic-creutzfeldt-jakob-disease-based-on-resting-state.md)
- [Predicted Brain Age Differences Across Nine Disorders: Case-Control MRI Study](https://medichelpline.com/clinical-feed/plos-medicine-1-brain-aging-patterns-among-nine-neurological-disorders-a-case-control-study.md)

## Navigation
- [← Back to Neurology Feed](https://medichelpline.com/clinical-feed/neurology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.