Peer-reviewed article
A Framework for Predicting Neurofeedback Treatment Response in ADHD Using EEG Functional Connectivity and Genetic Algorithm-Driven Channel Selection.
Ziyue Li, Zhinuan Zhu, Dong Li, Saiyu Huang ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This paper presents a computational method for predicting which children with ADHD will respond well to neurofeedback treatment using brain imaging data. The framework analyzes EEG signals from 60 children aged 6-12 years to identify patterns that distinguish treatment responders from non-responders. The method combines feature extraction, dimensionality reduction, and machine learning classification, achieving 84.72% accuracy in predicting treatment response.
Key points
- The framework uses functional brain connectivity (Phase Locking Value) from EEG recordings during early-to-mid neurofeedback treatment
- A genetic algorithm identified six key brain channels (C3, C4, Cz, Fz, Fp1, T6) from the original 32-channel setup as most predictive
- The study population was 60 children with ADHD aged 6-12 years
- Ensemble machine learning classifiers achieved 84.72% accuracy in predicting responder status
- The authors note the framework requires validation on independent clinical datasets before clinical application
Limitations
This paper presents a computational prediction tool but does not evaluate the effectiveness of neurofeedback treatment itself for ADHD outcomes. The study is methodological and technical rather than an evaluation of whether the predicted responders actually experience clinical improvement.
Abstract
In this article, we present a computational framework for predicting treatment response to neurofeedback (NF) among patients with Attention-Deficit/Hyperactivity Disorder (ADHD). The proposed framework uses functional brain connectivity analysis of electroencephalogram (EEG) signals acquired during an early-to-mid NF treatment window to classify participants as eventual responders or non-responders. The six-stage algorithm was evaluated using an open-access EEG dataset from the Mendeley Data repository comprising 60 children with ADHD aged 6-12 years. The framework includes a preprocessing pipeline designed to reduce EEG artifacts and noise. Next, spectral features, specifically of the alpha and beta frequency bands, were extracted from the noise-reduced signals. In the fourth stage, functional connectivity was estimated by calculating Phase Locking Value (PLV) between all electrode pairs, thereby quantifying inter-channel phase synchronization. The fifth stage, which is an essential stage, was dimensionality reduction to find the most discriminative features. Dimensionality reduction was achieved in a two-process manner; for the first process, statistical screening was performed using Welch's t-test with FDR correction, followed by GA-based channel selection to identify the most discriminative electrode subset. The GA analysis identified a compact six-channel subset consisting of C3, C4, Cz, Fz, Fp1, and T6 from the original 32-channel montage. In the last classification stage, the reduced feature vectors were input as part of an ensemble of machine learning classifiers to achieve classification. Model performance was evaluated using subject-wise grouped cross-validation, with the best configuration achieving an accuracy of 84.72%. These results suggest that the proposed data-driven framework may support future research on individualized NF response prediction, pending validation on independent clinical datasets.