Attention Deficit Hyperactivity Disorder Diagnosis using non-linear univariate and multivariate EEG measurements: a...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder in childhood.
• Resting-state Electroencephalography (EEG) offers a non-invasive method to study brain activity.
• Distinguishing ADHD from typical development using EEG requires sophisticated analytical techniques.

Purpose of the Study:

• To propose and evaluate two novel classification algorithms for discriminating ADHD from normal children using EEG signals.
• To investigate the utility of non-linear entropy measures as features for ADHD classification.
• To compare the performance of univariate and multivariate feature extraction methods.

Main Methods:

• Extracted linear and non-linear univariate features (e.g., Theta/Beta Ratio, Sample Entropy) from individual EEG channels.
• Extracted non-linear multivariate features (e.g., multivariate Sample Entropy) from brain lobes.
• Employed Support Vector Machines (SVM), k-Nearest Neighbor (kNN), and Probabilistic Neural Network (PNN) classifiers.
• Utilized Multiscale Sample Entropy (MSE) for complexity analysis and entropy mapping for visualization.

Main Results:

• ADHD children exhibited higher brain activity and Theta/Beta Ratio (TBR) compared to controls.
• A more regular neural system and reduced dynamical complexity were observed in ADHD.
• Classification accuracy reached 99.58% using univariate non-linear features with Radial Basis Function (RBF) SVM.
• Classification using multivariate features achieved 90.63% accuracy with PNN.

Conclusions:

• Non-linear entropy measures are effective biomarkers for ADHD detection in EEG.
• Univariate feature analysis, particularly with non-linear measures, demonstrates superior performance for ADHD classification.
• The findings support the use of advanced signal processing techniques for objective ADHD diagnosis.