EEG-Based ADHD Diagnosis Using Autoencoder and Reptile Search Algorithm Integrated with Machine Learning.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Biomedical Engineering

Background:

• Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder impacting cognitive and behavioral functions.
• Current diagnostic methods rely on subjective, time-consuming, and costly assessments like questionnaires and interviews.
• Limitations in traditional ADHD diagnosis hinder early detection and intervention.

Purpose of the Study:

• To develop and evaluate a machine learning-based approach for objective ADHD diagnosis using electroencephalogram (EEG) data.
• To compare the efficacy of Random Forest and AdaBoost classifiers in identifying ADHD patterns from EEG signals.
• To enhance feature extraction and selection for improved diagnostic accuracy.

Main Methods:

• Utilized electroencephalogram (EEG) data for ADHD diagnosis.
• Employed Random Forest and AdaBoost machine learning classifiers.
• Implemented Reptile Search Algorithm with an autoencoder for feature extraction and selection.
• Evaluated model performance using accuracy, precision, recall, F1-score, and AUC.

Main Results:

• Random Forest achieved 92.36% accuracy, precision, recall, and F1-score, outperforming AdaBoost (89.78%).
• Random Forest demonstrated superior effectiveness in distinguishing ADHD cases with an ROC AUC score of 0.93.
• The machine learning approach showed higher diagnostic accuracy compared to traditional methods.

Conclusions:

• Machine learning applied to EEG data provides a promising, objective, and reliable tool for ADHD diagnosis.
• This method offers an effective alternative to traditional assessments, facilitating timely intervention.
• The findings support the use of advanced computational techniques for improved ADHD diagnosis and management.