Related Experiment Video
Updated: Jan 15, 2026
08:17
A Semantic Priming Event-related Potential ERP Task to Study Lexico-semantic and Visuo-semantic Processing in Autism Spectrum Disorder
Published on: April 12, 2018
11.1K
A comprehensive EEG dataset and performance assessment for Autism Spectrum Disorder
Melinda Melinda1, Prima D Purnamasari2, Fahmi Fahmi3
1Department of Electrical and Computer Engineering, Universitas Syiah Kuala, Banda Aceh, Indonesia. melinda@usk.ac.id.
Scientific Reports
|October 7, 2025
Summary
Independent Component Analysis (ICA) and Discrete Wavelet Transform (DWT) improve Electroencephalography (EEG) signal quality for Autism Spectrum Disorder (ASD) detection. ICA excels at denoising, while DWT preserves signal characteristics for better ASD classification.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Early diagnosis of Autism Spectrum Disorder (ASD) is crucial for intervention and reducing healthcare costs.
- Electroencephalography (EEG) is a non-invasive tool for identifying neural patterns associated with ASD.
- Improving EEG signal quality is essential for accurate ASD detection.
Purpose of the Study:
- To evaluate the effectiveness of three EEG preprocessing techniques: Butterworth, Discrete Wavelet Transform (DWT), and Independent Component Analysis (ICA).
- To enhance EEG signal quality for improved Autism Spectrum Disorder classification.
- To compare the performance of these techniques using various signal quality metrics and Hjorth parameters.
Main Methods:
- Applied Butterworth, DWT, and ICA preprocessing to EEG signals from neurotypical and ASD individuals.
- Assessed performance using Signal-to-Noise Ratio (SNR), Mean Absolute Error (MAE), Mean Squared Error (MSE), Spectral Entropy (SE), and Power Spectral Density (PSD).
- Computed Hjorth parameters (activity, mobility, complexity) to analyze neural dynamics.
Main Results:
- ICA yielded the highest SNR (86.44 neurotypical, 78.69 ASD), indicating superior denoising.
- DWT demonstrated the lowest error metrics (MAE: 4785.08, MSE: 309,690 for ASD), preserving signal integrity.
- Hjorth parameters showed higher activity and complexity in neurotypical EEGs compared to ASD EEGs.
Conclusions:
- ICA is optimal for applications requiring high signal clarity in ASD EEG analysis.
- DWT provides a balanced approach, effectively preserving signal features for ASD classification.
- Findings support the development of advanced, EEG-based diagnostic tools for early ASD detection and clinical support.
