Related Concept Videos
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
AVaTER: Fusing Audio, Visual, and Textual Modalities Using Cross-Modal Attention for Emotion Recognition.
Latent Space Projections and Atlases, a Cautionary Tale in Deep Neuroimaging using Autoencoders.
Transformer-Based Anomaly Detection for Neurodegenerative Screening in MRI Images.
Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction.
Automatic Seizure Detection using Hierarchical Spectral-Temporal Feature Learning with an Imbalance-Aware Transformer.
Pyramid Vision Transformer-Enhanced Conformer Network for Epileptic Seizure Recognition Using MultiChannel EEG Signals.
A Time-Frequency Decoupled Contrastive Learning Framework for Electroencephalography-Based Parkinson's Disease Diagnosis.
Related Experiment Video
Updated: May 27, 2025

Investigating Social Cognition in Infants and Adults Using Dense Array Electroencephalography dEEG
Published on: June 27, 2011
Autism Spectrum Disorder Detection Using Prominent Connectivity Features from Electroencephalography.
Zahrul Jannat Peya1, Mahfuza Akter Maria1, Sk Imran Hossain1
1Computer Science and Engineering Department, Khulna University of Engineering & Technology, Khulna 9203, Bangladesh.
This study introduces a new method using electroencephalography (EEG) signals to detect Autism Spectrum Disorder (ASD). Mutual Information (MI) feature extraction from Connectivity Feature Maps (CFMs) showed the best results for distinguishing ASD from control subjects.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Computational Psychiatry
Background:
- Autism Spectrum Disorder (ASD) presents significant diagnostic challenges due to its variability and lack of definitive biomarkers.
- Electroencephalography (EEG) signals, reflecting brain activity, offer a promising avenue for neurodevelopmental disorder detection.
- Current diagnostic methods for ASD lack objective indicators, necessitating advanced analytical approaches.
Purpose of the Study:
- To develop an effective feature extraction method from EEG signals for discriminating between individuals with ASD and neurotypical controls.
- To evaluate the efficacy of various connectivity features in identifying ASD patterns within EEG data.
- To establish a robust classification framework for ASD detection using advanced machine learning techniques.
Main Methods:
- Extraction of six prominent connectivity features from EEG signals: Cross Correlation (XCOR), Phase Locking Value (PLV), Pearson's Correlation Coefficient (PCC), Mutual Information (MI), Normalized Mutual Information (NMI), and Transfer Entropy (TE).
- Construction of Connectivity Feature Maps (CFMs) to represent spatial information from extracted features.
- Classification of ASD and control subjects using Convolutional Neural Networks (CNN) applied to CFMs.
Main Results:
- Connectivity Feature Maps (CFMs) demonstrated superior performance in distinguishing ASD from control subjects due to their inherent spatial information.
- The Mutual Information (MI) feature exhibited the highest accuracy in categorizing ASD and control participants.
- Performance improvements were observed with increased sample size and data segmentation, particularly for the MI feature.
Conclusions:
- The proposed method utilizing CFMs and CNNs provides a promising approach for objective ASD detection from EEG data.
- Mutual Information (MI) emerges as a key feature for enhancing the accuracy of ASD classification.
- Further validation across diverse datasets and larger sample sizes is recommended to solidify the clinical applicability of this EEG-based diagnostic tool.

