Related Experiment Video
Updated: Jan 5, 2026
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
6.0K
Establishing functional brain networks using a nonlinear partial directed coherence method to predict epileptic
Qizhong Zhang1, Yuejing Hu2, Thomas Potter3
1Intelligent Control & Robotics Institute, College of Automation, Hangzhou Dianzi University, Hangzhou, China.
Journal of Neuroscience Methods
|October 16, 2019
Summary
This study introduces a novel method using nonlinear partial directed coherence (NPDC) to analyze functional brain networks (FBNs) for improved epileptic seizure prediction, achieving high accuracy and extended prediction times.
Area of Science:
- Neurology
- Biomedical Engineering
- Data Science
Background:
- Epilepsy is a neurological disorder causing unpredictable seizures.
- Electroencephalography (EEG) is cost-effective for long-term epilepsy monitoring.
- Existing EEG analysis methods overlook brain network alterations in epilepsy.
Purpose of the Study:
- To develop a novel method for epileptic seizure prediction.
- To incorporate functional brain network (FBN) information into seizure prediction models.
Main Methods:
- Nonlinear partial directed coherence (NPDC) was used to measure FBNs.
- Extracted FBN features were integrated into an extreme learning machine (ELM) for prediction.
Main Results:
- The proposed method achieved high performance across all EEG frequency bands.
- Best accuracy reached 89.2% in the beta band.
- Optimal prediction time was 1356.4 seconds in the delta band.
Conclusions:
- The NPDC-based FBN approach surpasses existing graph theory and nonlinear methods.
- The developed prediction strategy is effective for predicting seizure onset.
