Related Experiment Video
Updated: Jun 4, 2025
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
15.6K
Non-parametric full cross mapping (NFCM): a highly-stable measure for causal brain network and a pilot application
Danni Yang1,2, Wentao Lin3, Minghui Liu3
1School of Materials Science and Engineering, Lanzhou University of Technology, Lanzhou 730050, People's Republic of China.
Journal of Neural Engineering
|December 18, 2024
Summary
A new parameter-free method, non-parametric full cross mapping (NFCM), accurately measures causal brain networks. This technique identifies higher causal coupling in successful epilepsy surgery zones, aiding in precise treatment for drug-resistant epilepsy in children.
Area of Science:
- Neuroinformatics
- Computational Neuroscience
- Systems Neuroscience
Background:
- Causal brain network analysis from neurophysiological signals is crucial but limited by traditional algorithms' computational demands and instability.
- Existing methods often require time-consuming parameter tuning, hindering practical application.
Purpose of the Study:
- To introduce a novel, parameter-free technique for robust causal brain network measurement.
- To enhance the accuracy and stability of brain network analysis compared to existing methods.
- To validate the technique's clinical utility in pediatric drug-resistant epilepsy.
Main Methods:
- Developed 'non-parametric full cross mapping' (NFCM), improving phase-space reconstruction and utilizing simplex projection for cross-mapping estimates.
- Conducted numerical experiments to assess NFCM's stability and accuracy against noise and baseline methods.
- Applied NFCM to stereoelectroencephalogram data from children with drug-resistant epilepsy, analyzing brain network dynamics during seizures.
Main Results:
- NFCM demonstrated superior quantization stability and lower coefficient of variation, even under system noise, outperforming six baseline methods.
- Analysis of pediatric drug-resistant epilepsy revealed significantly higher average causal coupling in epileptogenic zones of successful surgical outcomes (0.81 ± 0.04) compared to non-epileptogenic zones (0.40 ± 0.03) (P<0.001).
- No significant difference in causal coupling was observed between successful and failed surgical outcomes in the DREC cohort.
Conclusions:
- The NFCM technique provides a stable and accurate method for measuring causal brain networks.
- Causal brain network analysis using NFCM serves as a credible biomarker for localizing epileptogenic zones in pediatric drug-resistant epilepsy.
- This advancement holds promise for improving precision medicine approaches in DREC treatment.
