Related Experiment Video
Updated: Jan 3, 2026
07:52
Multiscale Investigations of Cortical Processing by Integrating Laminar Polytrodes and Optogenetics with Micro Electrocorticography in Rodents
Published on: May 23, 2025
737
Large-scale cortical travelling waves predict localized future cortical signals.
David M Alexander1, Tonio Ball2, Andreas Schulze-Bonhage3
1Perceptual Dynamics Laboratory, Brain and Cognition Research Unit, KU Leuven, Leuven Belgium.
Plos Computational Biology
|November 16, 2019
Summary
Predicting future brain activity is now more accurate using large-scale spatiotemporal dynamics. This method utilizes Fourier and principal component analysis to model brain waves, improving prediction accuracy for electrocorticography (ECoG) and magnetoencephalography (MEG) signals.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Signal Processing
Background:
- Predicting future brain activity is crucial but challenging.
- Existing methods often struggle with complex spatiotemporal dynamics.
Purpose of the Study:
- To develop a novel method for predicting future cortical activity phase.
- To leverage large-scale spatiotemporal dynamics for improved prediction accuracy.
Main Methods:
- Extracted brain signal dynamics using Fourier analysis and principal component analysis (PCA).
- Developed a data model to predict future signal phase at specific sites and frequencies.
- Analyzed electrocorticography (ECoG) and magnetoencephalography (MEG) data.
Main Results:
- The model identified dominant eigenvectors as smoothly propagating waves.
- Achieved low mean phase prediction errors (as low as 0.5 radians) in ECoG and MEG data.
- Prediction accuracy was highest in delta to beta bands and during high global power episodes.
Conclusions:
- Large-scale, low spatial frequency traveling waves are effective for predicting future brain activity.
- This approach surpasses current state-of-the-art prediction models.
- The method allows estimation of prediction errors due to irreducible activity dynamics.
