Related Experiment Video
Updated: May 26, 2026
09:32
Cortical Source Analysis of High-Density EEG Recordings in Children
Published on: June 30, 2014
Performance evaluation of the Champagne source reconstruction algorithm on simulated and real M/EEG data
Julia P Owen1, David P Wipf, Hagai T Attias
1Biomagnetic Imaging Laboratory, Dept. Radiology and Biomedical Imaging, UCSF San Francisco, CA, USA.
Neuroimage
|January 3, 2012
Summary
Champagne, a new source localization algorithm, accurately pinpoints brain activity from M/EEG data. It outperforms existing methods, even with noisy or overlapping brain signals, for better neuroscience research.
Area of Science:
- Neuroscience
- Biophysics
- Signal Processing
Background:
- Source localization is crucial for understanding brain function using M/EEG data.
- Existing algorithms struggle with correlated sources and noise.
- A novel approach is needed to improve the accuracy and robustness of source localization.
Purpose of the Study:
- To evaluate the performance of the Champagne algorithm for M/EEG source localization.
- To compare Champagne against benchmark algorithms using simulated and real data.
- To demonstrate Champagne's robustness to correlated sources and noise.
Main Methods:
- Champagne algorithm developed within an empirical Bayesian framework for sparse inverse problem solutions.
- Testing on simulated M/EEG data with challenging source configurations.
- Validation on real MEG and EEG data, including analysis of correlated brain activity.
Main Results:
- Champagne significantly outperforms benchmark algorithms in source localization accuracy and time course estimation in simulations.
- The algorithm demonstrates superior robustness to correlated brain activity in real MEG data.
- Champagne successfully resolves distinct and functionally relevant brain areas in real MEG and EEG data.
Conclusions:
- Champagne offers a significant advancement in M/EEG source localization.
- Its ability to handle noise and correlated sources enhances its utility for neuroscience research.
- Champagne provides more accurate and reliable mapping of brain activity.
