Related Experiment Video
Updated: Apr 19, 2026
11:00
Reliable Acquisition of Electroencephalography Data during Simultaneous Electroencephalography and Functional MRI
Published on: March 19, 2021
5.3K
Reliability of information-based integration of EEG and fMRI data: a simulation study
Sara Assecondi1, Dirk Ostwald, Andrew P Bagshaw
1School of Psychology, University of Birmingham, Birmingham, B17 2TT, U.K. sara.assecondi@me.com.
Neural Computation
|December 17, 2014
Summary
Estimating information in electroencephalographic-functional magnetic resonance imaging (EEG-fMRI) data requires careful consideration of binning strategies. The equipopulated binning method provides the most accurate and consistent results for analyzing complex neural data.
Area of Science:
- Neuroimaging
- Computational Neuroscience
- Biophysics
Background:
- Simultaneous electroencephalographic (EEG) and functional magnetic resonance imaging (fMRI) data analysis typically uses linear correlation methods.
- Information-theoretic measures like mutual information and entropy can capture higher-order correlations but are sensitive to estimation parameters.
- Previous research on information estimation accuracy has focused on invasive neurophysiological data, leaving a gap for EEG-fMRI.
Purpose of the Study:
- To systematically evaluate the accuracy of information-theoretic estimates for multivariate EEG-fMRI data.
- To investigate the influence of sample size, variable correlation, and binning strategies on information estimation.
- To compare different bias correction methods for information estimation in simulated EEG-fMRI data.
Main Methods:
- Generated simulated bivariate and trivariate distributions mirroring EEG-fMRI data statistical properties.
- Compared estimated information shared between simulated variables against their true numerical values.
- Assessed the impact of various binning strategies and estimation methods on accuracy.
- Evaluated the performance of bias correction techniques: asymptotically debiased (TPMC), jackknife debiased (JD), and best upper bound (BUB).
Main Results:
- The accuracy of information estimates is significantly influenced by the interplay between the chosen binning strategy and the estimation method.
- The equipopulated binning strategy demonstrated superior and consistent performance across different distributions and bias correction techniques.
- TPMC, JD, and BUB bias correction methods yielded comparable and consistent results across the tested distributions.
Conclusions:
- Equipopulated binning is recommended for accurate information estimation in EEG-fMRI analyses.
- Standard bias correction techniques (TPMC, JD, BUB) are reliable for EEG-fMRI data.
- This study provides crucial insights for advancing the analysis of complex, multimodal neuroimaging data.
