Related Experiment Video
Updated: Feb 28, 2026
11:00
Reliable Acquisition of Electroencephalography Data during Simultaneous Electroencephalography and Functional MRI
Published on: March 19, 2021
5.1K
Predicting EEG single trial responses with simultaneous fMRI and relevance vector machine regression.
Federico De Martino1, Aline W de Borst, Giancarlo Valente
1Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, The Netherlands. f.demartino@maastrichtuniversity.nl
Neuroimage
|August 10, 2010
Summary
This study explores combining electroencephalography (EEG) and functional Magnetic Resonance Imaging (fMRI) for brain dynamics. Multivariate regression effectively predicts EEG signals from fMRI data, linking brain activity across modalities.
Area of Science:
- Neuroscience
- Cognitive Science
- Biomedical Engineering
Background:
- Simultaneous electroencephalography (EEG) and functional Magnetic Resonance Imaging (fMRI) offer high temporal and spatial resolution for studying brain dynamics.
- EEG-fMRI trial-by-trial coupling aims to integrate these modalities for enhanced analysis, requiring simultaneous acquisition.
- Predicting signals between EEG and fMRI modalities is crucial for fully leveraging combined acquisitions.
Purpose of the Study:
- To investigate the prediction of EEG signals from fMRI data using multivariate regression.
- To explore the relationship between fMRI activation patterns and simultaneously recorded EEG responses during a complex cognitive task.
- To assess the feasibility of linking spatially distributed fMRI activation to specific EEG signal components.
Main Methods:
- Utilized multivariate Relevance Vector Machine (RVM) regression to model the relationship between fMRI and EEG data.
- Acquired simultaneous EEG-fMRI data during a cognitive task involving auditory cues, visual mental imagery, and visual targets.
- Analyzed evoked and induced oscillatory EEG responses in relation to fMRI time series.
Main Results:
- Multivariate regression successfully predicted evoked and induced oscillatory EEG responses from fMRI time series.
- EEG prediction from fMRI was significantly influenced by the hemodynamic response function's filtering effects.
- A small but significant contribution of single-trial modulations was identified in auditory evoked responses, linking fMRI patterns to EEG components.
Conclusions:
- Multivariate regression is a valuable method for predicting EEG from fMRI data.
- Simultaneous EEG-fMRI acquisitions enable the linkage of spatially distributed brain activity to specific electrophysiological signals.
- This approach facilitates a more comprehensive understanding of brain dynamics by integrating high-temporal and high-spatial resolution data.
