Bayesian M/EEG source reconstruction with spatio-temporal priors

• Published in: NeuroImage (Orlando, Fla.) Vol. 39; no. 1; pp. 318 - 335
• Main Authors: Trujillo-Barreto, Nelson J., Aubert-Vázquez, Eduardo, Penny, William D.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 2008; Elsevier Limited
• Subjects: Algorithms; Bayes Theorem; Bayesian models; Brain Mapping - methods; Computer Simulation; Diagnosis, Computer-Assisted - methods; Electroencephalography - methods; Ensemble learning; Estimates; Evoked Potentials, Visual - physiology; Face; Generators; GLM; Humans; M/EEG source localization; Magnetoencephalography - methods; Models, Neurological; Pattern Recognition, Automated - methods; Pattern Recognition, Visual - physiology; Random variables; Spatio-temporal priors; Tomography; Variational Bayes; Bayesian models; GLM; M/EEG source localization; Variational Bayes; Ensemble learning; Spatio-temporal priors
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2007.07.062

This article proposes a Bayesian spatio-temporal model for source reconstruction of M/EEG data. The usual two-level probabilistic model implicit in most distributed source solutions is extended by adding a third level which describes the temporal evolution of neuronal current sources using time-domain General Linear Models (GLMs). These comprise a set of temporal basis functions which are used to describe event-related M/EEG responses. This places M/EEG analysis in a statistical framework that is very similar to that used for PET and fMRI. The experimental design can be coded in a design matrix, effects of interest characterized using contrasts and inferences made using posterior probability maps. Importantly, as is the case for single-subject fMRI analysis, trials are treated as fixed effects and the approach takes into account between-trial variance, allowing valid inferences to be made on single-subject data. The proposed probabilistic model is efficiently inverted by using the Variational Bayes framework under a convenient mean-field approximation (VB-GLM). The new method is tested with biophysically realistic simulated data and the results are compared to those obtained with traditional spatial approaches like the popular Low Resolution Electromagnetic TomogrAphy (LORETA) and minimum variance Beamformer. Finally, the VB-GLM approach is used to analyze an EEG data set from a face processing experiment.