Bayesian inference applied to the neural electromagnetic inverse problem

The problem of estimating the current distribution in the brain from surface EEG or MEG measurements (the so called neural electromagnetic inverse problem) is mathematically ill-posed; it has no unique solution in the most general, unconstrained case. We have developed a new probabilistic approach to the electromagnetic inverse problem, based on Bayesian inference. Unlike almost all other approaches to this problem, our approach does not result in a single "best" solution to the problem. Rather we estimate a probability distribution of solutions upon which all subsequent inferences are based. This distribution tabulates the multiple solutions that can account for any set of surface EEG/MEG measurements. Furthermore, features of these solutions that are highly probable can be identified and quantified. We applied this method to MEG data from a visual evoked response experiment in order to demonstrate the ability of the method to detect known features of human visual cortex organization. We also examined the changing pattern of cortical activation as a function of time.