Layman’s abstract for paper on a Potts‐mixture spatiotemporal joint model for combined magnetoencephalography and electroencephalography data

Every few days, we will be publishing layman’s abstracts of new articles from our prestigious portfolio of journals in statistics. The aim is to highlight the latest research to a broader audience in an accessible format.

The article featured today is from the Canadian Journal of Statistics, with the full article now available to read here.

Song, Y., Nathoo, F. and Babul, A. (2019), A Potts‐mixture spatiotemporal joint model for combined magnetoencephalography and electroencephalography data. Can J Statistics, 47: 688-711. doi:10.1002/cjs.11519

Magnetoencephalography (MEG) and electroencephalography (EEG) have been widely used to study brain activity over the past several decades by non-invasive measurements obtained from sensor arrays placed at various locations on the scalp (EEG) or above the scalp (MEG). These sensor arrays can capture the time-varying electromagnetic field that exists around the head, which can be used to estimate sources of neural activity inside the brain. This source localization problem is one the most challenging statistical problems in imaging neuroscience and represents an ill-posed inverse problem.
Often, the solution can be improved by incorporating data from more than a single modality. Therefore, a new Bayesian spatial finite mixture model that combines the data from three different modalities, MRI, MEG, and EEG is proposed. This new model builds on the mesostate-space model developed by Daunizeau and Friston (2007) with major extensions that consider the combination of different modalities for a joint model as well as the spatial dependence across different brain regions. An efficient procedure for simultaneous point estimation and model selection based on the iterated conditional modes algorithm combined with local polynomial smoothing has also been derived. The proposed method results in a novel estimator for the number of mixture components and is able to select active brain regions. This new methodology is evaluated through extensive simulation studies and an application examining the visual response to scrambled faces.