Each week, 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 in Early View here.

Mohammed, S. and Dey, D.K. (2021), Scalable spatio‐temporal Bayesian analysis of high‐dimensional electroencephalography data. Can J Statistics. https://doi.org/10.1002/cjs.11592

This paper presents a novel modeling approach to identify brain regions that respond to a certain stimulus and uses those regions to distinguish between groups of subjects. It specifically deals with multi-subject electroencephalography (EEG) data to distinguish between alcoholic and control (non-alcoholic) groups. The EEG data contains measurements taken for each subject at different locations on the scalp across different times, thus having a complex structure with both spatial and temporal attributes. The selection of the brain locations is based on a Bayesian variable selection approach that utilizes the spatial and temporal characteristics of EEG data. A divide-and-conquer strategy is used for analysis and a computationally scalable algorithm is developed to deal with the challenges arising due to the high-dimensionality of EEG data. Simulation experiments are performed to demonstrate the efficiency of the algorithm both in terms of estimation and fast computation. Results using the highly scalable approach are presented for a case study on multi-subject EEG data.