Bayesian modelling for brain networks

• Date: 6 August 2020
• Time: 17:30 (Bangladesh standard time)
• Venue: Zoom online platform

Abstract:

Brain network data – measuring anatomical interconnections among a common set of brain regions – are increasingly collected for multiple individuals, and recent studies provide additional information on the brain regions of interest.

These predictors typically include the 3-dimensional anatomical coordinates of the brain regions and their membership to hemispheres and lobes. Although recent studies have explored the spatial effects underlying brain networks, there is still a lack of statistical analyses on the net connectivity topology, after controlling for spatial constraints.

We answer this question via a Bayesian latent space model for network data, obtaining a meaningful representation for the net connectivity architecture via a set of latent positions which are assigned mixtures of Gaussian priors.

This allows flexible inference on brain network structures not explained by closeness in the anatomical space and facilitates clustering among brain regions according to their latent positions

Biography:

Emanuele Aliverti is a post-doctoral research fellow at the Department of Statistical Sciences of the University of Padova. He obtained his Ph.D. in Statistics at the University of Padova (2020), supervised by Bruno Scarpa and David Dunson. The focus of his research is on developing flexible Bayesian methodologies and scalable algorithms to perform inference for high-dimensional structured data; some examples include networks, large contingency tables, functions, and curves.