Bayesian Inference about the Directional Brain Network Modeled by Damped Harmonic Oscillators for Intracranial EEG Data

• Activity Number: 332 - SPEED: Section on Bayesian Statistical Science
• Type: Contributed
• Date/Time: Tuesday, August 1, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yinge Sun* and Tingting Zhang and Qiannan Yin and Brian Scott Caffo and Dana Boatman-Reich
• Companies: University of Virginia and University of Virginia and University of Virginia and Johns Hopkins and Johns Hopkins University
• Keywords: Bayesian inference ; ODEs ; cluster structure ; directional brain network ; damped harmonic oscillator

Abstract

We use ordinary differential equations (ODEs) to model the human brain as a continuous-time dynamic system consisting of biophysically interacting brain regions. In contrast to existing ODE models for fMRI and EEG data that focus on directional connectivity among only a few brain regions, we propose a high-dimensional ODE model motivated by statistical considerations to explore connectivity among many small brain regions recorded by intracranial EEG (ECoG). The new model is widely applicable to characterize various brain regions' oscillatory activity and their network of connectivity in a cluster structure. We develop a unified Bayesian framework to quantify the inadequacy in the proposed ODE model for the complex brain system, identify clusters and strongly connected brain regions, and map the brain's directional network where each network edge denotes a significant directional effect exerted by one brain region over another. We apply the proposed ODE model and Bayesian method to an EGoG data set collected in an auditory study, mapping the brain networks under different auditory stimuli. Our method revealed different brain networks under regular tone stimuli and speech stimuli.