Evaluating Dynamic Correlations in fMRI

• Activity Number: 431
• Type: Invited
• Date/Time: Wednesday, August 6, 2014 : 8:30 AM to 10:20 AM
• Sponsor: SSC
• Author(s): Martin Lindquist*+
• Companies: Johns Hopkins Bloomberg School of Public Health
• Keywords: fMRI ; brain ; connectivity ; dynamics

Abstract

Recently, there has been significant research focused on quantifying changes in functional connectivity over time, as it is thought this may provide insight into the fundamental properties of brain networks. The most common approach towards investigating the dynamic nature of functional connectivity has been the so-called sliding window approach, where the correlation is computed over a window of fixed length that is moved across time, creating a time-varying measurement of the correlation. This approach towards studying dynamic correlations has a number of shortcomings, which we illustrate in this talk. We will also use discuss the fact that the problem of studying time-varying variances and correlations between multivariate time series has been extensively studied in the finance literature, and we will look at some of the solutions from literature as inspiration for analyzing fMRI data. In this work we investigate the properties the some of these methods in simulation studies and application to real data.