Tutorial 28: Connectivity

[TUTORIAL UNDER DEVELOPMENT: NOT READY FOR PUBLIC USE]

Authors: Francois Tadel, Esther Florin, Sergul Aydore, Syed Ashrafulla, Elizabeth Bock, Sylvain Baillet

During the past few years, the research focus in brain imaging moved from localizing functional regions to understanding how different regions interact together. It is now widely accepted that some of the brain functions are not supported by isolated regions but rather by a dense network of nodes interacting in various ways. In order to quantify the amount of information exchanged between regions, experts in signal processing developed metrics to compare signals recorded in these different regions.

These inter-regional measures can help us explore the brain dynamics by understanding if two regions are activated synchroneously during a task (functional connectivity) or linked by causal interactions (effective connectivity). This tutorial introduces the measures and display tools available in Brainstorm to explore this inter-regional connectivity.

Connectivity processes

In the Process1 and Process2 tabs, the menu "Connectivity" contains the following options:

• Correlation
• Coherence (Imaginary coherence and Magnitude-squared coherence)
• Granger causality
• Granger causality (spectral)
• Phase-locking value (PLV)

Each of these metrics offer several variations:

• [1xN]: Connectivity between one signal and all the other signals in the same set of signals
  (ie. one sensor and all the other sensors in the same file, one source and all the other sources, etc). The output file has the same dimensions as the input files and can be visualized in a similar way.

• [NxN]: Connectivity between all the possible pairs of signals in the input files (recordings, source maps or scouts). The output file is full connectivity matrix [Nsignals x Nsignals x Ntime x Nfreq].

• [AxB]: The Process2 tabs offers more flexibilty. You can select different types of files in the two lists FilesA and FilesB. For example, by selecting recordings in FilesA and sources in FilesB, you can compute the connectivity between one sensor and all the sources.

Simulate auto-regressive signals

Process: Simulate > Simulate AR signals

Process: Simulate > Simulate AR signals (random)

Process: Simulate > Simulate generic signals

Example 1: Correlation sensor-sources

• In Process1, select

Example 2: Coherence scout-scout

Method: Correlation

Method: Coherence

Method: Granger causality

Method: Phase locking value

Method: Phase Transfer Entropy

This method is a directed measure of connectivity (akin to Granger in the frequency domain), which has had some success lately [Lobier et al., 2014]. The output values are normalized between -0.5 and 0.5, with 0 meaning no preferred directionality. So positive values indicate inflow (B->A) and negative ones indicate outflow (A->B).

Unconstrained sources

Describe how the three orientations are handled.

On the hard drive

Document the file tags

Document how to extract the connect matrix

How to input your own connect matrix

Additional documentation

Articles

• Phase transfer entropy: Lobier M, Siebenhühner F, Palva S, Palva JM

  Phase transfer entropy: A novel phase-based measure for directed connectivity in networks coupled by oscillatory interactions, NeuroImage 2014, 85:853-872

Forum discussions

• Forum: Connectivity matrix storage:http://neuroimage.usc.edu/forums/showthread.php?1796

• Forum: Comparing coherence values: http://neuroimage.usc.edu/forums/showthread.php?1556

• Forum: Reading NxN PLV matrix: http://neuroimage.usc.edu/forums/showthread.php?1681-PLV-NxN-Read-matrix

• Forum: Scout function and connectivity: http://neuroimage.usc.edu/forums/showthread.php?2843

• Forum: Unconstrained sources and connectivity: http://neuroimage.usc.edu/forums/t/problem-with-surfaces-vs-volumes/3261

• Forum: Digonal values: http://neuroimage.usc.edu/forums/t/choosing-scout-function-before-or-after/2454/2