Effective Connectivity in the Brain – An analysis on MEG data using causal discovery
Neuroimaging techniques have led to an increase of our knowledge of the brain, in particular with regard to connectivity between different brain regions. This study investigates how causal discovery is used for finding effective connectivity in task based MEG data. The PC (Peter-Clark) algorithm operates under the correctness of causal discovery and tests on simulated data show that it is feasible for finding effective connectivity on large sparse graphs. Data of two different tasks have been analysed for connectivity using the PC algorithm, a presented visual stimulus and the remembering of that visual stimulus. This study shows that differences in connectivity are found between the tasks, showing distinct different patterns.
Faculteit der Sociale Wetenschappen