MEG-SIM Portal: Reconstructions from Realistic Simulations of Sensory and Cognitive Processing
To enable and encourage comparisons between MEG analysis algorithms, a partnership was formed between 5 research institutions. MEG/MRI data was acquired from 5 subjects at 3 of these institutions using 3 different MEG systems and paradigms from 3 sensory modalities. A grant from NIMH (R21MH080141) allowed us to create realistic simulated data from the data acquired through this partnership. Visual simulated data sets, along with tools to read these data, have been placed on a collaborative portal established and maintained at MRN (http://portal.mind. unm.edu/megsim). Additional simulated data sets are added as they are developed. The present study reports on the construction of three different types of realistic simulated data sets and results using a semi-automated, multidipole, spatio-temporal (CSST) algorithm for source localization. Our most recent data sets supply a variety of source configurations in multiple sensory modalities, as well as simulating functional connectivity. The results show that CSST accurately reconstructs both temporal and spatial characteristics of the simulated data sets.
KeywordsSimulations single trials working memory modeling oscillatory
Unable to display preview. Download preview PDF.