Localisation and time courses of CMV generators from MFT analysis of average MEG signals

Dammers, Jürgen (2000). Localisation and time courses of CMV generators from MFT analysis of average MEG signals. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.0000f969

Abstract

The research presented here is divided into two parts. The first part addresses questions relating to the localisation capability of magnetoencephalography (MEG), with emphasise on testing the accuracy of typical MEG systems before any source reconstruction is applied. The second and main part is concerned with localisation and time courses of neuronal generators contributing to both the slow and fast neuromagnedc field changes associated with the contingent negative variation (CNV) in the normal human brain. We use multi-channel and full head MEG systems to study the magnetic counterpart of the CNV, the contingent magnetic variation (CMV). MEG analysis of such near DC-like signals requires advanced source reconstruction that is able to identify widely distributed as well as focal sources which often fire simultaneously from cortical as well as from deeper brain structures. We use magnetic field tomography (MFT) to extract time courses of regional brain activity. Results are presented from a multi-subject CMV study performed using the BTi MAGNES II system (Experiment 1) and a single subject experiment, using the CTF whole cortex system (Experiment 2).
From Experiment 1 we identified four different CMV generators (auditory cortex, sensorimotor cortex, inferior prefrontal cortex, posterior inferior parietal area) and observed priming of the auditory cortex as part of the early CMV complex and the priming of the sensorimotor cortex as part of the late CMV complex. These results were confirmed by Experiment 2, where the full head coverage also revealed two additional areas, the supplementary motor area and the posterior cingulate cortex, which were dramatically reduced when identical runs were repeated. The SMA activity has been notoriously difficult to identify non-invasively from electrophysiological data, especially from MEG, so our success in identifying them clearly and showing how they change with repetition can he considered as the highlight of our project.