http://cogimage.dsi.cnrs.fr/seminaires/soutenance_jerbi_2006.htm
Résumé :
To fully describe functional cerebral networks, we need to localize the structures involved and characterize the mechanism by which they interact. The biological hypothesis put to the test in the present thesis is that distant neural populations interact by synchronization of their rhythmic activity. The investigations were carried out, both on methodological and experimental levels, using the framework of non-invasive lectromagnetic imaging at high spatio-temporal resolution. In the first part of this work, we developed new models for the analysis of magnetoencephalography (MEG) data. The proposed multipolar model extends previous equivalent source models by accounting for spatially extended cortical activity and thus provides a more realistic description of the sources underlying the measured data. Our results show that the developed multipolar model outperforms the classical dipole model. Next, we set out to investigate large-scale brain integration and evaluate MEG’s ability to detect the neural mechanisms involved in the latter. To do so, we set up a sustained visuomotor coordination experiment to study neural interactions between multiple areas across the brain. By combining source localization, advanced spectral analysis tools and statistical inference techniques, we revealed several phenomena involved in sensorimotor integration and motor control. Our results provide novel evidence for task-related frequency-specific modulations of phase synchronization between distant but functionally related brain areas and, interestingly, also between cerebral activity and behavioral signals (such as the speed of the active hand). The patterns of long-range synchronization and the sources of task-related neural oscillations revealed in this thesis vary over a large a range of frequencies (from 2 Hz to 90 Hz). Taken together, our findings provide substantial evidence for the functional role of synchronized oscillatory activity between distinct brain areas and