Functional connectivity of the brain with wavelets

old_uid2458
titleFunctional connectivity of the brain with wavelets
start_date2007/03/19
schedule12h
onlineno
detailsinvité par Line Garnero
summaryIn this talk, I will present a method to analyse fMRI or MEG data, in resting state or task related, in order to characterize the network of connections in the brain. I will show how to measure the functional connectivity in the brain at different frequency bands, using a wavelet correlation measure. In the particular case of multivariate long memory processes, the wavelet correlation function reaches an asymptote depending on the difference of the memory parameters of each process. Then, significant correlations are marked as effective connections, this allows us to construct a graph. Finally, topological parameters of these graphs, such as efficiency, are extracted, and comparisons between populations are conducted. I will illustrate this method on a new data set with young and old people at resting state. We show that graph-efficiency is reduced disproportionately to cost in older people and the detrimental effects of age on graph-efficiency are localised at frontal and temporal cortical and subcortical regions.
responsiblesBaillet