|
From resting fMRI to directed functional networks?| old_uid | 9589 |
|---|
| title | From resting fMRI to directed functional networks? |
|---|
| start_date | 2011/01/28 |
|---|
| schedule | 15h |
|---|
| online | no |
|---|
| details | invité par Sylvain Takerkart |
|---|
| summary | I will outline our recent work on the use of resting FMRI data to estimate brain networks. We
recently showed that ICA at low and medium levels of decomposition gives very similar gross
functional networks in resting FMRI data, and (separately) derived from a wide range of task FMRI experiments. We then proposed to use high-dimensional ICA to determine a much higher level of functional parcellation (~150 functional regions) and use the associated timecourses to estimate functional networks between these network "nodes". In order to learn what measures of connectivity are most appropriate, given functional MRI timeseries, we carried out comprehensive BOLD simulations, showing some surprising results for certain analysis methodologies. Finally, we show some initial results from applying the most successful network analysis methods to resting FMRI data, and discuss whether estimating directionality in these networks is possible from such data. |
|---|
| responsibles | Riehle |
|---|
| |
|