help
help > RE: Smoothed and unsmoothed data
Apr 1, 2018 12:04 AM | Larry Lai
RE: Smoothed and unsmoothed data
Dear Experts
What's the rationale behind using the smoothed data as voxel time series, while the unsmoothed data as seed timeseries for statistical analyses?
Thanks!
Best
Larry
Originally posted by Alfonso Nieto-Castanon:
What's the rationale behind using the smoothed data as voxel time series, while the unsmoothed data as seed timeseries for statistical analyses?
Thanks!
Best
Larry
Originally posted by Alfonso Nieto-Castanon:
Hi
Kaylah,
Seed-to-voxel analyses will look at the correlation between a "seed timeseries" and a "voxel timeseries". If you use all default settings in CONN (including the default MNI-space preprocessing pipeline, which includes spatial smoothing as the last preprocessing step) then seed-to-voxel analyses will use:
1) smoothed data at each voxel, to define each "voxel timeseries"
and 2) unsmoothed data aggregated across all voxels within each seed ROI, to define each "seed timeseries"
Just to clarify, this behavior can be modified by the user. You may enter in Setup.Functional (dataset 0) whichever data that you want to use to define the "voxel timeseries" above; and you may specify in Setup.ROIs "from functional dataset ##" whichever data you want to use to define each "seed timeseries" above (even differently for each ROI if you want).
Hope this helps
Alfonso
Originally posted by Kaylah Curtis:
Seed-to-voxel analyses will look at the correlation between a "seed timeseries" and a "voxel timeseries". If you use all default settings in CONN (including the default MNI-space preprocessing pipeline, which includes spatial smoothing as the last preprocessing step) then seed-to-voxel analyses will use:
1) smoothed data at each voxel, to define each "voxel timeseries"
and 2) unsmoothed data aggregated across all voxels within each seed ROI, to define each "seed timeseries"
Just to clarify, this behavior can be modified by the user. You may enter in Setup.Functional (dataset 0) whichever data that you want to use to define the "voxel timeseries" above; and you may specify in Setup.ROIs "from functional dataset ##" whichever data you want to use to define each "seed timeseries" above (even differently for each ROI if you want).
Hope this helps
Alfonso
Originally posted by Kaylah Curtis:
Hi Alfonso,
We are trying to run seed-to-voxel analysis using an unsmoothed dataset. If we run the analysis under the default setup, with ROI-level BOLD signal extraction set to use unsmoothed data, will seed-to-voxel use smoothed or unsmoothed data?
Thanks for your help!
We are trying to run seed-to-voxel analysis using an unsmoothed dataset. If we run the analysis under the default setup, with ROI-level BOLD signal extraction set to use unsmoothed data, will seed-to-voxel use smoothed or unsmoothed data?
Thanks for your help!
Threaded View
| Title | Author | Date |
|---|---|---|
| Peter McColgan | Jan 30, 2015 | |
| Alfonso Nieto-Castanon | Feb 8, 2015 | |
| Kaylah Curtis | Nov 11, 2016 | |
| Alfonso Nieto-Castanon | Nov 12, 2016 | |
| Larry Lai | Apr 1, 2018 | |
| Peter McColgan | Feb 13, 2015 | |
| Alfonso Nieto-Castanon | Feb 18, 2015 | |