help > Stripe artifacts in rest
Showing 1-2 of 2 posts
Dec 6, 2013 03:12 AM | benjamin sinclair
Stripe artifacts in rest
Hi All,
I have stripy artifacts in the slice direction in some of my resting data (see attached).
It only appears after removing covariates (any of WM, CSF, motion), and doesn't appear to be motion related (subjects have interscan motion < 0.2mm/TR).
Acquisition was interleaved on Siemens/Bruker 4T, however light/dark stripes are not alternating.
Has anybody seen such effects before?
Regards
Ben
I have stripy artifacts in the slice direction in some of my resting data (see attached).
It only appears after removing covariates (any of WM, CSF, motion), and doesn't appear to be motion related (subjects have interscan motion < 0.2mm/TR).
Acquisition was interleaved on Siemens/Bruker 4T, however light/dark stripes are not alternating.
Has anybody seen such effects before?
Regards
Ben
Dec 7, 2013 04:12 AM | Alfonso Nieto-Castanon - Boston University
RE: Stripe artifacts in rest
Hi Ben
Many different forms of artifacts may produce similar sorts of "stripe" effects. The first thing I would try would be to look for potential outlier scans. I would suggest using art (http://www.nitrc.org/projects/artifact_d...). Look at the "voxel-wise variability" or "voxel-wise SNR" displays (it probably displays the same sort of stripe patterns if you don't apply any threshold to the data) and play with the threshold options to see if scrubbing the data reduces/eliminates these effects. If it does, you may simply export the list of outlier scans into the CONN toolbox (as a first-level covariate) to remove those outlier scans from the analyses.
Hope this helps
Alfonso
Originally posted by benjamin sinclair:
Many different forms of artifacts may produce similar sorts of "stripe" effects. The first thing I would try would be to look for potential outlier scans. I would suggest using art (http://www.nitrc.org/projects/artifact_d...). Look at the "voxel-wise variability" or "voxel-wise SNR" displays (it probably displays the same sort of stripe patterns if you don't apply any threshold to the data) and play with the threshold options to see if scrubbing the data reduces/eliminates these effects. If it does, you may simply export the list of outlier scans into the CONN toolbox (as a first-level covariate) to remove those outlier scans from the analyses.
Hope this helps
Alfonso
Originally posted by benjamin sinclair:
Hi All,
I have stripy artifacts in the slice direction in some of my resting data (see attached).
It only appears after removing covariates (any of WM, CSF, motion), and doesn't appear to be motion related (subjects have interscan motion < 0.2mm/TR).
Acquisition was interleaved on Siemens/Bruker 4T, however light/dark stripes are not alternating.
Has anybody seen such effects before?
Regards
Ben
I have stripy artifacts in the slice direction in some of my resting data (see attached).
It only appears after removing covariates (any of WM, CSF, motion), and doesn't appear to be motion related (subjects have interscan motion < 0.2mm/TR).
Acquisition was interleaved on Siemens/Bruker 4T, however light/dark stripes are not alternating.
Has anybody seen such effects before?
Regards
Ben