help
help > RE: Realignment and Unwarp for Right-Left phase encoding
Nov 15, 2020 08:11 PM | Alfonso Nieto-Castanon - Boston University
RE: Realignment and Unwarp for Right-Left phase encoding
Hi Martin,
Yes, that sounds perfectly reasonable (and the effect of the additional interpolation step should be very small compared to the standard smoothing, but if you are concerned about that you could also use the 'functional Realignment without reslicing' option in CONN which is equivalent to SPM realign-estimate step)
And regarding the question about how to include the information of phase encoding direction into the distortion correction via the CONN GUI, CONN is going to look for that information directly in the field 'PhaseEncodingDirection' of your functional data sidecar .json file (when directly entering a VDM file, CONN assumes it represents the spatial distortions along the same encoding direction as your functional data).
Best
Alfonso
Originally posted by Martin Domin:
Yes, that sounds perfectly reasonable (and the effect of the additional interpolation step should be very small compared to the standard smoothing, but if you are concerned about that you could also use the 'functional Realignment without reslicing' option in CONN which is equivalent to SPM realign-estimate step)
And regarding the question about how to include the information of phase encoding direction into the distortion correction via the CONN GUI, CONN is going to look for that information directly in the field 'PhaseEncodingDirection' of your functional data sidecar .json file (when directly entering a VDM file, CONN assumes it represents the spatial distortions along the same encoding direction as your functional data).
Best
Alfonso
Originally posted by Martin Domin:
My solution:
Calculating and applying the VDMs via script externally and using the resulting unwarped time series as the base functional dataset in CONN. I changed the "indirect" pipeline by removing VDM calculation und Realign&Unwarp and inserted a simple Realign step. Not optimal, because of the additional interpolation, but perhaps the only reasonable way.
Best,
Martin
Calculating and applying the VDMs via script externally and using the resulting unwarped time series as the base functional dataset in CONN. I changed the "indirect" pipeline by removing VDM calculation und Realign&Unwarp and inserted a simple Realign step. Not optimal, because of the additional interpolation, but perhaps the only reasonable way.
Best,
Martin
Threaded View
| Title | Author | Date |
|---|---|---|
| Martin Domin | Nov 5, 2020 | |
| Alfonso Nieto-Castanon | Nov 6, 2020 | |
| Martin Domin | Nov 10, 2020 | |
| Martin Domin | Nov 12, 2020 | |
| Alfonso Nieto-Castanon | Nov 15, 2020 | |