open-discussion
open-discussion > RE: Eddy current correction
Mar 13, 2013 01:03 PM | Saad Jbabdi
RE: Eddy current correction
Hi - If you want to find out whether your data has eddy current
distortions, you need to look at the data (e.g. with a movie loop
after motion correction). Even the twice refocused sequence can
have some residual EC distortions, depending on the specifics of
your sequence/scanner.
Note that the new recommendation in FSL is to use "eddy", not "eddy_correct".
Eddy comes with a number of advantages:
- it has a model for the data, so can deal with high bvalues (doesn't use registration to the b0s to drive the correction)
- you can feed it the results of EPI distortion corrections (from the topup tool), so it can differentiate between the two types of distortions
- it properly rescales the signal using the Jacobian of the warps
- etc.etc.
Cheers
Saad
Originally posted by Paul Enggruber:
Note that the new recommendation in FSL is to use "eddy", not "eddy_correct".
Eddy comes with a number of advantages:
- it has a model for the data, so can deal with high bvalues (doesn't use registration to the b0s to drive the correction)
- you can feed it the results of EPI distortion corrections (from the topup tool), so it can differentiate between the two types of distortions
- it properly rescales the signal using the Jacobian of the warps
- etc.etc.
Cheers
Saad
Originally posted by Paul Enggruber:
Hey! Thanks for your replies.
Ok to make a general statement, it is always useful to make a motion correction, but eddy correction usually is not required.
In terms of FSL where both is made at the same time by "eddycorr" it would be more promising to skip that step and just use FLIRT to do the Motion correction between the different timepoints. Am I correct?
Ok to make a general statement, it is always useful to make a motion correction, but eddy correction usually is not required.
In terms of FSL where both is made at the same time by "eddycorr" it would be more promising to skip that step and just use FLIRT to do the Motion correction between the different timepoints. Am I correct?
Threaded View
| Title | Author | Date |
|---|---|---|
| Paul Enggruber | Mar 12, 2013 | |
| Niels Foit | Mar 13, 2013 | |
| Paul Enggruber | Mar 13, 2013 | |
| Simon Warfield | Mar 13, 2013 | |
| Paul Enggruber | Mar 13, 2013 | |
| Saad Jbabdi | Mar 13, 2013 | |