Posted by bob cox on November 07, 2002 at 08:55:22:
In Reply to: Re: Importing registered data back into afni posted by Edward J. Butterworth on November 06, 2002 at 17:41:18:
The distortions in echo-planar images are due to nonuniformities in the magnetic field, which are in turn caused by nonuniformities in the contents of the scanner (i.e., the subject). The best way to correct for this is to use a special pulse sequence to measure these magnetic field distortions, and then use that information during the image reconstruction to make non-distorted images.
However, this option is not a standard pulse sequence/reconstruction combination that the manufacturers supply. The result is that this "best" method is confined to sites that can implement these things themselves - sites with a strong MRI physics staff.
This state of affairs leads to research into other ways to un-distort EPIs when the magnetic field distortion data is unavailable. That's not so straightforward, since the needed information is missing, which explains why there is a plethora of papers on this problem. (I like alliteration.)
If you have a 3 T scanner and are doing 1282 EPI on patients with massive tumors, I can well believe you have serious distortions in the images.
bob cox