Dear SPMs,
I tried to realign 1 series of about 200 fmri images using SPM99b
(trilinear interpolation). Everything works fine, except that it was
terribly slow, about 3 hours on an Ultra 60's. I tried the same using SPM96
(bilinear interpolation) on the same platform and it took about 20
minutes!!.
>From my understanding, differences may be explain by the fact that
realignement in SPM96 is limited in 8 iterations, whereas SPM99b has to
reach a criteria with no limitations on iterations. I checked the results
for both and differences in movement estimation were very small
(fortunately I would say). Only for the last images in the series,
differences become systematic, though still small (which makes sense since
8 iterations may be a bit too small to realign the last images on the first
one).
May be I did something wrong. If no, realignement in SPM99b may be
difficult to use in practice for big analyses.
Did someone experienced similar problems? I would be happy to get get some
help and share comments. Thank you.
Etienne.
Etienne Koechlin, PhD.
---
Cognitive Neuroscience Section
National Institute of Health
MNB/NINDS/Bdg 10, Room 5c206
BETHESDA, MD 20892, USA
Tel: (301) 402-6390
Fax: (301) 480-2909
email: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|