I would suggest that you estimate the spatial normalization parameters via the Segment button. It works rather better than the Normalise button (although you'll need to write the spatially normalized data via the Normalise button).
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Jay Hegdé
Sent: Monday, June 18, 2007 5:57 AM
To: [log in to unmask]
Subject: [SPM] Image distortions during normalization
Hi SPMers,
I'm pretty new to SPM. I have a question about normalization artifacts.
I have T1 and functional data from a normal adult female subject
collected in a Siemens 3T Trio. When I normalize the data (in SPM5), I get
weird distortions: the normalized brain looks pancaked (please see
attached). The T1 images (not attached) are also distorted in a very
similar (i.e., coregistered) fashion after normalization.
All other data sets I've tried normalize without a hitch.
So my questions are:
(1) Does anyone know why these distortions happen?
(2) Are there ways to avoid such distortions?
Thank you very much,
Jay Hegdé
U. Minnesota
Minneapolis, USA
P.S. The following are my normalization settings, after the batch scripts of
Dr. Steve Fix of Northwestern U.:
-----------------------------------------------------------
T1 settings:
jobs{5}.spatial{1}.normalise{1}.write.roptions.preserve = 0;
jobs{5}.spatial{1}.normalise{1}.write.roptions.bb = [-78 -112 -50; 78 76 85];
jobs{5}.spatial{1}.normalise{1}.write.roptions.vox = [1 1 1];
jobs{5}.spatial{1}.normalise{1}.write.roptions.interp = 1;
jobs{5}.spatial{1}.normalise{1}.write.roptions.wrap = [0 0 0];
-----------------------------------------------------------
%Functionals settings:
jobs{5}.spatial{1}.normalise{2}.write.roptions.preserve = 0;
jobs{5}.spatial{1}.normalise{2}.write.roptions.bb = [-78 -112 -50; 78 76 85];
jobs{5}.spatial{1}.normalise{2}.write.roptions.vox = [3 3 3];
jobs{5}.spatial{1}.normalise{2}.write.roptions.interp = 1;
jobs{5}.spatial{1}.normalise{2}.write.roptions.wrap = [0 0 0];
-----------------------------------------------------------
