Dear experts,
We have recently switched our imaging protocol from a 64*64 matrix (i.e. 3.5mm isotropic voxel size) to a 96*128 matrix (2.5 mm isotropic voxel size, 15% slice gap). During preprocessing, we apply a non-linear spatial normalization to a scanner-specific EPI template, after which the images are resliced to a 2mm isotropic voxel size.
When checking our images after the normalization step, there appears to be high spatial frequency noise following a ‘checkerboard’ pattern, particularly evident on the edges of brain. This pattern is not obvious in the raw data. We’ve tried reslicing to different voxel sizes, but this didn’t seem to have any effects. If anybody could provide us with a hint on the cause and possible solutions of this problem, we would be very grateful.
Yours,
Rogier & Wim
________________________________________
Rogier B. Mars
F.C. Donders Centre for Cognitive Neuroimaging
& Nijmegen Institute for Cognition and Information
Adelbertusplein 1
6525 EK Nijmegen
The Netherlands
Phone: +31 (0)24 3610980
Fax: +31 (0)24 3610989
E-mail: [log in to unmask]
________________________________________
Wim A.C van de Riet
F.C. Donders Centre for Cognitive Neuroimaging
Adelbertusplein 1
6525 EK Nijmegen
The Netherlands
Phone: +31 (0)24 3668291
Fax: +31 (0)24 3610989
E-mail: [log in to unmask]
&
Tilburg University
Department of Cognitive and Affective Neuroscience
P.O.Box 90153
5000 LE Tilburg
The Netherlands
Phone: +31 13 4662339
E-mail: [log in to unmask]
|