Sorry to clutter your mailboxes with more questions about nifti, but I
am trying to make sure that I am correctly understanding things. We
have a preprocessing script that was originally developed to work with
Analyze images, but some of our users have since used it with Nifti as
their output type. I am certain that the resulting orientation is
correct, but I want to make sure that I completely understand what is
happening when people go from analyze to nifti. Here is the stream:
1. Convert from DICOM to Analyze 3D using the LONI Debabeler
- this creates images stored in radiological convention, but they do
not have the customary negative X dimension value for FSL.
2. create 4D file
- when this was done using fsl 3 with analyze as the output type, the
files would automatically have the x dimension set to negative, and
thus header and data were in sync
- however, when output type is set to nifti, the resulting nifti
images have a positive sform and are classified by fslorient as
neurological; header and data are thus out of sync
3. do preprocessing and first-level analysis
- throughout these steps, the derivative nifti images retain the
neurological format
4. create registered images at 2nd level
- at this point, the resulting registered images now have a header
that specifies radiological storage. it would appear that data and
header are back in sync, assuming that no flipping occurred during
registration.
What I need to confirm is that the second-level analysis is NOT
flipping the actual data in the move from neurological to radiological
convention. I am pretty certain that this is the case (since the
motor cortex activation shows up on the correct side in the registered
maps) but I want to make absolutely sure that there is not something
else happening that I am missing.
thanks!
russ
---
Russell A. Poldrack, Ph.d.
Associate Professor
Wendell Jeffrey and Bernice Wenzel Term Chair in Behavioral Neuroscience
UCLA Department of Psychology
Franz Hall, Box 951563
Los Angeles, CA 90095-1563
phone: 310-794-1224
fax: 310-206-5895
email: [log in to unmask]
web: www.poldracklab.org
|