What is the benefit when converting *.imas to SPM with visionToSPM ?
The basic idea of this tool is preserve all the information (orientation/position
of a volume scan too) that is contained in the raw image files (*.imas)
when the data is transformed to the analyse-format used by SPM. This can
make a coregistration of anatomical and functional data in SPM-preprocessing
dispensable:
If the experimental scanning-procedure ensures that the first functional
image is recorded immediately after the anatomical scan (MPRage, FLASH)
has finished, a perfect coregistration between anatomical and functional
data is obtained by transforming the orientation/position of the scans
to the SPM-world too and realigning all the functional images onto the
first functional scan.
When can I use visionToSPM ?
First at all you need to have recorded your functional data with an SIEMENS
Magnetom vision scanner. Image files must be in ACR-NEMA file format (*.ima),
not DICOM (this would be a nice extension to the tool. If you add it -
inform me please !).
There is no experience with other MR-scanner from SIEMENS at the moment.
But I suppose that even on a SIEMENS Symphony or something similar it
will not work without some adjustments.
Second the tool is tested only for a few sequences (MPRage, T1_se_fast,
EPI_mosaic). The basic problem is that there is no documentation of raw
image file format provided by SIEMENS. Thus the parameters needed by visionToSPM
may read from the wrong place in header, when using other sequences. In
this case you must enter some parameters manually (see below).
Installation
Make the directory visionToSPM in your matlab-working space and unzip
the files into this directory.
Add this directory to the path in matlab
Running the program
Let us assume you would like to transform an anatomical volume scan (MPRage)
recorded first and a functional session (EPI_mosaic) recorded subsequently.
We assume further that the subject has not moved his head between the
anatomical and the functional scan !
start the tool by typing: visionToSPM
A) conversion of the anatomical data:
Choose from select ima type the item normal (anatomical)
and enter a name for the output file. If the slices were scanned interleaved
select it (you MUST have used top to bottom interleaved mode !).
From the menu file choose select ima files.
To select the images select one of the image files and click OK. visionToSPM
selects all the files that are numbered continiously in the last number
of the filename. If the number of file visionToSPM founds is correct,
click yes. If not, check whether al the ima-files of the volume scan are
in the directory.
Check whether the scan-parameters given in the left frame are
correct. If not check enter Ima parameters manually and
enter those parameters which are wrong by your own (the image displayed
on the right is the one you selected before).
Choose the information that sould be transformed (bitmaps ad geometry,
bitmaps only or geometry only) and start the conversion. Three files named
as output-Filename.hdr, output-Filename.img and output-Filename.mat are
generated.
B) conversion of the functional data:
Choose from select ima type the item mosaic (EPI). If the
slices were scanned interleaved select it (you MUST have used top to bottom
interleaved mode !).
When converting the functional data the mat-file prefix
determines after which step of the preprocessing in SPM ( a=slicetiming,
ra=realigning, nra=normalization or snra=smoothing) the geometrical information
(orientation/position) obtained from the raw image files is added to the
flow of data-processing. In particular if you plan to use the spm_batch
tool this must be 'ra'.
Now, from the menu file choose select ima files.
To select the volume-scans of a session select one of the volume-scans
and click OK. If the imaType is mosaic, visionToSPM selects
all the files that are numbered continiously in the last two numbers of
the filename.
Again, check whether the scan-parameters given are correct and enter
them manually if not.
Start the conversion. Three files per volume-scan will
be generated, a *.hdr, a *.img and a ra*.mat file.
If you perform the preprocessing in SPM manually (not using spm_batch),
separate the ra*.mat files of the functional data from the *.hdr and *.img
files by moving them into a different directory before you start with
preprocessing in SPM. Now perform slice-timing and realigning (coregister
& reslice). After realigning copy the ra*.mat files generated by visionToSPM
into the directory of the ra*.img and ra*.hdr files.
A coregistration between anatomical and functional data is now obtained.
|
