This describes the process of localising the actual voxel location used during the acquisition onto the participants FSPGR scan. The output is a binary voxel mask, in the space of the FSPGR. This can then be combined with outputs from segmentation (Fast) to calculate the tissue composition of the voxel (wm, gm, csf)
1) Copy native FSPGR dicoms
A copy of the native dicom files (dcm) is required as these contain the position information of the slices in the FSPGR. The step to create the nifti removes these, so first a copy of the FSPGR dicoms should be made, e.g from the directory of the scan data;cp Series_0000n FSPGR_dcm_folder
where n is the series number of the
FSPGR.
2) Create FSPGR nifti
Then you need to create the FSPGR: go into the Series_0000n folder that contains the FSPGR scan by typing;geprepfunct 172 1 [name of FSPGR].
3) Run VoxelOnFSPGR
Then, in Matlab, run the following function:
VoxelOnFSPGR(pfile,
ImagingDataDirectory, MRS_Rot_folder,... FSPGR_dcm_folder, FSPGR_nifti)
VoxelOnFSPGR is in the same location as Gannet, so ensure that you have Gannet in your matlab path for this to work (/cubric/software/Gannet/GannetExtras).
e.g.
VoxelOnFSPGR( ...
'~/mrs/somato_motor/data/3143_6_P12800.7', ...
'/cubric/home/sapje1/mrs/somato_motor/3143_081007-1', ...
'/cubric/home/sapje1/mrs/somato_motor/3143_081007-1', ...
'Series_00005', ...
'Series_00004_bak', ...
'fspgr.nii.gz')
Pfile = full
path to pfile (e.g.
'~/mrs/somato_motor/data/3143_6_P12800.7')
ImagingDirectory = the directory containing imaging data, unpacked by sortdicomball. The directory '/cubric/home/sapje1/mrs/somato_motor/3143_081007-1' in the above example should contain the Series_00001, ... directories and the nifti FSPGR.
MRS_Rot_folder = folder containing the
rotator/oblique localiser scan for the voxel, path from ImagingDirectory (e.g. ‘Series_00005’)
FSPGR_dcm_folder = folder containing the FSPGR native dicom files (e.g. Series_00004_bak')
FSPGR_nifti = the name of the FSPGR you
created (e.g. ‘fspgr.nii.gz’)
This will create a voxel mask, in the FSPGR space, for the MRS voxel and place it in a created directory named MRSVoxelReg.
The filename is
examnumber_series_Mask.nii.gz so e.g. 3999_08_Mask.nii.gz.
Below is a the output from VoxelOnFSPGR (e.g. 3999_08_Mask.nii.gz, but from an occipital voxel in this case) overlaid on the participant's FSPGR.
Below is a the output from VoxelOnFSPGR (e.g. 3999_08_Mask.nii.gz, but from an occipital voxel in this case) overlaid on the participant's FSPGR.