Hi Mark,
On Tue, 26 Mar 2013 22:02:32 +0000, Mark Jenkinson <[log in to unmask]> wrote:
>Hi,
>
>Your image seems to have the Superior-Inferior axis along the voxelwise y-axis (look at this in FSLView if you are not aware of this). The default for the main field axis in b0calc is the z-axis, and unless you have a very strange scanner, the main field direction needs to be the Superior-Inferior direction.
>
Yes, I did notice the slight differences but didn't pay too much attention originally... Anyway thanks for the tip and after changing the orientation, the results are more reasonable-looking now.
>Also, once you've done this you will need to worry about matching shim terms (linear and quadratic fields) in order to get a field which is a good match for a real, experimental fieldmap.
>
Where do I use these shim terms??
Cheers,
Louis.
>All the best,
> Mark
>
>
>On 25 Mar 2013, at 16:02, Louis Shue <[log in to unmask]> wrote:
>
>> Hi,
>>
>> I have been trying to compute the fieldmap for EPI correction using the command b0calc. The following are the commands I used (binaries are from neurodebian repository). The skull+tissue vs air segmentation image (EM_Map) was created from a T1-weighted MRI image.
>>
>> fsl5.0-b0calc -i EM_Map_mask -o EM_Map_b0field --b0=3
>> fsl5.0-fslmaths EM_Map_b0field -mul 42575000 -mul 6.28318 EM_Map_b0field
>>
>> Here are the links to the files in question. I have omitted the file $BMAP because of its size. EM_Map_fmap.nii.gz was extracted from EM_Map_b0field using a brain mask.
>>
>> http://dl.dropbox.com/u/8460189/EM_Map_mask.nii.gz
>> http://dl.dropbox.com/u/8460189/EM_Map_b0field.nii.gz
>> http://dl.dropbox.com/u/8460189/EM_Map_fmap.nii.gz
>>
>> However it seems that I must have made a mistake somewhere because in the final image, the light and dark areas don't appear to be in the right places? This is most obvious when comparing against the fieldmaps in for example Poyton et al, "Fieldmap-Free Retrospective Registration and Distortion Correction for EPI-Based Functional Imaging", 2009.
>>
>> Any pointers are very much appreciated!
>>
>> Cheers,
>> Louis.
|