Hi,
It looks like you are using MATLAB functions to convert the coordinates, which are not part of FSL. It is therefore likely that these functions are not compatible in some way, or that you are not comparing things correctly. If you are still worried that something is wrong then we would need the full details about how you are calling img2imgcoord (and for MNI images you may need img2stdcoord or std2imgcoord instead). You can also check the relationship between voxel and mm coordinates within a standard space image alone using FSLView (which shows both coordinates).
All the best,
Mark
On 5 Jun 2014, at 14:19, xuemei <[log in to unmask]> wrote:
> Hi,
> Many thanks for your reply.
>
> In order to verify its correctness, I drew a spherical ROI in MNI standard space.
>
> Then I read the ROI and converted the voxel coordinates into mm coordinates (using ind2sub and vox2mni). That is to say, these mm coordinates are (x’,y’,z’) coordinates in the reference image.
>
> Next, img2imgcoord command was called to acquire corresponding native space mm coordinates, which correspond to (x,y,z) coordinates in the original image .
>
> In order to see if the (x,y,z) is corresponding to (x’,y’,z’), I converted these mm coordinates into voxel coordinates (using mni2vox and sub2ind) and wrote it as ROI.
>
> Theoretically, the two ROIs should be similar with each other both in shape and location. However, the two ROIs are heterochiral when opened with MRIcro.
>
> I don’t know why this hanppens.
> Hope for your help.
>
> xuemei
|