Hello,
Depending on your image size, the command may be doing a lot more calculation than required if you just want to make the point larger. You could alter the bounds of the roi option, to give a larger ( box ) ROI in voxels, or you could use the -dilM option rather than fmean, which should be faster but still use a spherical kernel.
Hope this helps,
Matthew
> Using the following command line I can define a point, e.g. the tip of an ICP catheter in situ:
>
> fslmaths <filename.nii.gz> -mul 0 -add 1 -roi x 1 y 1 z 1 0 1 <output_filename1> -odt float
>
> and then draw a ROI around this very point using this command line:
>
> fslmaths <output_filename1> -kernel sphere 5 -fmean -odt float
>
> The problem I have now is that this last procedure takes 10 minutes to calculate/save and creates a 40 MB . nii file when building a sphere, but only seconds for building a box.
>
> Does anyone have an idea what's going on here and how to
> change/optimize it?
>
> Ruling out any significant hardware issues, I tested this on my labtop (8 GB of RAM, i7) running Ubuntu 15.04 with fsl-complete from neuro.debian.net which has not shown any improvement in speed for creating the sphere.
>
> Thanks for your help.
>
> Patrick
|