Hi,
To start with, FIRST can really only reliably be used on images with contrast
like a T1-weighted image. However, we have found that even for these images
the segmentation of the ventricles is not robust, and consequently we have
removed this as an option from the latest release of FIRST. Hence I would
not recommend you using FIRST to segment the ventricles. If you are
interested in getting a ventricle segmentation then I would try using FAST
and a standard space mask to separate ventricular CSF from other CSF.
All the best,
Mark
On 22 Feb 2011, at 12:09, Chaarani Bader wrote:
>
> Hello FSL experts,
>
> i'm trying to calculate the volume of human lateral ventricles using different MRI sequences ( Flair, coronal 3D, T2*, diffusion ). For this purpose, i 'm running the following script on my different subjects :
>
> run_first_all -i (name of the subject).nii - L_Late,R_Late -o output
>
> then i calculate the ventricles volumes using fslstats on the generated "first_seg" image. The results are not very convincing because when viewing first_seg i notice that the ventricles are not 100% segmented (highlighted).
>
> I just want to know if it's right to use this script (that does a registration with a default MNI brain) and if anyone can propose a better way to calculate the volumes by adding options to the script or proceeding step by step instead of using run_first_all for example
>
> Thanks in advance,
>
> Bader
>
>
>
>
>
>
>
> --
>
|