Should we add choroid plexi to lateral ventricle volume?

1

We use Freesurfer v6 for volumetric analyses. I think people use the lateral ventricle volume as produced by FS, but looking at the segmentations I remember seeing that choroid plexi are a separate label. I am thinking to perhaps add the choroid volume to the lateral ventricles’ volume to obtain a more accurate volume. The reason is that I am not very sure we can say that those voxels are 100% choroid plexi but are rather affected also by partial volume. The geometry of the ventricle without the choroid plexi doesn’t make sense either, while it is nice and round when considering the plexi.

What are your thoughts on the matter? Do you add the plexi in lateral ventricle volumes, and why?

Thank you.

Example

Plexi
Plexi374×520 10.1 KB

segmentation from FS showing the plexi label:

2

Anyone has any suggestion?

3

Hi Dorian,

I have a related question instead of an answer. I am primarily using FSL, and never used Freesurfer. With Freesurfer, to be able to extract volumes (such as ventricle vol), do you have to run all of recon-all (took me ~2hr/nii), or is there a faster way to extract the volumes?

Thank You,
Lunky

4

I don’t have a good answer to your question. We run all recon steps typically. I remember that cortical segmentation goes the surface processing route, while subcortical is volume based. So I think cortical volumes will need almost all steps while subcortical volumes might be obtained with shortcuts, but not sure.