Hi,
From SIENAX you can get a scaling factor - VSCALING - and this is the number that can be multiplied by the volumes from FIRST to get normalised volumes. You do not multiply by the whole brain volume.
For comparison, it is possible to compare un-normalised volumes, but it would be more advisable to compare the volumes that are corrected for head-size (unless you correct for this in the statistical test in some other way). If you don't do any correction then you run the risk that the difference is primarily driven by differences in head-size, which are known to strongly correlate with the volumes of subcortical structures.
All the best,
Mark
> On 18 Jan 2017, at 20:49, Angela Spirou <[log in to unmask]> wrote:
>
> Hi Eelke,
>
> Thank you for your input! I have a question about using SIENAX. The volume that we now have that we need to be normalized is that of the different subcortical structures. SIENAX gives normalized whole brain volume and not that of the structures specifically. In the thread you kindly instructed me to look at, it instructed one to take the output of SIENAX and multiply it to the volume that one has. Would this make sense though in this situation since it would be, for example, whole brain volume x volume of putamen?
>
> And just to double check, I cannot compare the FIRST data between groups with the output I currently have, thus needing to run SIENAX?
>
> Your help is very much so appreciated!
>
> -Angela
|