Hi,

I'm not sure what you are trying to normalise.
If it is something like subcortical structure volumes, then normalising by the ICV (that includes CSF) or NBV is very similar unless you have an unusual clinical population where the brain is much smaller, and more variable, than the ICV.

I hope this helps.
All the best,
Mark


On 26 Sep 2012, at 11:40, Patricia Pires <[log in to unmask]> wrote:

Hello Mark,

thanks for your answer. Then, NVB doesn't include CSF but i am wondering which value should i use in order to normalize the structures (e.g. WM). I had understood it is necessary to normalize through the three values ​​(gray, white and CSF). I send you an sienax photography in which has been segmented the three values​​. What I have done is from a formula which has been normalized by the total sum. It should therefore only normalize the sum of the gray and white matter?

Thanks again,

Patricia.

2012/9/25 Mark Jenkinson <[log in to unmask]>
Dear Patricia,

The brain volume is just GM + WM and does not include CSF.
As it is difficult to distinguish CSF from skull/scalp/membranes, depending on the contrast and fat saturation used in different acquisitions, it is not generally reliable to try and calculate ICV, and instead we calculate NBV, which excludes CSF.

All the best,
        Mark



On 25 Sep 2012, at 14:34, SUBSCRIBE FSL Patricia Pires <[log in to unmask]> wrote:

> Hello jiscmail,
>
> I'm doing sienax segmentation for white matter, gray matter and CSF and I'm normalizing these values to the sum of the three values ICV = GM + WM + CRF. Is it right to proceed in this way or is it better to use the value Estimated Normalised brain volume (NBV) to normalize?
>
> Thanks in advance,
>
> Patricia
>

<sienaxAxial.png>