Thank you Mark.

After including the debug option I do not see many problems (some meningeal inclusions but nothing major). I do have a couple more quick questions though.

I was wondering if SIENAX corrects for signal intensity bias? If not, would you recommend I correct for this before running SIENAX? If so, it seems like I would have to enter the FAST corrected bet'd image into SIENAX (although with my data it seems like SIENAX works better when I include the -c x y z function rather than starting with a bet'd image).

I was also wondering if using the FAST corrected and bet'd image with the -S "-i 20" option would be the way to go (and not including the -c x y z option)?

Lastly, I was wondering if I was calculating the ICV correctly. For the values below my ICV would equal 1372475.44. Is this correct?

VSCALING 1.2707000446

tissue             volume    unnormalised-volume
GREY               617324.75 485814.69
WHITE              755150.69 594279.26
BRAIN              1372475.44 1080093.95

Cheers!

From: Mark Jenkinson <[log in to unmask]>
To: [log in to unmask]
Sent: Thu, September 9, 2010 2:00:59 AM
Subject: Re: [FSL] ICV and Age?

Hi,

The scaling derived by SIENAX is based on an affine registration to the standard
space that takes into account the skull estimates provided by BET. I would also
not expect to see a strong correlation between this and age, but it is possible that
the estimate of the skull is not very accurate, especially if there is strong atrophy
or unusual dura. If you run it with the debug option and look at the _brain_skull
images then it might give you an indication of whether there are any problems.
Note that the skull estimate is always noisy though, and it would only be a problem
if there were large areas that were coherently and incorrectly classified.

All the best,
Mark

On 7 Sep 2010, at 21:00, Forrest Johnson wrote:

> Hello FSL Experts,
>
> I have used SIENA (sienax <T1_image> -o <output_name> -B 88 85 149) to capture ICV in several subjects and have found an unexpected relationship between age and ICV (a strong inverse relationship with ICV decreasing as age increases). I typically would not expect to observe such a relationship (unless my older subjects have smaller heads), so I was wondering if my methodology was correct (sienax <T1_image> -o <output_name> -B 88 85 149)? A more likely scenario might be a decrease in brain volume with age, but I was under the impression that when I multiplied the total brain volume by the VSCALING factor, that I was capturing ICV. Is this correct?
>
> Some of the captured ICV values are listed below.
>
> Cheers!
>
> ICV Age
> 1519642.91 62.92
> 1372345.71 69.83
> 1566289.91 61.08
> 1381052.66 68.33
> 1498448.09 58.83
> 1437447.51 67
> 1500638.8 69.33
> 1413323.2 76.08
> 1668326.34 61.42
> 1626539.29 65.25
> 1520028.79 68.83
> 1439787.42 77.42
> 1367213.24 75.5
> 1462530.25 76.83
> 1392069.23 82.25
> 1391631.46 83.5
>