I ran sienax on a normal volunteer.
The FSL developers did a great job, thank you.
I have a few questions: I notice in the periph_render images that
some of the medial temporal cortex is not counted as cortex (I assume it
ends up in white matter).
The ventricle csf includes some extracerebral CSF along the midline
and in the superior cistern above the cerebellum (I would like to include
only lateral and 3rd ventricles).
A small part of the cerebellar vermis was cut off.
I thought of repeating the analysis adding the -a or -A option for
fast (currently running) to address this issue. What is the difference
between the -a and -A options in FAST in a few more words than "for
initialization and posteriors", better still what are initialisation and
posteriors, respectively?
Are there options I can use to correct these inaccuracies or do they
depend on the fact that my T1 weighted images are 2 mm thick? I have some
masks derived from the icbm 151 database (is this the same as the MNI 152
standard brain you use?) that include lateral ventricles only. Can I use
those with the -lm option in some way (although it looks like that option
works for gray matter)?
Concerning the sienax output:
1) are the measurement units of brain, white, gray etc. mm3?
2)I assume that tissue 1 and tissue 2 stand for grey and white matter
(non-normalized) volumes, but I am surprised that these volumes are almost
equal (689862 and 650991, respectively), while I thought white matter volume
would be larger than gray matter, generally (I may be wrong about this).
3) I was surprised to see that the product of the scaling factor by tissue1
and tissue2 was slightly inferior to grey and white, respectively. Why is
that?
Is there a way to run a series of subjects through sienax as a
batch?
Finally, I wonder if there is a way to obtain further sub-regional
volumetric measures (like basal ganglia, thalamus and hippocampal gray for
example....). Also, are the regional measures as accurate as the gray and
white matter measures you report in the 2002 Neuroimage paper (Smith et al.
NI 17, 479)?. Is the addition of these measures planned for future versions
or is there some hard limitation to the segmentation from T1 maps that makes
it impossible?
Stefano Marenco, MD
Senior Staff Fellow
NIMH/CBDB
10 Center Drive, Room 4S235
Bethesda, MD 20892
tel (301) 435-8964
fax (301) 451-4269
-----Original Message-----
From: Mark Jenkinson [mailto:[log in to unmask]]
Sent: Wednesday, August 27, 2003 1:01 PM
To: [log in to unmask]
Subject: Re: [FSL] ventricle volume
Hi,
Yes, there's a very easy way.
Just use the -r option in sienax and it will print out
the normalised ventricular volume as vcsf.
All the best,
Mark
Marenco, Stefano (NIH/NIMH) wrote:
> Is there an easy way to measure ventricle (NOT CSF) volume in two
> cohorts with the FSL tools (e.g. sienax)?
>
>
>
> Stefano Marenco, MD
>
> Senior Staff Fellow
>
> NIMH/CBDB
>
> 10 Center Drive, Room 4S235
>
> Bethesda, MD 20892
>
> tel (301) 435-8964
>
> fax (301) 451-4269
>
>
>
|