If you take into account the volume of each voxel, then you should find
more similar results. The probable volume is from summing the
probability of each voxel bing a particular tissue class, times the
volume of a voxel.
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Marta Moraschi
Sent: Wednesday, January 31, 2007 12:54 PM
To: [log in to unmask]
Subject: [SPM] modulated images
Dear SPMers,
I encountered a problem using the SPM segmentation module in a DTI
study in
which i need to determine WM volume of each subject.
I segmented a T1-weighted volume saving both WM volume in native space
(c2Img.*) and in normalized space (mwc2Img.*), but modulated in
intensity.
I read that modulation is a process which multiplies the segmented
normalized volume of interest by the jacobian, to correct for effects of
volume change.
Given this, i determined the sum of all voxels' intensity for volume
both in
native space and in normalized modulated space, and i was expecting for
these two values to be near equal.
On the contrary, i found that the sum of all voxels' intensity for the
normalized modulated space is less (about a half) than that of native
space,
even setting all NaN of the modulated volume equal to 1.
Does anyone encounter the same problem? and which is the better value
(native or normalized modulated) for estimating WM volume?
Thank you for your time
Marta
|