Dear SPM-ers,
I am beginning to work with Tumor Patients and I would like to implement
the unified segmentation technique [from Crinion et al (2007)] as part of
the preprocessing steps for my tumor patients data, following which I can
analyze their functional (T2*) images.
What are the exact steps I must follow in order to implement the above
mentioned technique?
The following are the preprocessing steps (without segmentation) that I’ve
been following for my tumor data, up till now using SPM5:
1. Realign: Estimate + Reslice (Reslice options: All Images + Mean Images)
2. Coregister: Estimate (Reference: mean realigned functional ; Source: T1
anatomical image; Other images: resliced functional images)
3. Smooth (files to smooth: resliced functional images)
*[In all of the above steps, for the Estimation, Reslice options I have
retained the default SPM5 values. ]*
Q1: At what stage in this preprocessing pipeline must I include the
segment option? (After coregistration, is it not?)
Also, what set of files must I input for the segment option? The
functional T2*’s or the coregistered T1??
Q2: Must I follow up segmentation with normalization (Normalize:Write
option)??
Any suggestions regarding tumor data preprocessing would greatly help!
-Keerthi Ramanujan (Rookie SPM-er)
--
Keerthi Ramanujan
Project Assistant
SALLY (Speech and Language Laboratory)
National Brain Research Center
Manesar, India
[log in to unmask]
|