Regarding DARTEL

Submitted by accyang on Mon, 03/12/2012 - 21:35

Hi All,

Suppose I have 20 subjects and split them into two groups (10 each). Instead of running DPARSFA with DARTEL on all 20 subjects at one time, I ran DPARSFA twice with DARTEL option for each group. Can obtained normalized images be still compared between groups?

Thanks

-Albert

Submitted by YAN Chao-Gan on Mon, 03/12/2012 - 22:12

I think they are not comparable since now you registered to two different group-specific template.
I don't know why you use DARTEL twice, probably you can also ask on SPM's maillist for confirming.

Submitted by accyang on Mon, 03/12/2012 - 22:46

Dear Dr. Yan,

Thanks for prompt reply. This is what I suspected and worried about. For a whole dataset with different groups (like disease or age group), can we still run DPARSF for all subjects at one time (using DARTEL option)? Is that reasonable to generating a DARTEL template based on a heterogeneous cohort? And what's the best solution of this (instead of conventional way). Sorry that the question may sound wiered but thanks for your advise in advance.

-Albert

Submitted by ZHANG_RESTadmin on Mon, 03/26/2012 - 11:15

Unless your patient cohort has lesion (tumor, etc.), putting all subjects into one preprocessing procedure is the common way.
For example you can check out Fair et al 2007, PNAS, Development of distinct control networks through segregation and integration.

Submitted by accyang on Mon, 03/12/2012 - 23:04

I found a question similar to mine in SPM maillist. Although it does not solve my question about the rationale of using DARTEL in heterogeneous groups, it looks like we need to input all subjects at once to ensure everyone is on the same DARTEL template.

This also means we need to run DPARSF every time for all when adding a new case. It's very time consuming anyway.

Link1: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1009&L=spm&P=R110&1=spm&9=A&J=on&d=No+Match%3BMatch%3BMatches&z=4

Link2: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1009&L=spm&P=R1312&1=spm&9=A&J=on&d=No+Match%3BMatch%3BMatches&z=4

Submitted by YAN Chao-Gan on Mon, 03/12/2012 - 23:14

I agree.
If you use two different group-specific templates, then the significant difference you found is difficult to attribute to the difference in templates or the true brain activity differences.

Submitted by cara on Thu, 03/15/2012 - 12:23

You have a very good site, well constructed and very interesting i have bookmarked you hopefully you keep posting new stuff.Fine information, thanks to the author. This work is really useful and significant.Resources like the one you mentioned here will be very useful to me! I will post a link to this page on my blog. I am sure my visitors will find that very useful
Pass4sure HP0-M46
Pass4sure CSSBB
Pass4sure 1Y0-A21
Pass4sure 7003.1
Pass4sure 350-040
Pass4sure E20-018