Structural PLS
Ansgar
Posted on 10/01/08 15:44:37
Number of posts: 2
Hi everybody,
I wondered whether it is statistically sound to run a structural PLS with MRs from different scanners. I would like to compare different patient groups. Each patient has one MR but unfortunately they are not all from the same scanner and don't have all the same field strenght. I know that for voxel-based morphometry this would be normally a "no- go" but was hoping that PLS can handle this :-)
Thanks so much,
Ansgar
(UC Berkeley)
Untitled Post
I'm Online
nlobaugh
Posted on 10/01/08 16:03:23
Number of posts: 229
Hello Ansgar..
Differences in resolution (voxel size), image intensity, contrast, etc. across MR machines and/or field strength make a very difficult situation for any voxel-based method, PLS included. When you add in differences due to clinical issues, it becomes even more challenging to sort out what differences are due to disease vs due to scanner properties.
Sorry... but PLS is not going to help much.
nancy