Dear SPMers,
I am very interested in using the recently-developed option to use multispectral data (T1, T2 and Flair) to segment MRIs. I have carefully read SPM12 manual, searched through this email list, and performed a few tests, but some points remain unclear at this point.
Notably, I am unsure what I should enter as the number of Gaussians for each tissue compartment. Is the default (2, 2, 2, 3, 4, 2) acceptable or should it be adapted when using multispectral data? As anyone looked at this specifically? I could not find papers about this.
My second question is related to this point too. I study elder participants, and some of them have (vascular-related) white matter lesions (hyperintensities on T2 and FLAIR, sometimes even showing hypointensities on T1). Should the number of gaussians entered for WM tissue be increased to enable SPM to classify these voxels as WM?
Thanks for your help,
Best
Renaud
|