 |
 |
The initial affine registration of SPM5 (and the new segmentation algorithm in SPM8) do not use template data such as the T1 average, T2 average etc. Instead, it aligns the tissue probability maps to the data, using a form of mutual information. The initial alignment is much more robust for the new SPM8 segmentation, as it uses tissue probability maps of various non-brain tissues. It also extends the old mutual information approach to incorporate a kind of non-parametric tissue classification. Best regards, -John > I am analyzing an elderly patient population (55+ yrs) of ALS and PLS > patients and Normal Volunteers and decided after the advice of many > individuals to create my own template. I used the VBM toolbox in SPM2 to > create this template. I am using SPM5 to carry out the segmentation steps. > How can I set SPM5 to use the custom T1 template that was generated by the > SPM2 VBM toolbox? I was able to set the priors to the grey, white matter, > and csf custom priors, but in the Affine Regularisation option of the > segment job I cannot choose my custom T1 template. Do I need to use this > Affine Regularisation option? Will someone please inform me on which setting > should I choose and an explanation why that setting is appropriate- Average > brain size, no affine transformation, no affine regularization, ICBM > European brains? -- John Ashburner <[log in to unmask]>