Dear Hong,

The method in this paper were implemented in SPM96 but was dropped from subsequent versions as it was not as flexible as the other RFT methods (i.e. it only worked for Gaussian images).

SnPM, the nonparametric toolbox for SPM, has the ability to give you 'combining inferences', combining cluster size and height (as was offered in SPM96), but, even better 'cluster mass', the integral of the statistic values above the cluster-forming threshold. There is some work (Bullmore et al, 1999, IEEE TMI 18:32-42; Hayasaka & Nichols, 2004, NI 23:54-63) that suggests that the cluster mass test is more powerful than cluster size or joint-cluster-size-height. We're also working on a RFT version of cluster mass [ http://www.bepress.com/umichbiostat/paper76/ ] which should eventually be available as a SPM toolbox.

-Tom

On Feb 15, 2008 5:53 PM, Hong Gu <[log in to unmask]> wrote:

Hi,

I am new to SPM. Previously I use AFNI to analyze my fMRI data and when do
multiple comparison correction, I used AFNI's AlphaSim.

I found this paper from SPM group:
"Combining spatial extent and peak intensity to test for activations in
functional imaging", J-B. Poline et al, NeuroImage 5, 83-96.

Can some one tell me whether SPM implemented this method and which command I
should use or what document can I look at to use that command?

Thanks in advance,
Hong

--
____________________________________________
Thomas Nichols, PhD
Director, Modelling & Genetics
GlaxoSmithKline Clinical Imaging Centre

Senior Research Fellow
Oxford University FMRIB Centre