Hi.
On Mon, 17 Dec 2001, David Glahn wrote:
> When using FEAT4 to do group analysis, spatially normalized cope images
> are used as inputs. However, these cope images seem to go through a
> number of corrections before use in the group model. Reviewing the
> ContrastMgr.cc code, it seems that both the corrections image and neff
> images are created. Why exactly are these corrections used and what is
> accomplished with them?
>
> If I were to take cope images as inputs into some other program for
> group analysis without performing these corrections, would the analysis
> be valid?
"corrections" (created by the linear modelling (film_gls), before the
contrast_mgr contrast program is run) is a list of the variance correction
matrix, used when contrast_mgr creates the varcope (variance of contrasts
of parameter estimates)
"neff1" etc _is_ created by contrast_mgr and is not used for anything (it
is the variance correction applied at each voxel and has historically been
outputted for our own research interest - we should probably stop
outputting it)
so - almost certainly all you would ever want in order to pass relevant
stuff into other programs (eg group stats) is the cope and varcope images
- ie it's all quite strightforward.
Hope this all makes sense! Thanks, Steve & MarkW
Stephen M. Smith
Head of Image Analysis, FMRIB
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|