Let me clarify: you can still look at relative changes between conditions
using the FEAT example. This may be all you want.
However, if you want to say that A is *deactivating* more than B, then I
believe you may want to confirm that both conditions are in fact
deactivations -- and this is where Featquery comes in... The subtraction can
obscure the direction of the main effects.
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Rebecca Leigh
Sent: Tuesday, August 11, 2009 3:26 AM
To: [log in to unmask]
Subject: Re: [FSL] within group deactivation
Thank you for your response David.
I have previously only used Featquery for ROI analysis - but I am unfamiliar
in how would I use it to examine whole brain activation - what file should I
use instead of a mask?
On Mon, 10 Aug 2009 10:56:53 -0400, David V. Smith <[log in to unmask]>
>On the example page
>C1 should give regions whose activation is greater in A compared to B
>whereas C2 will give the inverse. But, to say whether something is
>"activating" or "deactivating", you'll need to look at the parameter
>estimates (see Featquery).
>From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
>Of Rebecca Leigh
>Sent: Monday, August 10, 2009 10:42 AM
>To: [log in to unmask]
>Subject: [FSL] within group deactivation
>I want to run a two sample paired t-test, but looking for deactivation
>patterns and I am unsure where to put the -1 for this purpose.
>Any help is appreciated