Hi,
On 13 Mar 2009, at 09:12, Rebecca Leigh wrote:
> Hi Steve,
> thank you for the reasurrance that my designs are ok.
>
> However, I am still confused about the contrast masking. As I
> mentioned
> previously I have read the manual and also other postings on the
> issue, but
> remain confused and was hoping you could clarify.
>
> For example I want to look at Pre (Left > Right) Vs Post (Left >
> Right). I
> therefore want to know where the Pre group had greater activation
> than the
> post group when looking at Left vs Right.
>
> so if I assume that if I have the contrasts
>
> C1 = Pre (left>right)
> C2 = Post (left>right)
> C3 = Pre>Post
> C4 = Post>pre
>
> would I contrast C3 with C1 AND C2? And then do the same for C4?
I presume you mean _mask_ C3 by C1 and C2? Yes, that's right, if that
gives you what you want - i.e. ensure that both C1 and C2 are positive
in order to help interpret the C3 result.
> I am unsure as to when I contrast with one or both masks.
> And presumably if I use the Z>0 I am only looking at positive
> activation?
> Is this correct?
Yes, though if C1 was already a contrast (L>R) I wouldn't call it
positive _activation_ though.
Cheers, Steve.
>
> Thanks in advance for any clarification and I again apologise for my
> basic
> understanding of analysis.
>
> Rebecca
>
> On Wed, 11 Mar 2009 06:37:15 +0000, Steve Smith
> <[log in to unmask]> wrote:
>
>> Hi,
>>
>> You can certainly run any contrasts between these groups with FEAT,
>> but I'm still not sure in what way it makes sense here to use the
>> controls, given that you didn't also give them the treatment. You can
>> run the following contrasts that make some sense:
>>
>> controls-patientsPRE (unpaired t-test model)
>> controls-patientsPOST (unpaired t-test model)
>> patientsPOST-PRE (paired t-test)
>>
>> See the FEAT manual for example models.
>> Hopefully this will give you the results you need?
>>
>> Cheers.
>>
>>
>>
>> On 10 Mar 2009, at 08:30, Rebecca Leigh wrote:
>>
>>> Hi Steve,
>>> sorry I should have been clearer about my groups.
>>>
>>> My controls are healthy volunteers, my training group consist of
>>> patients.
>>> I therefore want to examine how patients differ to healthy controls
>>> prior to
>>> training and how they differ after training.
>>> Do my FEAT analysis seem reasonable if this is what I want to
>>> examine?
>>>
>>
>>
>> --------------------------------------------------------------------------
> -
>> Stephen M. Smith, Professor of Biomedical Engineering
>> Associate Director, Oxford University FMRIB Centre
>>
>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
>> +44 (0) 1865 222726 (fax 222717)
>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
>> --------------------------------------------------------------------------
> -
>
>
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|