ΕΒΏμΌ® wrote:
> Dear William D. Penny.
>
> Thank you for your comment. As you commented, I tested with -1 (negative contrast), and I could
> get the almost same result with F-contrast. And, I have two more questions.
>
> Q1)Is it a possible T-contrast with [-1 0 0] or [1 0 0 ]? (my simplized auditory fMRI design is 3
> stimulation [ 500 hz, 4khz, and rest ])
>
This all depends on how you have specified and modelled 'rest'.
Scenario A:
If by, 'rest' you
mean the *mean* activity at each voxel (which is modelled automatically
by SPM) then the contrast
1 0 0
will show you voxels that increase their activity in response to
500Hz auditory stimulation. 1 0 -1 is not a useful F- or t-contrast, here.
Scenario B:
If, however, 'rest' is defined using some other criteria and you actually
have 4 regressors (500hz, 4khz, rest, mean) then I imagine you are interested in
the differential effect. The t-contrast
1 0 -1 0
gives voxels showing greater activity during 500Hz stimulation than `rest'.
The corresponding F-contrast shows voxels showing different responses to
stimulus than rest - could be bigger or smaller ...
> Q2) And, what is the meanning F-contrast [ 1 0 -1 ](order of 500hz, 4khz, rest), and is it
> possible contrast?
>
> Your comment will end my endless questions.
Hey, it was only 2 !
Will :-)
>
> Sincerely.
>
>
>
> ----- Original Message ----- From: "Will Penny" <[log in to unmask]>, To: "Woo Suk, Tae"
> <[log in to unmask]>, Date: 2005-09-09 19:34:23 Subject: Re: [SPM] the fMR result of
> F-contrast could be reported for paper?
>
> Dear Woo Suk Tae,
>
>
>
> F contrasts look for two-sided effects ie. they
>
> can detect positive or negative effects.
>
>
>
> When you did the t-test did you look for just a positive effect (ie
>
> +1 contrast vector) ?
>
>
>
> If so, then it may be that what you have found is a deactivation.
>
> Check this using a `-1' t-contrast.
>
>
>
> Best,
>
>
>
> Will.
>
>
>
>
>
>
>
> Woo Suk, Tae wrote:
>
>
>
>
>> Dear SPMers.
>
>
>> Recently, I got the strange results in auditory fMRI study abou monotopy mappng.
>
>
>> The results of volunteers seems to good in F-contrasts,
>
>
>> but in T-contrast, the results seems to be not adequate.
>
>
>> Now, I am confused whether I report it with F-contrast or not until the results will be good
>> with T-contrast.
>
>
>> Do you have any suggestions?
>
>
>
>
>> -----------------------------------------------------------
>
>
>> Woo Suk, Tae
>
>
>> Senior Researcher
>
>
>> Neuroimaging Laboratory. Neurology,
>
>
>> Samsung Medical Center, Seoul, Korea
>
>
>>
>
>
>> Tel: 82-2-3410-2738
>
>
>> Fax: 82-2-3410-2759
>
>
>> email: [log in to unmask]
>
>
>> -------------------------------------------------------------
>
>
>
>
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7475
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|