Hi,
If you want to test differences in amplitude _or_ shape then the way to do that is with the F-test at the higher level, combining over all 22 FIR estimates. If you only want to test for amplitude, then you need to calculate the amplitude from this set in a way that is independent of shape and then pass the calculated value (in the form of an image per subject) into a higher level analysis using randomise (as the calculation of the amplitude will not be linear). One easy calculation that you can do is just using the maximum value across the 22 values (voxelwise, using fslmaths).
To investigate shape you will want to look at the 22 values plotted against time. If you do an fslmerge of all the 22 values in order then you can use FSLview to plot the timeseries interactively at each voxel. If you then want to obtain a mean over some ROI then you can use fslmeants for this.
I hope this helps.
All the best,
Mark
> On 10 Apr 2017, at 16:21, Pavel Hok <[log in to unmask]> wrote:
>
> Hello again,
>
> just a few corrections of wrong numbers: ON is 30 s and the shortest
> OFF is 15 s (not 18 s). The FIR setting was also wrong. The window
> should be 45 s and there are 18 EVs. Sorry for any confusion.
>
> I would still appreciate any comment on the higher-level analysis.
>
> Cheers
>
> Pavel
>
> On 10 April 2017 at 14:10, Pavel Hok <[log in to unmask]> wrote:
>> Dear FSL experts,
>>
>> I could not find an appropriate answer in previous posts. I am trying
>> to analyse a block design dataset with long epochs (30 s ON, at least
>> 18 s OFF, TR 2.5 s, no jitter with respect to TR). We know already
>> that the response does not follow the canonical HRF. I therefore
>> created a single subject design with 22 FIR basis functions with
>> window 55 s.
>>
>> Having a single contrast for activation, I ended up with 22 copes per
>> subject. Just for the record, each subject had two such runs, so I
>> averaged the copes in a middle-level FE analysis. Now, is there any
>> other way to perform group analysis other than feeding all the cope
>> images to the group analysis, making a contrast for average per bin,
>> and F-test across all contrasts?
>>
>> I wish to know a) where the brain activates at any timepoint (F-test)
>> and b) what is the shape of the response (single COPE or PE?).
>>
>> Thanks for any help!
>>
>> Cheers
>>
>> Pavel
>>
>>
>> --
>> -------------------------------------------------------------------------------
>> -- MUDr. Pavel Hok
>> -------------------------------------------------------------------------------
>> -- Laboratoř funkční magnetické rezonance
>> -- Neurologická klinika
>> -- Lékařská fakulta
>> -- Univerzita Palackého v Olomouci
>> -- Fakultní nemocnice Olomouc
>> -------------------------------------------------------------------------------
>> -- Laboratory of functional magnetic resonance imaging
>> -- Department of Neurology
>> -- Faculty of Medicine and Dentistry
>> -- Palacky University Olomouc
>> -- University Hospital Olomouc
>> -- Czech Republic
>> -------------------------------------------------------------------------------
>> -- I.P. Pavlova 6, 775 20 Olomouc
>> -- web: fmri.upol.cz
>> -- tel.: +420 588 443 418
>> -- e-mail: [log in to unmask]
>> -------------------------------------------------------------------------------
>
>
>
> --
> -------------------------------------------------------------------------------
> -- MUDr. Pavel Hok
> -------------------------------------------------------------------------------
> -- Laboratoř funkční magnetické rezonance
> -- Neurologická klinika
> -- Lékařská fakulta
> -- Univerzita Palackého v Olomouci
> -- Fakultní nemocnice Olomouc
> -------------------------------------------------------------------------------
> -- Laboratory of functional magnetic resonance imaging
> -- Department of Neurology
> -- Faculty of Medicine and Dentistry
> -- Palacky University Olomouc
> -- University Hospital Olomouc
> -- Czech Republic
> -------------------------------------------------------------------------------
> -- I.P. Pavlova 6, 775 20 Olomouc
> -- web: fmri.upol.cz
> -- tel.: +420 588 443 418
> -- e-mail: [log in to unmask]
> -------------------------------------------------------------------------------
|