Hello Marzia,
If the stimulus is in the design then yes, it will also be removed - you may want to rerun feat with a cardiac-and-respiratory only design so that any stimulus signal is retained in res4d.
Kind Regards
Matthew
> On 16 Mar 2017, at 12:01, Marzia Restuccia S224515 <[log in to unmask]> wrote:
>
> Yes, I'm doing this via Feat. What about the stimulus? Will it be removed too?
>
> Il 2017-03-16 12:54 Matthew Webster ha scritto:
>> Hello Marzia,
>> If you are doing this via FEAT, then res4d + mean_func ( using
>> fslmaths ) will effectively give you the cleaned data, since fslmaths
>> will add mean_func onto each timepoint in the res4d image.
>> Kind Regards
>> Matthew
>>> On 16 Mar 2017, at 11:35, Marzia Restuccia <[log in to unmask]> wrote:
>>> Thank you Anderson. I don't succeed to understand how can I get a 4D "clean" dataset (I need this for a spectral analysis).
>>> I am running a GLM including 5 regressors for a task-based fMRI study.
>>> My model is Y = x_stim*B_stim + x_c1*B_c1 + x_c2*B_c2 + x_r1*B_r1 + x_r2*B_r2 + e, where:
>>> x_stim : stimulus time course convolved with HRF
>>> x_c1 and x_c2: cardiac regressors
>>> x_r1 and x_r2: respiratory regressors
>>> I would like now to get the "clean" data, and what I mean is the data from which I removed the contribution of the cardiac and respiratory regressors. In order to do this I applied the following formula:
>>> Yclean = Y - ( x_c1*.B_c1 + x_c2*.B_c2 + x_r1*.B_r1 + x_r2*.B_r2 )
>>> Anyway, from the comparison between the clean data and the model fitting I'm convinced that there's something wrong. Could you please tell me how would you do to reach this purpose?
>>> Thank you and sorry for bothering you!
|