Thank you so much. Very helpful!
Any ideas, btw, how to do the same for contrasts, or am I asking too much?
Thank you
Misha
p.s. It seems that SPM subtracts the mean of regressed values SPM.Sess(1).U(n).P.P (P-mean(P)) and then uses _these_ values to arrive at the X's for the appropriate HRFs. Specifically, if we have
SPM.xX.X(:,i) corresponding to a trial with no regressor
>> SPM.Sess.U(4).P
ans =
P: [10x1 double]
name: 'Inv'
h: 1
i: [1 2]
>>
and another corresponding to the regressor P,
then
SPM.xX(:,j)/SPM.xX(:,i) for appropriate j's and i's is not simply, per each of ten rounds, the value in
SPM.Sess.U(4).P.P above, but rather the values SPM.Sess.U(4).P.P-mean(SPM.Sess.U(4).P.P).
Is there any specific reasoning for this? Is it an option?
p.p.s. Is there a good way to go between SPM.Sess.U(n) structures and the indices i in SPM.xX.X(:,i). It seems that there can be more than one event in each SPM.Sess.U(n), and thus the n's and i's don't match. Thank you!
James Lee wrote:
> Misha,
>
> To find the regressors used in the design matrix:
>
> disp(SPM.xX.X(:,i)), where i is the column of the design matrix
>
> To find trial onsets:
>
> disp(SPM.Sess(1).U(1).ons) for 1st run, 1st trial type
> disp(SPM.Sess(1).U(2).ons) for 1st run, 2st trial type
>
> And to find durations:
>
> disp(SPM.Sess(1).U(2).dur) for 1st run, 1st trial type etc.
>
> Hope this helps!
>
> Jim Lee
> Univ. of Utah
>
>
>
>
>
> On 2/22/10, Misha Koshelev <[log in to unmask]> wrote:
>> Dear All:
>>
>> I am very sorry to bother you. However, I had a (dumb) question regarding
>> the SPM.mat file that accompanies some beta images, and am not able to find
>> the answer unfortunately. Your help much appreciated!
>>
>> Specifically, I have a directory that contains:
>> (a) a beta image for each individual
>> (b) an SPM.mat file from the generation (?) of these Beta images
>>
>> I would like to use the SPM.mat file to extract:
>> (i) onsets/durations of events which have been regressed to create the Beta
>> values
>> (ii) whether the regression was on HRFs * constant or HRFs * some
>> round-dependent value
>>
>> I believe these should be in SPM.Sess fields (at least my understanding is
>> that this is where they are placed upon creation of the Beta maps), but
>> unfortunately there is neither an SPM.Sess _nor_ SPM.xBF structure in the
>> appropriate SPM.mat files.
>>
>> Thank you. Please reply to myself as well as the list if possible.
>> Misha
>>
>> p.s. I have also attached a very short (one page) document detailing my
>> understanding of how the beta maps are mathematically derived for at least
>> one multi-round economic exchange game in our laboratory. If there is
>> something incorrect or inaccurate in my understanding, I would very much
>> appreciate hearing about it. Thank you!
>> --
>> Misha Koshelev
>> MD/PhD Student
>>
>> Human Neuroimaging Laboratory
>> One Baylor Plaza
>> S104
>> Baylor College of Medicine
>> Houston, TX 77030
>>
--
Misha Koshelev
MD/PhD Student
Human Neuroimaging Laboratory
One Baylor Plaza
S104
Baylor College of Medicine
Houston, TX 77030
|