On Sat, 30 Aug 2008 02:45:53 +0100, Christopher Benjamin
<[log in to unmask]> wrote:
>Hi John
>
>With the caveat that I've limited experience, I think it is best to keep all the
volumes cf
>power.
IMHO, power isn't the issue---if enough volumes are "bad" that it would
impact power, then there are so many tainted volumes that the data should
be discarded.
The reason volumes must be kept is that SPM views the data not as a series
of isolated observations, but as a timecourse. Censoring volumes would leave
holes in the data that SPM isn't set up to deal with.
>One way around the SOTs problem is to manually specify the regressor.
That
>way you can specify how ever many or few SOTs as you want, convolve
them with the
>canonical HRF (spm_hrf), and sample this time course at the time of ALL
scans. You
>would of course need to create a separate regressor for and model all
effects you would
>be expecting to influence blood flow - i.e., one regressor for 'accurate'
scans, another for
>'inaccurate' scans.
>
>Cheers
>
>Christopher
>
>
>On 30/8/08 6:17 AM, "<John> <Gelburg>" <[log in to unmask]> wrote:
>
>> Hi,
>>
>> I am creating my ER design with optseq. Suppose some volumes I don’t
want to
>> include in my analysis (for example, incorrect subject’ behavioral
>> response). Is it possible just to remove the volume and numbers of
>> consecutive events will be shifted? I would assume this to be problematic
>> from ER deconvolution point of view. Then, is it’s possible, what is the
>> correct way to do this.
>>
>> Thanks!
|