yes, you would convolve those since they relate to the neural process
that you are modeling, which would have its effect via the hemodynamic
response. Anytime you think the covariate is having its effect by
changing neural processing (as opposed to directly changing the MR
signal, e.g., by motion) then you should convolve.
cheers
russ
Soohyun Cho wrote:
> Thank you so much, Mark.
> That helps a lot.
>
> One last thing..
> I'd like to make sure that I understand what it means by the covariate
> being "related" to the stimulus.
> What about covariates that are 'related' to the stimulus, but are not of
> interest ?
> For example, let's say I have a memory task using words. If I want to
> regress out the effect of word concreteness (not of interest, but
> related to the stimulus) for each trial, should I convolve this
> regressor or not?
> Also, if I want to regress out the response time for each trial (not of
> interest, but related to the stimulus in that it reflects the cognitive
> difficulty of the trial), should I convolve this covariate or not?
>
> Looking forward to your answer.
> Thank you in advance.
>
> Soohyun.
>
>
> On Nov 19, 2007, at 11:53 PM, Mark Jenkinson wrote:
>
>> Hi,
>>
>> The "particular waveform" is the custom 1-entry waveform specified by
>> the file.
>> Usually you do want to convolve this with the HRF as the values in the
>> 1-entry
>> file often specify stimulus-related events. One example of this is a
>> single-event
>> stimulus where the 1's in the waveform represent single stimuli and
>> the remaining
>> values are 0. In this case you want to account for the haemodynamic
>> delay and
>> dispersion by convolving with the HRF in order for this model to fit
>> the data.
>>
>> An example of when you do not want to use convolution is when the
>> entries relate
>> to something other than stimuli - something that directly affects the
>> signal,
>> without being subject to haemodynamic processes. This is often the
>> case for
>> artefacts such as those caused by motion, respiratory or cardiac
>> processes. If
>> you have recordings of these and you want to regress them out of the
>> signal, then
>> you can use the custom 1-entry file but you should not convolve with
>> the HRF.
>>
>> I hope this makes things clearer.
>> All the best,
>> Mark
>>
>>
>>
>> On 20 Nov 2007, at 06:19, Soohyun Cho wrote:
>>
>>> Hello FSL experts,
>>>
>>> Can I get some clarification about some explanations in the feat manual?
>>>
>>> In the feat in detail user guide, it says:
>>>
>>> " For a single-event experiment with irregular timing for the
>>> stimulations, a custom file can be used.
>>> With Custom (1 entry per volume), you specify a single value for each
>>> timepoint.
>>> The custom file should be a raw text file, and should be a list of
>>> numbers, separated by spaces or newlines, with one number for each
>>> volume (after subtracting the number of deleted images).
>>> These numbers can either all be 0s and 1s, or can take a range of
>>> values. The former case would be appropriate if the same stimulus was
>>> applied at varying time points; the latter would be appropriate, for
>>> example, if recorded subject responses are to be inserted as an
>>> effect to be modelled.
>>> Note that it may or may not be appropriate to convolve this
>>> particular waveform with an HRF - in the case of single-event, it is."
>>>
>>> 1. what is "this particular waveform" referring to in the last
>>> sentence? (the latter? or both the former and the latter ?)
>>> 2. what does it mean by "in the case of single-event" in the last
>>> sentence?
>>> 3. can you give me some more examples of when it is appropriate to
>>> convolve a covariate with an HRF and when it is not?
>>>
>>> thank you for your help in advance,
>>>
>>> -Soohyun.
>>
>>
--
Russell A. Poldrack, Ph.d.
Associate Professor
Wendell Jeffrey and Bernice Wenzel Term Chair in Behavioral Neuroscience
UCLA Department of Psychology
Franz Hall, Box 951563
Los Angeles, CA 90095-1563
phone: 310-794-1224
fax: 310-206-5895
email: [log in to unmask]
web: www.poldracklab.org
|