Print

Print


Dear Oshrat,

It is a bit unclear to me what exactly you want to test for / find out. 
Do I understand correctly that you want to compare pre- vs post swallow 
blocks between water and dry swallows? Anyway, my 2cts: with an external 
marker like EMG electrodes you are sure when somebody swallowed, with 
just a visual cue you are not 100% sure , and the delay between the cue 
and the swallow will vary. In the latter case you could get some clues, 
however, from the registration parameters; unfortunately, swallows will 
often show up quite clearly. This is also the reason that your data 
obtained during (and possibly some time after) swallowing will be messed 
up. One practical detail is to ensure that your subjects keep their chin 
up a bit, or at least not keep it too low; this makes it easier to 
swallow in a supine position without tilting your head.
As for the analysis, quite some studies have used a basic canonical hrf 
for looking at taste activation, but a FIR model is certainly a good 
idea. Why not compare basic hrf, hrf + derivatives and FIR?

Good luck,

Paul

Oshrat Sella wrote:
> <pre wrap>Hello
>
> I`m fairly new in the area of fMRI and while working on my research 
> design I
> had a few questions. I`m going to use event related design, with 2 
> types of
> swallowing tasks: water and dry swallows, randomized in order. The 
> tasks are
> going to be visually cued, and the time between each swallow is 45 sec. We
> will compare the brain activity pre and post swallowing treatment. We are
> going to use 3T scanner.
> My questions are regarding the analyses of the fMRI data. According to the
> literature some researchers used an external marker for the swallowing 
> event
> (surface EMG electrodes or pressure belt on the thyroid cartilage) and 
> some
> did not use any marker, and I guess that they used the timing of the 
> cues as
> the marker.I didn`t find any works that compared the use of external 
> markers
> and timing of the cues, so I`m struggling with this issue.
> Does anyone have experience with this kind of work? If so, what is your
> opinion about using or not using an external marker?
> From what I have understood the peak of the BOLD response happens 9-12 sec
> after the swallow, unlike other tasks where the peak happens after 6 sec.
> What model should I use in interpretation the data? we were thinking about
> the FIR model, and since we haven`t used that in our lab yet, any feedback
> can help.
> What are the Lags I should enter into the model, since there is 
> variation in
> the peak timing?
>
> Thanks for any suggestions,
> Oshrat Sella