[Homer-users] A few very basic questions about HRF timing and deconvolution

[Ng Eric]

Hello,

The hemodynamic signals are something quite new to me. I am grateful if you
can offer me some help on understanding the HRF timing in Homer:

1) What does the HRF timing stands for? At the first glance the pre and post
Time reminded me of the idea of baseline and epoching in ERPs, and it is
said in the user manual that the post time epoch would be compared to the
pre-time epoch for statistical results. But after reading some fundamental
concepts of fMRI i began to think whether that describes the convolved HRF
function... I searched the archive and found something relevant, saying that
there should be no overlapping between the HRF timing epochs, so I assume
it's more likely to be the former meaning?

2) We ran a blocked-design-like finger tapping experiment. Participant
tapped after seeing a visual cue. In one NIRS file there were 6 conditions
(different tapping patterns), each with 6 trials, the SOA between two cues
was 5 seconds. Then there would be 20 seconds of resting period between each
condition. In this case, how should we tag our cues? Would it be problematic
if we set up a trigger for each cue, which would only be about 5 seconds
after the previous cue?

3) Would deconvolution be more beneficial than a block average in 2)? Is
deconvolution dependent on the HRF timing, too?

4) I encounter the out of memory problem quite often... are there any
specific reasons (e.g. too many stimuli/sessions)?

Thank you very much!

Best regards,
Eric

Psychology Department,
National University of Singapore
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/homer-users/attachments/20090318/998e3f17/attachment.html