Hello Alfonso,

I have a task-fMRI experiment applying sparse temporal sampling (TR = 9 s, scanning time of a volume = 2 s). I want to do gPPI by CONN. I don' t know whether some options or parameters should be changed, compared to continuous sampling. (for example, to my knowledge, some papers suggested to elide "slice timing" step in pre-processing, when sparse sampling is used). Do you have some suggestions about it?

Many thanks in advance.

Best,
Arya

Hi Arya,

Yes, in CONN you may select in the Setup tab the option 'sparse acquisition' (instead of the default 'continuous acquisition') and that will take care of some of the largest issues (it will mainly skip steps that involve convolution with a canonical hrf). In addition to this you are also right that often slice-timing correction is skipped during preprocessing of these datasets, since sequential acquisitions may be too far apart for effective interpolation. Similarly, band-pass filtering during denoising may also be skipped, particularly for non-uniform sampling cases. 

Hope this helps
Alfonso
Originally posted by arya:

Thank you a lot for your kind reply, Alfonso! It is helpful. I am learning CONN and will try to do the gPPI in this way.

Best, 
Arya