Dear SPM users,

It was a sparse temporal sampling for auditory fMRI experiment. Data was analysed with SPM12. A simple diagram of scanning is attached. Each trial was 9s and a volume was acquired in last 2s. I have 66 sounds and 67 images in one session. The number of images were one more than sound is due to a preceding scan before the first sound was presented. The design was specified in "scans". My questions are:

(1) when I specified onset of the first sound in first-level analysis, is it 1 or 2 (i.e. Should I regard the first scan that was not involved in analysis as 0 or 1)?

(2) I have tried Finite Impulse Response model (FIR) and canonical HRF model as basis function respectively. Is there any difference to specify onsets when I used FIR model or HRF model? If so, how to specify onsets in these two models?

Any help would be much appreciated !

Junxian Wang

School of Psychological and Cognitive Neurosciences

Peking University