 |
 |
Dear FSL experts, I just wanted to run by some ideas to see if it sounds feasible. Currently, my project is interested in functional connectivity (seed or dual-regression) between a brainstem nuclei with cortical regions. According to the brainstem fMRI literature, it has been suggested that the conventional 5~6mm FWHM spatial smoothing is not feasible, due to relatively small size of the nuclei. It recommends around 3mm FWHM as increasing SNR and keeping the signal of interest at the same time. It goes on to say that If nonparametric test (randomise) is to be used, since Random Field Theory is not being used for assessing significance, using small kernel is acceptable as we don't have to worry about the size of RESEL. This implies that smaller spatial smoothing will also be applied to the cortical regions. Will 3mm FWHM make much difference from the conventional 5~6mm kernel? I am a bit cautious as we might not see as much increase in SNR. Let me know what you think, Thanks,