Hey AFNI Gurus!
We just got our Siemens Prism installed and the possibilities of simultaneous multi slice (SMS) / multi band (MB) imaging just opened up to us (SMS and MB is the same thing, right?). We are used to evaluate fMRI scan settings and comparing different coils using tSNR maps. If I remember correctly these tSNR maps should represent the average signal divided by the average noise between two time points (TRs). The tSNR description in AFNI just says that TSNR is calculated by the average signal divided by the standard deviation of the noise (which is the de-trended data).
Q1. Does the AFNI way of calculating tSNR equal using the average noise between two TRs, as described above?
Q2: If we want to use e.g. an SMS factor of 2 to lower our TR from 2 to 1 second, is it fair to use plain tSNR maps to compare this sequence to a non SMS sequence with a TR of 2 seconds? If we use a shorter TR the individual data points should be of lower “quality” but this is compensated by increased BOLD contrast due to the increase in data points (and faster sampling). Is the tSNR taking this into account? I.e. is a larger number of noisy data points going to give a lower tSNR than a smaller number of less noisy data points?
We just don’t want to accidentally dismiss an SMS sequence due to lower tSNR comapred to a standard non SMS sequence due to a missunderstanding of the mechanicns of the tSNR meassure. Maybe a lower sub cortical tSNR is expected but also an increased BOLD contrast.