 |
 |
The numbers entered as "Reference slice" correspond to the spatial position of the slice, not the temporal one. For example on SIEMENS scanners the default acquisition scheme is ascending interleaved with 2 4 6 ... 30 1 3 5 ... 29 in case of an even number of slices, here 30 1 3 5 ... 29 2 4 6 ... 28 in case of an odd number of slices, here 29 If you want to correct onto the temporally first slice you would thus go with "2" in the first case and "1" in the second. In both instances you should use a microtime onset of 1 during single-subject model specification (reflecting the fact that the data is aligned onto the beginning of the TR). In general, it should be better to use the temporally middle slice because the required adjustments are smaller overall, which should be more precise. You would have to go with either 30 or 1 in the first case or 29 in the second. The microtime onset should be set to 8 then (in case of the default 16 time bins as microtime resolution, reflecting the middle of the TR). If you adjust the microtime resolution according to the number of acquired volumes then you also have to adjust the microtime onset so that it corresponds to the middle time bin again.