Hi,
I'm trying to use fslsplit/fslmerge to separate different echoes from a multi-echo dataset, but am having issues with the resulting header. The TR of my data is 2.2s. When I run fslhd on the raw data, I get pixdim4 = 2200.0000000000, time_units = ms, which is correct. However, when I split the data across time using fslsplit, I get pixdim4 = 2.2000000477, time_units = s. The added .0000000477 is causing errors for downstream scripts that use fslval to read TR from the header of the data. This altered TR remains after using fslmerge to re-concatenate the data, even if I specify the TR as 2.2 or 2.2000000000 using the -tr flag. Any idea what's going on or whether there's a fix for this? thanks,
Ben
|