Hi, everyone,
I’m working on a set of old P files acquired using GE 1.5T PROBE. The protocol in our record showed TR/TE = 1500/35; NEX =8; CV4 (total number of scans) = 192.
When the data is loaded into Osprey, the number of averages and rawAverages are both 16. When I load the same P file into LCModel, The NS shown in the output is 128.
Does anyone know how the number of transients is read in the header of P files?
Thanks,
Steve
By default, the GE PROBE implementation sums transients across the phase cycle – so you’ll get 128 / NEX = 16 spectra stored in the P-file. It sounds like Osprey is reading this correctly, ie loading all the data in a consistent way, but maybe mis-reporting rawAverages (depending how exactly one defines that variable). This is probably only an issue for reporting, such as the final MRS-in-MRS table.
The fix isn’t entirely trivial (eg, multiplying by NEX or looking at a different header field) since the behaviour can change according to a few other settings: edited sequences usually disable this summation for example, and some of the reported dimensions include water reference scans. You could look at the Gannet reader, which handles most of the common cases correctly.
Sorry, I mis-read the total number of scans from your original message.
At this stage, I would be looking for a different tool to verify the number of scans. Do you know how old the data is (which software release), and do you have access to GE SAGE (on the scanner console)?