Hello!
OK, as I promised that the last question would be the final one, I will have
to do some careful grammar.
A scanner from a given vendor, not having undergone hard- or software
updates and not having had a physicist play around with it, will run with
"standard" gradient tables for n directions, as cost minimized by the
algorithm most liked by the developers. Using exactly these tables or those
bvecs corrected for imaging grid rotation is then broad as long (for FA
calculation).
Regards...
Stefan
>Hi
>
>>This notwithstanding, I do understand you correctly, that if one uses a
>>given algorithm for cost minimization for any n number of gradients, the
>>resulting "standard" vectors are those one could use as "bvecs" for FA
>>calculation irrespective of any rotations caused by actual i.e. patient
>>positioning etc.?
>
>If I understand what you're saying correctly, then the answer is no!
>If you acquire some data with a given set of bvecs, these are the bvecs you
>need to use to fit a model to the data (eventually rotated if the imaging
grid has
>been rotated by the radiographer). You cannot use another set of (optimised or
>not) bvecs even if they have the same number of orientations.
>
>Cheers,
>Saad.
|