Hi Dan
how do you do the B1 correction? And why would we need to know the field strength? The T1 is certainly a function of the field strength, but we will just find the T1 for whatever strength field the input data was acquired with.
And yes, we search in 5ms increments. You could certainly change that if you want (it's a constant in the code), but I don't think we have the sensitivity to distinguish 5ms differences in T1 from typical input data.
The bigger problem is getting stable estimates of T2* from a max echo time of <15ms. The maps we have generated are *really* noisy. Note that by default we don't remove the T2* component from the PD unless you specify -correct on the command line, in which case we will write out a PDcorrected.mgz as well that is divided by the T2* decay term.
I guess in the end I'm not clear what problem you are trying to solve. If the discretization only has a small effect on the estimates, do you need to fix anything? Bruce
On Thu, 25 Apr 2013, dgw wrote:
I have run mri_ms_fitparms (version 5.1.0) on phantom data from two multiecho flash volumes (5 and 20 degrees with 8 echos: 1.67ms - 14.83ms in 1.88ms steps) at 7T utilizing a b1 map for correction.
How does mri_ms_fitparms know the field strength (I can't figure out where the information is when I read the mgz files using MRIread.m)?
I notice that the T1 volume seems to have graded steps of 5ms. Is this due to the table of 1ms stepped T1 values in the table described in Fischl et al. 2004 (and the data in this case being collected at 7T)?
Is there any way to increase the sensitivity?
Interestingly, this effect seems to only have a small effect on the values of the PD.mgz; however, from the 2004 Fischl paper, I would have expected it to have a large effect.
What concerns should I have about using the PD.mgz to represent the Equilibrium Magnetization in terms of accuracy of the values?
Thank You, Dan