[MINC-development] N3-1.12 changed behaviour using a mask

[Claude LEPAGE]

Hi,

N3-1.10 had an effective damping of about 3.5e-05 at 1mm voxel for
a masked human brain in MNI space. 1.0e-7 might be a little small
as the suggested default value in N3-1.11 and 1.12. I don't remember
exactly the details leading to my estimation of the damping in the
old code. What you can do is run N3 at different values of damping
and plot the correction field (you can extract it if you look at
the options in nu_correct). If damping is too small, you will get
numerical instabilities (ill-conditioned matrix to invert). 

Claude

> Hi Claude,
>
> I have a follow-up question related to the lambda value. Do you know the 
> formula that we can use to compute the lambda value for the current 
> version of N3 which will give the same result as using 1.0e-7 in N3 1.10 
> given the other settings?
>
> Thanks,
> Matthijs