[MINC-users] validating nu_correct

[Jon Frederick]

I'm trying to prove to myself that nu_correct is working. My results are not 
encouraging. Can anyone suggest what I might be doing wrong?

My method was to introduce an artificial bias field using (slice number from 0 
to 255) * (1.0/256.0) + 0.5 to get a fractional number between 0.5 and 1.5, and 
multiplying the pixel intensity at each slice by that number. The bias was 
clearly visible in the sagittal and axial views (brighter in the frontal 
lobes). However, this did not visibly change after running nu_correct -stop 
.000001 -iterations 200

Here are my rawtominc and minctoraw steps (after stripping the headers):

dd conv=swab if=tempbias.raw | rawtominc -coronal -byte -unsigned -ystep 1 
-zstep 1 -xstep 1 -mri -origin 0 0 0 -range 0 255 -real_range 0 255 -clobber 
tempbias.mnc 256 256 256

minctoraw -range 0 255 -normalize tempcorr.mnc | dd conv=swab of=tempcorr.raw