[MINC-users] rawtominc & nu_correct difficulties

[Andrew Janke]

Hi Michel,

nu_correct is very picky about ranges within data. Is the data
perchance all squashed into a few bits of precision?

I also seem to remember that it prefers unsigned short data from some
reason but I could be wrong on that.  Given that the ranges are going
to be rescaled anyhow, I would run it through something like mincnorm
with an output range of 0 - 100 to ensure that you have no restriction
of range effects.

send me a reply if you dont have a copy of mincnorm.


a

On 8/1/07, Audette, Michel <Michel.Audette at medizin.uni-leipzig.de> wrote:
> Dear folks,
>
> subsequent to yesterday's discussion, I'm trying to run nu_correct on a volume converted from tiff to raw to minc, as the volume has significant intensity inhomogeneity, but there is apparently something that I've neglected to do in the conversion. See below, where I've used some of Vladimir Fonov's parameters.
>
> maudette at icaw164201:~/data/ENTdata> cat HumanS16885_Stack.gray | rawtominc -clobber -byte -range 0 255 -orange 0 255 -real_range 0 255 -unsigned -zyx HumanS16885_Stack2.mnc 340 366 489
> maudette at icaw164201:~/data/ENTdata> nu_correct  HumanS16885_Stack2.mnc HumanS16885_Stack2_nuc.mnc
> nu_estimate_np_and_em: crashed while running volume_stats (termination status=11)
> nu_correct: crashed while running nu_estimate_np_and_em (termination status=65280)
>
> Can anyone spot the missing parameter? The program nu_correct otherwise runs fine on other MINC volumes, which have been converted from DICOM.