[MINC-users] mnc2nii

[Andrew Janke]

On 10/15/07, Soren Christensen <sorench at gmail.com> wrote:
> Hi,
>  I noticed that FSL 4.0 complains over the .nii files created by mnc2nii.
> ++ WARNING: nifti_read_buffer(tmp.nii):
>    data bytes needed = 2097152
>    data bytes input  = 2097148
>    number missing    = 4 (set to 0)
>
> Apparently it is triggered by the use of a new nifti library in the
> FSL package. Does anyone know if there is a bug (in FSL or mnc2nii)
> and if it has any consequences (a pixel or 2 are nulled pherhaps)?

Hi Soren,

Yes this is correct from what I understand. The new version of MINC
(in the 2.X chain) fixes this by linking with the new niftilib as
well.

The simple fix is to use .hdr/.img files instead of a .nii from what I
understand it is because of a ambiguity in the spec WRT NifTI/Analyze.


a