27 Nov
2010
27 Nov
'10
5:34 p.m.
Hi all,
I'm using mri_compute_seg_overlap to calculate dice coefficients for two asegs, and the numbers look reasonable. However, for two of my comparisons, mri_compute_overlap reports zero overlapping voxels between the segmentations. Do these two commands compute overlapping voxels in a different way? Is there another reason why these two commands might report conflicting information?
Thanks much for any advice, Jeff