Multi atlas bad naming diagnosis needed

NIRALUser / AutoSeg

AutoSeg is a C++ based application developed at UNC-NIRAL that performs automatic brain tissue classification and structural segmentation, including multi-atlas segmentation

http://www.nitrc.org/projects/autoseg

Apache License 2.0

4 stars 1 forks source link

Multi atlas bad naming diagnosis needed #3

Open styner opened 8 years ago

styner commented 8 years ago

the multi-atlas checking does not report to the user if there is an issue with missing labels, i.e. if there are no label files found the computing does not complain (but of course does not generate any correct output)

juanprietob commented 8 years ago

All files should have the same number of labels and if not labels found, it should notify the user? Should that be the behavior?

styner commented 8 years ago

the same number of labels or parcellations as atlases (one of the two). Also, it is multi-modality, there should be both t1 and t2 images for all cases.