I would like to ask and clarify some doubts regarding the processing pipeline of CAT12 for volume estimations.
I've used the CAT 12.7 version.
CAT12 outputted higher WM volume estimations for gray matter regions from the CoBra Atlas, and in the manual I didn't find any information on what to do with these estimations.
If CAT12 outputs a volume of 0.2 of GM and 1.2 of WM, for a known GM region, how should one interpret this result?
Are the WMV to be discarded or should be taken into consideration?
Also, can this result be associated with atlas/subject misalignment or incorrect voxel tissue classification? Is it possible to track the error?
Also, if one decides to consider only the GM estimations portions, because it knowns apriori that region X is suppose to be only GM, arent' the volumes estimations biased because they represent a random portion of voxels that CAT12 actually managed to identify correctly as GM?
In my specific case, for example, CAT12 estimates globus pallidus almost always as a majority of WMV, but in CoBra atlas they say it's a GM region.
Also, I would expect that the results outputted from CAT12 wouldn't require any manual inspection as there is no warning about this in the manual. If CAT12 outputs WM for a region I would expect that this region would contain white matter.
If someone who doesn't know anything about the brain uses the software, they won't know that the WMV estimates outputted by CAT12 are "not supposed to be used or are wrong" because it won't know that region X is supposed to be WM or GM. Would be useful to add some information about this in the manual.
Does CAT12 have issues, that are known, in properly segmenting deep brain nuclei structures?
I would like to ask and clarify some doubts regarding the processing pipeline of CAT12 for volume estimations.
I've used the CAT 12.7 version.
CAT12 outputted higher WM volume estimations for gray matter regions from the CoBra Atlas, and in the manual I didn't find any information on what to do with these estimations.
If CAT12 outputs a volume of 0.2 of GM and 1.2 of WM, for a known GM region, how should one interpret this result? Are the WMV to be discarded or should be taken into consideration? Also, can this result be associated with atlas/subject misalignment or incorrect voxel tissue classification? Is it possible to track the error?
Also, if one decides to consider only the GM estimations portions, because it knowns apriori that region X is suppose to be only GM, arent' the volumes estimations biased because they represent a random portion of voxels that CAT12 actually managed to identify correctly as GM?
In my specific case, for example, CAT12 estimates globus pallidus almost always as a majority of WMV, but in CoBra atlas they say it's a GM region.
Also, I would expect that the results outputted from CAT12 wouldn't require any manual inspection as there is no warning about this in the manual. If CAT12 outputs WM for a region I would expect that this region would contain white matter. If someone who doesn't know anything about the brain uses the software, they won't know that the WMV estimates outputted by CAT12 are "not supposed to be used or are wrong" because it won't know that region X is supposed to be WM or GM. Would be useful to add some information about this in the manual.
Does CAT12 have issues, that are known, in properly segmenting deep brain nuclei structures?