Hello everybody,
I would be interested in your thoughts on the following issue:
The Juelich Histological Atlas is useful when it comes to delineating the amygdala and its subcomponents in fMRI. However, there seems to be partially a low correspondence between this atlas and the Harvard-Oxford Cortical-Subcortical Atlas, in particular regarding more exterior portions which are sometimes allocated to the amygdala in the Juelich Atlas (e.g., >80% probability basolateral amygdala), but to white matter in the Harvard-Oxford atlas (e.g., >80%). I wonder if the Juelich Atlas still could be trusted in this case and what the causes of the discrepancies could be. Furthermore, it seems hard to delineate the centromedial amygdala at all (as definded via the Juelich Atlas) by using the Harvard-Oxford-Atlas instead (since the probabilities of being amygdala are there very low).
Thanks, Stefan
Hy Stefan,
why don't you just ask the authors/editors of these atlases directly?
Sincerely Yours
Stefan
Hi Stefan,
I already did, but I am interested in different perspectives or experiences. If I receive useful information I will post it here since this is probably an interesting matter for many.
Best, Stefan
One major factor seems to be the different methods used to delineate brain structures. The Juelich Atlas is based on cytoarchitectonic/microscopic data, whereas the Harvard-Oxford-Atlas is based on segmentation of T1-weighted MR images, where it might be harder to delineate small nuclei.
Thus, I think the Juelich Atlas is particularly useful for delineating the amygdala (but also hippocampus and other areas or white-matter tracts), whereas one strength of the Harvard-Oxford-Atlas is its better coverage of the whole brain.
check out Tyszka and Pauli's new set of amygdala parcellations:
http://evendim.caltech.edu/amygdala-atlas/
discrepancies with other masks are discussed in the related paper
if you use them, however, make sure you have normalized your data non-linearly. furthermore, preferably not with FNIRT since it seems to perform worse than Dartel, 3dQwarp or ANTs
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology