Hello!! I just wanted to verify some information that I am confused about! My understanding is that there are two dHCP surface template atlases: the asymmetric Bozek et al. (2018) and the symmetric dhcpSym40. For clarity: the Bozek et al. template has vertex-to-vertex correspondence with the HCP fsLR 32K space while the dhcpSym40 template does not. Also the dHCP release xfm folder’s surface transformation “from-native_to-dhcpSym40” is a transformation to this symmetric dhcpSym40 template. Is that all correct?
Additionally—if the dhcpSym40 template does not have vertex-to-vertex correspondence with the HCP fsLR space, would it be possible for me to generate a transformation to resample the dhcpSym40 template into fsLR 32K space? Thank you all so much!
I am also working on a similar research topic. My understanding is that the authors indeed provide two surface atlases. The dhcpSym40 atlas has left–right vertex-wise correspondence, whereas the Bozek atlas does not.
However, I am also a bit confused about whether these two atlases truly correspond to the HCP fs_LR 32k space. My impression is that they are defined in a neonatal template space rather than the adult fs_LR space, and therefore are not directly comparable between neonates and adults.
Regarding the “from-native_to-dhcpSym40” transformation, I agree with your interpretation. This transformation can be used to project native cortical surfaces into the dhcpSym40 space. The paper (Structural and functional asymmetry of the neonatal cerebral cortex | Nature Human Behaviour) provides more detailed information on this point, which may be helpful to refer to.
Hi both, we are using dhcpsym40, neither have vertex correspondence with the HCP but it is easy enough to do yourself through running MSM. Note none of these templates are compatible with outputs from the new deep learning pipeline because the spherical projection changed. If you want that we can send.