Does dhcpSym_32k space has correspondence with fs_LR_32k space?

1

Hi guys,
I am little confused about the relationship between template space of dhcpSym_32k and fs_LR_32k of HCP. Are these two have vertex-to-vertex correspondence? If no correspondence relationship remains, what should I do to build this correspondence?

Yao Meng

2

I think have found the answer, which is yes. According to Jelena Bozek et al. paper.
image
Bozek, J., Makropoulos, A., Schuh, A., Fitzgibbon, S., Wright, R., Glasser, M.F., Coalson, T.S., O’Muircheartaigh, J., Hutter, J., Price, A.N., et al. (2018). Construction of a neonatal cortical surface atlas using Multimodal Surface Matching in the Developing Human Connectome Project. NeuroImage 179, 11–29.

3

Still waiting someone to confirm that :joy:

4

Hi @YaoMeng,

The dhcpSym template is not in the same space as the Bozek et al. template, so I don’t think the excerpt from the text applies.

Best wishes,

Logan

5

Thanks Logan,
I must mix things up. So the dhcpSym doesn’t have vertex-to-vertex correspondence, is that you mean? Do you know some resources that describe the generation of dhcp_Sym space?

Yao

6

Hi @YaoMeng,

If the Bozek et al atlas does have vertex correspondence with the HCP template, then the dhcpSym template won’t. However, the dhcpSym template has left-right vertex correspondence - more information about the template generation can be found in the methods section of the preprint here.

Best wishes,

Logan

7

Thans Logan,

That's clear from your preprint I think, the dhcp doesn't in the same space with fs_LR but Bozek et al atlas does. Very much of thanks for you reply.

Yao

8

Hi @YaoMeng and @lzjwilliams , I came across the same question. First, I’m working with the 4th release of the dHCP dataset. In the anat folder, there is a file: hemi-right_from-native_to-dhcpSym40_dens-32k_mode-sphere.surf.gii My understanding is that this sphere is used to map data to the dhcpSym template, but I’d like to confirm—could it be referencing the Bozek et al. atlas instead?

Second, is there a known way to map the dhcpSym template to fs_LR 32k (HCP standard space), especially given that there is no direct vertex correspondence between the two surfaces? I found this dHCP_HCP-YA/CorticalSurfaceRegistration.sh and
dHCP_HCP-YA.MSMStrain.sphere.reg.surf.gii from your manuscript @lzjwilliams, I think I can use this directly? Would really appreciate your help!

Thanks in advance!