We are pleased to announce the release of an improved Group Average Dense Connectome based on resting state fMRI data from 468 subjects that were part of the 500 Subjects Release (the R468 group).
A number of processing pipelines are currently being implemented and refined by the HCP that carry out further analyses at the group level. As of April 2015, with improved analysis methods detailed on the HCP website: Correcting for the rfMRI Mound-and-Moat Effect, we have updated the group-average functional connectivity dataset we are distributing on the 500 Subjects Release R468 group (468 subjects, including many subjects that are related, with complete resting state fMRI data).
The group-average data are available for download through the links on the WU-Minn HCP Project page in ConnectomeDB. One can view the subjects included in this analysis using the “Open group” function on the ConnectomeDB dashboard.
The group-average rfMRI data includes:
- Group-average functional connectivity matrix (“dense” functional connectome, the grayordinate × grayordinate full correlation matrix), for the R468 group. Because of its large size (33 GB) this dense functional connectome file is released separately from the rest of the group average data.
- Group-PCA eigenmaps for the R468 group. These can be used as input to group-ICA. They can also be used to generate the dense connectome, but to do this optimally is not trivial, and requires following the procedures outlined in Correcting for the rfMRI Mound-and-Moat Effect.
If you prefer to view the dense connectome file in Connectome Workbench (recommended), you do not need to download it. The data are accessible in Workbench by remote access (requires internet connection and ConnectomeDB login), using the following URL: https://db.humanconnectome.org/spring/cifti-average?resource=HCP_Resources:GroupAvg:HCP_S500_R468_rfMRI_MIGP_d4500ROW_zcorr
We are pleased to announce the release of the “Netmats MegaTrawl”, a set of webpages inside ConnectomeDB that summarizes analyses of the relationships between imaging and non-imaging measures in the Human Connectome Project.
Multivariate-prediction and univariate-regression were used to relate 187 non-imaging behavioral and demographic subject measures (SMs: age, sex, education, tobacco use, fluid intelligence (IQ), reading ability, etc.) to 461 individuals’ functional connectivity resting-state fMRI (rfMRI) data.
Multivariate analyses seek to model (across subjects) a given subject measure, finding a set of edge weights (functional connections between brain regions, or parcels) in the data that can partially explain that subject measure. Results of this analysis are presented as web pages that enable visualization of the group-ICA parcellations utilized for a given MegaTrawl analysis, group-average netmats (network functional connectivity matrices), heritability calculations, and multivariate prediction and univariate regression results for each subject measure with thumbnail volume images showing the edges (node – pairs) whose connection most strongly correlates with the variable.
To get access, login into ConnectomeDB (sign up for an account if you don’t have one) and click Open Dataset for the WU-Minn HCP Data project. The Netmats MegaTrawl link is about halfway down the project page. Or use https://db.humanconnectome.org/data/projects/HCP_500#megatrawl and login when prompted.
Full documentation on the NetmatsMegatrawl release is at https://db.humanconnectome.org/megatrawl/HCP500_MegaTrawl_April2015.pdf (requires login).
In a post in honor of Brain Awareness Week, Tom Insel, M.D., Director of the National Institute of Mental Health (NIMH) raised the Human Connectome Project as one of this year’s major advances in brain science:
This year, this is a good time to note a few recent advances. In a few months, the Human Connectome Project will complete its multimodal study of 1,200 healthy adults, including 300 twin pairs. Already, data on over 500 subjects have been made public, creating an unprecedented treasure trove for students who want to explore individual variation in brain pathways. Like the Human Genome Project that created a fundamental map of our genetic sequence, the Human Connectome Project will provide a reference atlas of macro-level brain connections that can be used to study development, diseases, and species differences. Developmental connectomes and disease connectome projects will follow soon.
Insel mentions the surprising rise of interest and investment in brain mapping, both in the U.S., with the President’s BRAIN Initiative, and in other major efforts worldwide. The first 58 projects funded by the BRAIN Initiative have now begun and plans for funding new research and clinical tools through the BRAIN Initiative are unfolding.
The detailed information HCP is producing on connections and variability in the healthy human brain is providing a baseline set of knowledge on which applied projects in the BRAIN Initiative may develop tools for defining and ultimately treating mental disorders in patients.