Untamed: Cortical Parcellation Atlases from 1400 Subject fMRI

Atlases generated from resting-state fMRI of the 1400 subjects in the Brain Genomics Superstruct Project (GSP) dataset. The data is available in fsaverage6 and fsLR 32K surface space and was authored by Yijun Liu in 2026.

Use Cases

• Validate cortical parcellation algorithms based on the provided brain atlases.
• Study brain connectivity patterns based on resting-state fMRI data from the GSP cohort.
• Benchmark graph node embedding techniques for neuroimaging data.
• Compare parcellation results across different surface spaces (fsaverage6 and fsLR 32K).

Strengths

• Derived from a substantial cohort of 1400 subjects from the Brain Genomics Superstruct Project.
• Provides data in two standard neuroimaging surface spaces (fsaverage6 and fsLR 32K).
• Has a clear citation trail to a 2026 technical report from Wiley Online Library.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Brain Genomics Superstruct Project (GSP) dataset.

Collection Method

Generated via unconstrained tensor decomposition and graph node embedding applied to resting-state fMRI.

Time Range

The atlas generation method was published in 2026; underlying GSP data collection timeframe is unspecified.

Freshness

Last updated 2026-05-21 05:37:28; freshness should be verified.

Geography

Spatial coverage of the underlying GSP cohort is unspecified.