Skip to content

In Vivo Human Whole-Brain Connectom Diffusion MRI Dataset at 760 µm Isotropic Resolution

Choose a tag to compare

View all tags
@data-hcp data-hcp released this 29 Oct 14:42
· 8 commits to main since this release
mgh-760
d86107c
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

This dataset provides a whole-brain in vivo diffusion MRI (dMRI) acquisition at 760 µm isotropic resolution, sampled at 1,260 q-space points across nine 2-hour sessions on a single healthy adult subject.
It demonstrates the limits of ultra-high-resolution diffusion imaging achievable with advanced Connectom-class hardware, optimized acquisition protocols, and reconstruction pipelines.

The dataset enables exploration of fine-scale in vivo brain structures, fiber microarchitecture, and serves as a benchmark for developing high-resolution modeling, sub-sampling, denoising, and reconstruction algorithms.

License

Creative Commons Zero v1.0 Universal (CC0-1.0) — public domain dedication

Citation

Wang, F., Dong, Z., Tian, Q., Liao, C., Fan, Q., Hoge, W. S., Keil, B., Polimeni, J. R., Wald, L. L., Huang, S. Y., & Setsompop, K. (2021).
In vivo human whole-brain Connectom diffusion MRI dataset at 760 µm isotropic resolution) [Data set]. Dryad.
https://doi.org/10.5061/dryad.rjdfn2z8g
https://datadryad.org/dataset/doi:10.5061/dryad.nzs7h44q2

Source

https://doi.org/10.5061/dryad.rjdfn2z8g
Contact: kawin@nmr.mgh.harvard.edu
Institutions: Massachusetts General Hospital (MGH) & Harvard Medical School
Supported by: NIH BRAIN Initiative & MGH A.A. Martinos Center for Biomedical Imaging
Related work: 10.5061/dryad.nzs7h44q2

Dataset Information

Category Details
Subject Single healthy adult volunteer
Scanner 3 T Connectom MRI system (300 mT/m gradient strength)
Coil Custom-built 64-channel phased-array head coil
Resolution 0.76 mm × 0.76 mm × 0.76 mm isotropic
Sessions 9 sessions × 2 hours = 18 hours total acquisition
q-space Sampling 1,260 directions total
Sequence SNR-efficient diffusion-weighted spin-echo EPI
Reconstruction Parallel imaging with ghost-artifact reduction and SENSE acceleration
Stabilization Personalized motion-robust head stabilizer
Additional Data Submillimeter T1w and T2w anatomical MRI, field maps, and preprocessing code
Data Format NIfTI (.nii.gz) and supporting metadata (.bval, .bvec, JSON)

Experimental Setup

This dataset was acquired using state-of-the-art Connectom hardware and optimized dMRI sequences to push spatial resolution while maintaining high SNR.

Key elements:

  • High gradient strength (300 mT/m) allowing short diffusion times and strong diffusion weighting.
  • Custom 64-channel phased-array coil for increased parallel imaging acceleration and sensitivity.
  • Advanced ghost reduction algorithms during reconstruction.
  • SNR-efficient acquisition design minimizing echo-train blurring.
  • Nine separate scanning sessions ensuring full q-space coverage while minimizing motion contamination.

Preprocessing and Code

The dataset includes accompanying preprocessing scripts and documentation for:

  • Denoising
  • Eddy-current and motion correction
  • Gradient-nonlinearity correction
  • Signal normalization and registration to T1/T2 anatomy

All reconstruction and preprocessing steps are described in detail within the included README and supplementary code repository.

Purpose

This dataset represents one of the highest-resolution in vivo human dMRI datasets publicly available.
It is intended to support:

  • Development and benchmarking of microstructural modeling and tractography algorithms
  • Validation of denoising and sub-sampling frameworks
  • Studies of fine-scale cortical and subcortical connectivity

Keywords

Diffusion MRI • High Resolution • Connectom Scanner • 760 µm • Human Brain • Microstructure • q-space Sampling • Parallel Imaging • Ghost Reduction • Tractography • Martinos Center • BRAIN Initiative

Assets 6
Loading