Skip to content

3T Diffusion MRI Inter-Site Reproducibility Dataset for Cerebral Small Vessel Disease

Choose a tag to compare

View all tags
@data-hcp data-hcp released this 29 Oct 14:30
isd-svd
d418a73
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 multi-site diffusion MRI data from 10 patients diagnosed with the genetically defined small vessel disease CADASIL, scanned on two different 3 T Siemens MRI scanners (Magnetom Prisma and Magnetom Skyra) within 24 hours.
It enables evaluation of inter-scanner reproducibility of diffusion metrics in cerebral small vessel disease (SVD) and supports harmonization studies for multi-center neuroimaging trials.

The dataset was used in the publication:
Konieczny & Dewenter et al., Neurology, 2020 — https://doi.org/10.1212/WNL.0000000000011213

License

Creative Commons Attribution – NonCommercial 4.0 International (CC BY-NC 4.0)

Citation

Dewenter, A., Gesierich, B., & Duering, M. (2020).
3T diffusion MRI inter-site reproducibility dataset for cerebral small vessel disease [Data set]. Zenodo.
https://doi.org/10.5281/zenodo.16925507

Source

https://doi.org/10.5281/zenodo.16925507
Contact: marco.duering@med.uni-muenchen.de
Institution: Institute for Stroke and Dementia Research (ISD), LMU Munich, Germany
Associated publication: Neurology, 96(5), e698–e708 (2020)

Dataset Information

Category Details
Subjects 10 patients with CADASIL (genetically confirmed or via skin biopsy)
Age (median [IQR]) 55.5 years (13.3)
Sex 5 female (50 %)
Hypertension / Hypercholesterolemia 3 / 5 patients
WMH volume [% ICV] median (IQR) 7.67 (2.62)
Lacune count median (IQR) 3 (7.75)
Brain volume [% ICV] median (IQR) 75.7 (7.03)
Scanner Types Siemens 3 T Magnetom Prisma and Skyra
Scan Interval < 24 hours
Protocol Matched b-values, gradient directions, and resolution between scanners
BIDS-Compliance Data organized per subject and session (/sub-XX/ses-prisma, /ses-skyra)
Preprocessing Tools dcm2niix, MRtrix (dwidenoise & mrdegibbs), FSL (TOPUP & EDDY)
Anonymization Fully de-identified and restricted to diffusion MRI only

MRI Acquisition Protocol

Parameter Magnetom Skyra Magnetom Prisma
Coil 64-channel head/neck 64-channel head/neck
TR (ms) 3800 3200
TE (ms) 104.8 74
Resolution (mm³) 2×2×2 2×2×2
b-values (s/mm²) 1000 / 2000 1000 / 2000
Directions (per b) 30 / 60 30 / 60
b = 0 images 10 10
Multi-band factor 3 3
Acceleration factor 2 2

Both scanners were located in separate hospital buildings with independent infrastructure, establishing a true inter-site condition despite institutional proximity.

Data Structure and Preprocessing

The dataset follows the BIDS organization scheme:

/sub-01
   /ses-prisma
       /dwi_raw
       /dwi_preprocessed
   /ses-skyra
       /dwi_raw
       /dwi_preprocessed
  • Raw data converted from DICOM via dcm2niix
  • Preprocessed data includes denoising (MRtrix dwidenoise), Gibbs correction (mrdegibbs), and motion/eddy distortion correction (FSL TOPUP, EDDY).

Purpose

This dataset serves as a benchmark for assessing inter-scanner reproducibility and harmonization of diffusion MRI measures across 3 T sites.
It is particularly relevant for researchers working on:

  • Cross-site calibration of diffusion metrics in small vessel disease
  • Methodological validation of preprocessing and harmonization pipelines
  • Clinical reproducibility of DTI/DKI biomarkers in multicenter trials

Keywords

Diffusion MRI • DTI • CADASIL • Small Vessel Disease • Inter-site Reproducibility • Siemens Prisma • Siemens Skyra • BIDS • FSL • MRtrix • Neurology 2020

Assets 64
Loading