INDI Consortium for Reliability and Reproducibility (CoRR)

The Consortium for Reliability and Reproducibility (CoRR) was created to provide an open-science resource for test–retest reliability and reproducibility studies in functional and structural connectomics.

This repository collects and documents CoRR test–retest datasets that have been mirrored and/or processed by the Pittsburgh Fiber Data Hub, with helper files to simplify downstream analysis.

Upstream data are shared via the International Neuroimaging Data-sharing Initiative (INDI) and the 1000 Functional Connectomes Project.

---

Goals of CoRR

CoRR was established to:

1. Establish test–retest reliability and reproducibility for commonly used MR-based connectome metrics.
2. Characterize variability across sites and designs, including differences in reliability across imaging centers, scanners, and retest intervals.
3. Provide benchmark datasets for evaluating new metrics, preprocessing pipelines, and analysis methods.

Because CoRR is primarily a retrospective aggregation, its phenotypic key emphasizes variables that are:

• Core: minimal variables necessary to characterize any dataset (e.g., age, sex)
• Preferred: recommended variables that are important and likely to be collected across sites
• Optional: dataset-specific variables contributed by a subset of sites (e.g., task performance, specialized scales)

---

What this repository provides

For each CoRR dataset mirrored under data-indi/corr, this repository aims to provide:

• Short dataset summaries (site, design, sample size, modalities, DOI)
• Pointers to upstream downloads (NITRC / INDI project pages)

• Helper artifacts when available, such as:

  • participants.tsv
  • qc.tsv
  • Tractography-ready files (*.gqi.fz, *.qsdr.fz, *.sz, *.dti.dz, *.gqi.dz)

All raw imaging data remain hosted by INDI / NITRC and are subject to the INDI / CoRR Data Use Agreement.

---

XHCUMS — 6-Month Test–Retest Dataset

Dataset: Xuanwu Hospital, Capital Medical University (XHCUMS) – 6-Month Test–Retest Dataset DOI: 10.15387/fcp_indi.corr.xhcums1 Site: Xuanwu Hospital, Capital Medical University, Beijing, China Sample: 24 healthy adults, 5 sessions per subject, 6-month intervals Modalities: T1-weighted anatomical, resting-state fMRI (EPI), diffusion MRI (DTI) Condition: Rest, eyes closed, blank screen, no task

This dataset was designed to evaluate long-term test–retest reliability of structural, functional, and diffusion MRI measures. Each participant was scanned five times over ~2 years, providing dense longitudinal data for:

• Morphometry and cortical thickness
• Intrinsic functional connectivity
• Diffusion-based white-matter mapping

Typical assets in this repo include multi-modal group derivatives such as corr-xhcums.dti.dz, corr-xhcums.gqi.dz, participants.tsv, qc.tsv, and subject-level *_dwi.*.fz files.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/xhcums | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/xhcums").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

SWU1 — Training Effects for Attention Blink Task

Dataset: Southwest University (SWU1) – Training Effects for Attention Blink Task DOI: 10.15387/fcp_indi.corr.swu1 Site: Faculty of Psychology, Southwest University, Chongqing, China Sample: 20 healthy adults Design: 2 sessions — pre-training and post-training on an attention blink paradigm Modalities: T1-weighted structural MRI, resting-state fMRI (two sessions) Condition: Rest, eyes closed, no task

Participants completed an attention blink training paradigm between scans, enabling:

• Pre/post comparison of resting-state networks
• Evaluation of training-induced plasticity
• Test–retest reliability analysis in a short longitudinal design

Assets in this repo typically include subject-level *_T1w.nii.gz, resting-state fMRI derivatives, tractography-ready .fz files, and QC tables.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/swu | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/swu").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

NKI — Test–Retest Multiband fMRI and DTI

Dataset: Nathan Kline Institute – Test–Retest Multiband fMRI and DTI Dataset DOI: 10.15387/fcp_indi.corr.nki1 Site: Nathan S. Kline Institute for Psychiatric Research, Orangeburg, NY, USA Sample: 24 participants, multi-session test–retest Modalities: Multiband resting-state fMRI (TR = 645 ms, 1400 ms, 2500 ms), DTI (137 directions) Condition: Resting-state, eyes open, fixation

This dataset served as a pilot for the Enhanced NKI–Rockland Sample (NKI–RS), enabling:

• Test–retest reliability analysis across multiple TRs
• Multimodal comparisons between diffusion and functional measures
• Methods development for multiband fMRI and dynamic connectivity

This repo includes tractography-ready .dti.dz, .gqi.dz, subject-level .fz files, and QC information.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/nki | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/nki").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

MRN — Repeat Resting Sample

Dataset: Mind Research Network (MRN) – Repeat Resting Sample DOI: 10.15387/fcp_indi.corr.mrn1 Site: Mind Research Network, Albuquerque, NM, USA Sample: 56 healthy adults, 2 sessions, ~4 months apart Modalities: T1-weighted anatomical MRI, resting-state fMRI, DTI Condition: Rest, eyes open, fixation cross

Participants were scanned twice with a ~four-month interval, providing:

• Reproducibility benchmarks for resting-state networks
• Stability measures for diffusion-based connectivity
• Data to evaluate intra-individual variability over longer intervals

Data here include group .dti.dz / .gqi.dz files, subject-level tractography-ready .fz data, and QC tables.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/mrn | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/mrn").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

IPCAS1 — Time Perception and Estimation Sample

Dataset: Institute of Psychology, Chinese Academy of Sciences (IPCAS 1) – Time Perception and Estimation Sample DOI: 10.15387/fcp_indi.corr.ipcas1 Site: Institute of Psychology, Chinese Academy of Sciences, Beijing, China Sample: 30 healthy adults, 2 sessions one week apart Modalities: Resting-state fMRI, T1-weighted anatomical MRI, DTI Condition: Rest, eyes open, fixation cross

This dataset supports work on:

• Neural mechanisms of time perception and estimation
• Short-term test–retest reliability (1-week interval)
• Structure–function relationships involving the amygdala and other regions

The repo includes multi-modal group .dz files, participants.tsv, qc.tsv, and subject-level .fz derivatives.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/ipcas | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/ipcas").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

HNU1 — One-Month Test–Retest & Dynamical Resting-State

Dataset: Hangzhou Normal University (HNU) – One-Month Test–Retest Reliability and Dynamical Resting-State Study DOI: 10.15387/fcp_indi.corr.hnu1 Site: Center for Cognition and Brain Disorders, Hangzhou Normal University, Hangzhou, China Sample: 30 healthy adults Design: 10 sessions per participant, spaced 3 days apart (~1 month) Modalities per session: EPI (resting-state fMRI), ASL, T1-weighted, DTI, T2-weighted Condition: Rest, eyes open, fixation cross, no intentional task

HNU1 is one of the most comprehensive CoRR resources for:

• Dynamic functional connectivity
• Intra-subject variability across many sessions
• Short-term reliability across multiple MRI modalities

This repo provides group .dti.dz / .gqi.dz, session-level .fz files, and QC/participants tables to facilitate longitudinal analyses.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/hnu | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/hnu").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

BNU1 — Connectivity-Based Brain Imaging Research Database (C-BIRD)

Dataset: Beijing Normal University (BNU) – Connectivity-Based Brain Imaging Research Database (C-BIRD) DOI: 10.15387/fcp_indi.corr.bnu1 Site: State Key Laboratory of Cognitive Neuroscience and Learning, and IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China Sample: 57 healthy young adults (30 male / 27 female), ages 19–30 Design: 2 sessions ~6 weeks apart (40.9 ± 4.5 days) Modalities: Resting-state fMRI, T1-weighted, T2-weighted, DTI Condition: Rest, eyes closed, no task

BNU1 supports:

• Reliability and temporal stability analyses of functional and structural connectivity
• Graph-theoretic connectivity and network topology studies
• Cross-session comparisons of resting-state networks and DTI-based connectomes

This repo includes BNU1 .dti.dz / .gqi.dz, participants.tsv, qc.tsv, and subject-level tractography-ready .fz files.

Download (Linux / macOS — bash)

curl -s https://api.github.com/repos/data-indi/corr/releases/tags/bnu | jq -r '.assets[].browser_download_url' | xargs -n1 -P4 curl -LO

Download (Windows PowerShell 5.x)

(Invoke-RestMethod "https://api.github.com/repos/data-indi/corr/releases/tags/bnu").assets | ForEach-Object { Invoke-WebRequest $_.browser_download_url -OutFile (Split-Path $_.browser_download_url -Leaf) }

---

How to access original data

To obtain original imaging and phenotypic data from CoRR / INDI (beyond the derivatives mirrored here):

1. Create an account on NITRC.
2. Request to join the INDI / 1000 Functional Connectomes Project group.
3. Navigate to the corresponding CoRR dataset page (e.g., XHCUMS, SWU1, NKI, MRN, IPCAS1, HNU1, BNU1).
4. Download imaging and phenotypic data from NITRC once access is approved.

The derivatives in this repository are intended to complement the official CoRR distributions and to facilitate diffusion/tractography and connectomics analyses.

---

Licensing and data use

Use of CoRR datasets is governed by the INDI / CoRR Data Use Agreement:

http://fcon_1000.projects.nitrc.org/indi/corr/html/data_citation.html

In general:

• Datasets are for research use only.
• Users must not attempt re-identification of participants.
• Users must follow any site-specific restrictions and citation requirements listed on the dataset page.

Check each project page for precise license terms (including acknowledgements and additional conditions).

---

How to cite

Always:

1. Cite the original dataset / DOI and associated publications (e.g., site-specific reliability papers).
2. Acknowledge CoRR, INDI, and the contributing site(s) as specified on the CoRR citation page.
3. When using derivatives or helper files from this repository, please add:

“Processed artifacts and curation were prepared by the Pittsburgh Fiber Data Hub (data-indi/corr) based on INDI/CoRR datasets.”

---

Disclaimer

• All data originate from independent sites and are shared as-is.
• Sequence parameters, scanner models, and acquisition instructions (eyes open/closed, fixation, etc.) differ across datasets.

• Users are responsible for:

  • Performing appropriate quality control (e.g., motion, artifacts).
  • Choosing analysis pipelines that account for site differences, TR differences, and multiband acquisition when applicable.