Resting-state fMRI
Build functional connectivity matrices from resting-state fMRI data, compute graph metrics, and compare connectivity patterns across cohorts.
Research question
How is functional connectivity organized in this resting-state scan, and how does it differ across patient and control groups?
Who this is for
- Cognitive and systems neuroscience labs running resting-state fMRI
- Psychiatry research groups studying default-mode and task-negative network dysconnectivity
- Clinical imaging teams needing reproducible FC pipelines without SPM/FSL scripting
Data requirements
| Data | Required | Purpose |
|---|---|---|
| Preprocessed NIfTI timeseries | Yes | ROI extraction and correlation |
| BIDS archive | No | Automated validation and metadata detection |
| Subject group labels | No (required for cohort) | NBS and group metric comparison |
| Confound regressors | No | Motion and physiological noise regression |
Workflow
Create study → Upload NIfTI or BIDS → Functional connectivity → Graph metrics → Cohort NBS (optional)
Step 1 — Upload and validate
Upload preprocessed resting-state NIfTI per subject, or a BIDS ZIP archive, then run Validate BIDS from the dataset workspace.
Review validation warnings for missing BIDS fields before running FC pipelines.
Step 2 — Functional connectivity
Launch functional connectivity from the dataset workspace:
| Parameter | Options | Purpose |
|---|---|---|
atlas_name | aal, schaefer_100/200/400 | ROI parcellation scheme |
confound_strategy | motion, compcor, none | Nuisance regression |
| Bandpass cutoffs | configurable | Resting-state frequency filtering |
Outputs include ROI timeseries, Pearson correlation matrix, and Fisher z-transformed connectivity.
Step 3 — Dynamic FC (optional)
Run dynamic functional connectivity for sliding-window state analysis:
- Windowed correlation matrices across the scan
- k-means state identification and state timecourses
- Transition statistics between connectivity states
Step 4 — Graph metrics
Compute BCT graph metrics on the resulting connectivity matrix to characterize network topology for each subject.
Step 5 — Cohort comparison (optional)
When multiple subjects are uploaded with group labels, proceed to Cohort comparison for NBS and group metric analysis.
Expected outputs
- ROI-to-ROI functional connectivity matrix per subject
- Dynamic FC state maps and timecourses (when dynamic pipeline run)
- Global and nodal graph metrics per subject
- BIDS validation report with missing-field warnings
Typical analyses
| Analysis | Comparison | Question |
|---|---|---|
| Default-mode dysconnectivity | Patient vs. control | Are DMN hub connections reduced in the patient group? |
| Treatment effect | Pre- vs. post-treatment | Does FC change after pharmacological intervention? |
| State dynamics | High vs. low symptom score | Do connectivity state transitions differ by clinical severity? |