Skip to content

simnibs-analyze

Documentation for version 0.2.0

Requires simnibs-reader >= 0.3.0. See the changelog.

Turn SimNIBS simulation outputs into cohort-level e-field statistics and figures — driven entirely by a YAML config.

simnibs-analyze sits on top of simnibs-reader: the reader handles NIfTI plumbing and ROI extraction, while this package handles the pipeline — feature extraction across subjects and conditions, inter/intra-subject analysis, clustering, and publication-ready visualisation.

Installation

pip install simnibs-analyze

This pulls in simnibs-reader >= 0.2.0 automatically.

Two commands

The package installs two console entry points, both config-driven:

simnibs-analyze --config config-analyze.yaml   # features + statistics
simnibs-viz     --config config-viz.yaml       # figures + cohort montages

See Getting Started for the full argument reference.

What the analysis pipeline does

Per (subject × condition × mode) it:

  1. locates the simulation folder via the reader,
  2. extracts the target ROI (sphere / atlas / mask) in the chosen space,
  3. post-processes (smoothing + outlier removal),
  4. computes intra-ROI stats, extra-ROI (complement) stats, and the intra/extra ratio,
  5. appends one row to all_features_space-<space>.csv.

It then runs the inter/intra-subject analysis and clustering, writing the summary CSVs to results_dir.

What it covers

  • Target definition — ROI masks from MNI coordinates or atlas parcels (sphere, atlas-based)
  • E-field preparation — smoothing, masking, intra/extra-ROI decomposition
  • Feature extraction — mean, median, std, min, max and focality ratio, per subject and condition
  • Group analysis — inter-subject summaries, intra-subject contrasts
  • Clustering — specific/diffuse × high/low labelling from the focality ratio
  • Visualisation — 2D overlays, 3D surface renders, lesion overlays, EEG electrode caps, cohort montages

Ecosystem

  • SimNIBS

    The core simulation platform for non-invasive brain stimulation.

    Documentation

  • simnibs-modular

    Modular pipeline components for SimNIBS workflows.

    GitHub Pages · Repo

  • simnibs-reader

    Structured access to SimNIBS output trees, ROIs and statistics.

    GitHub Pages · Repo

  • OptiStims

    End-to-end automated stroke lesion-aware tDCS pipeline.

    GitHub Pages · Repo

Citation

If you use this pipeline in your research, please cite it via the CITATION.cff file in the repository.

The pipeline depends on several open-source tools whose references are printed at the start of each run. For a machine-readable BibTeX summary:

pip install "simnibs-analyze[citations]"
DUECREDIT_ENABLE=1 simnibs-analyze --config config.yaml
python -m duecredit summary --format bibtex

Key tools to cite: SimNIBS, nilearn, nibabel, NumPy, pandas, matplotlib.