CEA.fr Health & Life Sciences BrainVISA Champollion Pipeline
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
BrainVISA
BrainVISA Neuroimaging Platform
NeuroSpin · Software Champollion Pipeline

NeuroSpin — CEA Paris-Saclay — BrainVISA 6.0

Champollion PipelineFrom raw MRI to cortical embeddings

An end-to-end pipeline that transforms T1 MRI images into compact embeddings of sulcal morphology — built on the BrainVISA / Morphologist ecosystem and ready for population-scale brain analysis and disease prediction.

Python ≥ 3.8 PyTorch Pixi CeCILL-B BrainVISA 6.0 Morphologist
56
Pre-trained model folds
(28 regions × 2 hemispheres)
42k+
UKBioBank reference
subjects
6.0
BrainVISA / Morphologist
release
Part of the BrainVISA 6.0 neuroimaging platform — providing Morphologist brain segmentation, Anatomist 3D visualisation, and the full sulcal analysis ecosystem.

About the pipeline

A complete, reproducible workflow for cortical sulcal analysis

The Champollion Pipeline orchestrates the full journey from raw anatomical MRI data to compact embeddings of sulcal morphology. It leverages Morphologist (part of BrainVISA 6.0) to extract sulcal fold graphs, then uses cortical_tiles to tile 28 sulcal regions, and finally generates embeddings using pre-trained Champollion V1 transformer models hosted on Hugging Face.

A live demo is available on Hugging Face Spaces for quick single-subject testing. For batch processing or production use, install locally to leverage all available CPUs and GPUs. BrainVISA 6.0, DISCO, and Champollion are all available for download together.

Pipeline steps

Seven stages, from installation to visualisation

01
Installation & Environment Setup
Clone the repository and run pixi run install-all. Pixi initialises git submodules (champollion_V1, cortical_tiles), installs all Python packages in editable mode, and creates the data directory automatically.
Pixi
02
Generate Morphologist Sulcal Graphs
Run morphologist-cli on T1 NIfTI files to extract sulcal fold graphs using Morphologist (BrainVISA 6.0). Supports serial and parallel execution via soma-workflow with a configurable CPU count.
MorphologistBrainVISA 6.0soma-workflow
03
Generate Sulcal Regions (cortical_tiles)
Extract 28 sulcal region crops from Morphologist output. Accepts wildcard path segments, optional QC files, and configurable input types: skeleton, foldlabel, extremities.
cortical_tilesBrainVISA
04
Generate Champollion Configuration
Auto-generate the YAML dataset config consumed by the embedding scripts. Supports read-only container environments (Apptainer) via the --external-config flag.
YAMLApptainer
05
Generate Embeddings
Run all 56 pre-trained model folds to produce one full_embeddings.csv per region. Models load from local paths, archives, or directly from the Hugging Face hub (neurospin/Champollion_V1). GPU and CPU modes supported.
PyTorchHugging Face
06
Combine Embeddings
Merge all 56 per-region CSV files into a single output directory using put_together_embeddings.py. Verify with a quick ls … | wc -l check (expected: 56 files).
CSV aggregation
07
Generate Visualization Snapshots
Produce sulcal graph meshes, cortical tile masks, and UMAP scatter plots using Anatomist. Each new subject is projected onto a pre-trained 2D UMAP fitted on 42,433 anonymous UKBioBank reference subjects.
AnatomistUMAP

Tutorials

Step-by-step Jupyter notebooks from the cortical_tiles and Champollion V1 repositories

The following notebooks live in the cortical_tiles and champollion_V1 repositories — installed locally as git submodules under external/. Click any card to open the notebook on GitHub.

Installation

Requires Pixi and Git — all other dependencies managed automatically

bash — Quick setup
# 1. Clone the repository
mkdir Champollion && cd Champollion
git clone https://github.com/neurospin/champollion_pipeline.git
cd champollion_pipeline

# 2. Install all dependencies (submodules + Python packages)
pixi run install-all

# 3. Activate the Pixi environment
pixi shell

# 4. (Optional) Run the test suite
pixi run test

Prerequisites

  • Pixi package manager
  • Git (with submodule support)
  • BrainVISA 6.0 (for Morphologist)
  • Python ≥ 3.8

Key dependencies

Try Champollion in your browser

The Hugging Face Spaces demo runs on 2 CPU cores — ideal for quick single-subject testing with no installation required.

Open Live Demo →
Partners & ecosystem