Tutorials and guides for Nobrainer — a deep learning framework for 3D brain image processing.
Tutorials¶
Follow these tutorials in order for a complete learning journey:
| # | Tutorial | What you’ll learn |
|---|---|---|
| 01 | Getting Started | Install, import, explore available models |
| 02 | Download & Inspect | Get sample brain MRI data, understand NIfTI format |
| 03 | Extract & Batch | Prepare training data with patches and datasets |
| 04 | Brain Segmentation | Train a UNet in 3 lines with the estimator API |
| 05 | Uncertainty | Bayesian inference with variance and entropy maps |
| 06 | Brain Generation | Synthesize brain volumes with Progressive GAN |
| 07 | Advanced Training | Custom PyTorch loops for full control |
| 08 | Model Management | Save/load with Croissant-ML metadata |
| 09 | Zarr Pipeline | Cloud-optimized multi-resolution storage |
| 10 | Multi-GPU | Scale training across GPUs with DDP |
| 11 | Contributing | Add models, write tests, submit PRs |
Run in Google Colab¶
Stable release (master):
Pre-release (alpha):
Branches¶
| Branch | Nobrainer version | Install |
|---|---|---|
| master | Stable | pip install nobrainer |
| alpha | Pre-release | pip install --pre nobrainer |
Run locally¶
uv venv --python 3.14
source .venv/bin/activate
uv pip install "nobrainer[bayesian,generative,zarr]" monai pyro-ppl nilearn matplotlib
for script in docs/nobrainer-guides/scripts/[01]*.py; do python "$script"; done