Notebook Gallery

The demo notebooks synthesize their own inputs and write temporary outputs under tmp/ or notebook-local demo folders. They do not require bundled datasets.

Notebooks are rendered inline via nbsphinx with execution disabled during docs builds. The downloads import the installed mrax package and therefore work outside a source checkout. Install notebook tooling and the optional example backends, then launch the directory containing the downloaded notebook:

pip install "mrax[notebooks,nufft,denoise-external]"
jupyter lab

The upstream BM3D/BM4D packages permit non-commercial use only. Method references for the models, regularizers, reconstruction methods, and denoisers used in the notebooks are collected in References.

Rendered notebooks

• Getting Started
• Pulseq Import Demo
• Reconstruction Pipeline
• Variational Reconstruction
• Analysis Toolbox
• Preprocessing Denoiser Benchmarks

Downloads

• getting_started.ipynb — synthetic pipeline

• pulseq_import_demo.ipynb — Pulseq Cartesian import and fitting walkthrough

• reconstruction_pipeline.ipynb — raw k-space and reconstruction blocks

• variational_reconstruction.ipynb — variational MRI reconstruction (TV + NUFFT)

• analysis_toolbox.ipynb — post-fit analysis demo

• preprocessing_denoiser_benchmarks.ipynb — denoiser benchmarks (metrics + images)