BART workshops and webinars#
The BART tutorial catalogue and webinar archive provide recordings, scripts and datasets. This reading order connects them to bartorch. Their command-line, MATLAB and Python scripts need their filenames, dimensions and option syntax translated before they run here.
Learning goal |
Upstream material |
bartorch progression |
|---|---|---|
First reconstruction |
Phantom, calibration, coil preparation, the |
|
Noise processing and parallel imaging |
Whiten and compress consistently; then measured data and g-factor. |
|
Non-Cartesian reconstruction |
Webinar 7 (December 2022), in the webinar archive |
The radial operator and its direct Fourier check, then density compensation, sampling weights and measured trajectories. |
Temporal subspaces and signal models |
The model-based example, then basis contraction and quantitative mapping. |
|
Dynamic imaging |
Webinar 2 (December 2020), in the webinar archive |
Explicit time axes, data-consistent regularization, then GRASP-like applications. |
Wave reconstruction |
The hybrid-space wave model and the measured-data recipe in Research recipes: EPI, wave encoding, shuffling. |
|
Learning-based reconstruction |
Webinar 6 (March 2022), in the webinar archive |
BART’s network workflows beside the gallery’s DeepInverse physics and autograd. |
Volumetric reconstruction |
Webinar 8 (April 2024), in the webinar archive |
Axis-aware examples in 3-D, with memory measured. |
Follow the upstream links for current data locations and licenses. The gallery scripts are newly written synthetic examples and redistribute no workshop data or figures.