MBIRJAX: High-performance tomographic reconstruction#
Warning
MBIRJAX is now legacy software. It has been replaced by MBIRTorch, which has the same API, runs on PyTorch, and is faster and more memory efficient. MBIRJAX will not be updated further. Users should migrate to MBIRTorch.
Repository: github.com/cabouman/mbirtorch
Documentation: mbirtorch.readthedocs.io
Migration guide: Migrating from MBIRJAX
MBIRJAX is a Python package for Model Based Iterative Reconstruction (MBIR) of images from tomographic data.
Key features:
Vectorized Coordinate Descent algorithm for fast, robust convergence [A1].
Automatic parameter selection, with fine-tuning using intuitive meta-parameters.
Support for Plug-and-Play prior models that can dramatically improve image quality [A4] [A3].
Modular, extensible, easy-to-use, object-oriented Python interface.
Fast, portable, seamless use on CPUs or GPUs through the use of JAX.
Reconstructions can be performed in just a few lines of python code.
Vectorized Coordinate Descent produces high-quality images quickly.
Based on JAX, MBIRJAX can easily run on CPU or GPU.