SETR

Synergistic Emission Tomographic Reconstruction

Tools for nuclear medicine image reconstruction combining PET/SPECT with CT anatomical information.

Contents

• Vectorial Total Variation regularization
• GPU-accelerated gradients and Jacobians
• Python wrapper for STIR Kernel EM
• Numba-optimized kernel operators
• Extended CIL functionality:
  • Preconditioners
  • Callbacks
  • Step size rules
• SIRF data loading utilities

Installation

# Install dependencies via conda / pip or from source
- SIRF
- CIL

# Install SETR
pip install py-setr

Note

In order to use the partitioner with SPECT, you will need to build stir with https://github.com/samdporter/STIR/tree/SPECT_subsets