Deep Learning Cone Beam Computed Tomography

This repository provides an implementation of a Cone Beam Backprojection for Tensorflow. It can be used to data-driven learn preprocessing steps in the projection domain. As an example, we include the training of redundancy weights, as reported with the accompanying paper.

How to build

mkdir build
cd build
cmake -DCMAKE_INSTALL_PREFIX=[...] ..
make && make install

References

In case you use this code for your own scientific work, please cite the following publication:

Würfl, Tobias, et al. "Deep learning computed tomography: Learning projection-domain weights from image domain in limited angle problems." IEEE transactions on medical imaging 37.6 (2018): 1454-1463.

BibTeX:

@article{wurfl2018deep,
  title={Deep learning computed tomography: Learning projection-domain weights from image domain in limited angle problems},
  author={W{\"u}rfl, Tobias and Hoffmann, Mathis and Christlein, Vincent and Breininger, Katharina and Huang, Yixin and Unberath, Mathias and Maier, Andreas K},
  journal={IEEE transactions on medical imaging},
  volume={37},
  number={6},
  pages={1454--1463},
  year={2018},
  publisher={IEEE}
}