This repository will contain the code for the article "Solving ill-posed inverse problems using iterative deep neural networks" published on arXiv.
The code contains the following
- Training using ellipse phantoms
- Evaluation on ellipse phantoms
- Training using anthropomorphic data. (Data not included for legal/privacy reasons)
- Evaluation on example head slice
- Reference reconstructions of the above using ODL.
The code is currently based on an experimental version of ODL, so that specific branch needs to be used for the code to work.
Jonas Adler, PhD student
KTH, Royal Institute of Technology
Elekta Instrument AB
[email protected]
Ozan Öktem, Associate Professor
KTH, Royal Institute of Technology
[email protected]
Development is financially supported by the Swedish Foundation for Strategic Research as part of the project "Low complexity image reconstruction in medical imaging" and "3D reconstruction with simulated forward models".
Development has also been financed by Elekta.