This is the repository for paper "Fluid Registration Between Lung CT and Stationary Chest Tomosynthesis Images" MICCAI 2020 Paper,Presentation
In this work we propose a differentiable projection operator which renders 2D projections from a 3D volume with given geometry parameters (pose of the emitters).
This operator can be used in a 2D3D deformable registration framework as shown below.
- Setup environment
cd lung_sdt_ct conda env create -f environment.yml - Install Mermaid
- Install this repo
python setup.py develop
To run the code, two setting files are required. One is for configuration of the mermaid library. The other is configuration file of the data. Sample setting files can be found at here
cd lung_sdt_ct
python registration/test_ct_registration.py -s ./settings/dirlab/lung_registration_setting_dct1.json -d [OUTPUT_FOLDER] -p [PREPROCESSED_DATA_OUTPUT_FOLDER]