Astra-toolbox-for-cone-beam

This is a collection of Python scripts for implementing ASTRA Toolbox for cone-beam X-ray CT reconstruction. The complete scripts consist of raw data loading, configuration loading, center of rotation determination, back-propagation/iterative reconstruction, edge noise removal, and 3D volume generation. The current script includes all functions but 3D volume generation. We will add denoising, 3D volume generation, and detailed tutorial later on.

If you think this is useful for your research, please star this repository to make it noticeable to more people.

Before executing the script, please download the dataset from here and save it in ./raw file. The image for determining the center of rotation will be saved to offset_determination. After the code is executed, the reconstructed data will be saved to ./recon. You can refer to the comments in the script to modify the configuration setting in terms of your source and detector context.

If this link expires, please reach out to me to access the new link to the dataset.

Requirements

• All scripts make use of the development version (more recent than 1.9.0dev) of the ASTRA toolbox. If you are using conda, this is available through the astra-toolbox/label/dev channel.

Setup environment

Update apt-get

sudo apt update

Setup conda env:

conda env create -f environment.yml
conda activate astra-cone-beam

Contributors

Mingdian Liu (lmdvigor@gmail.com)