Please cite the following paper in which VesselGen is described:
Shen, J., Faruqi, A. H., Jiang, Y., & Maftoon, N. (2021). Mathematical Reconstruction of Patient-Specific Vascular Networks Based on Clinical Images and Global Optimization. IEEE Access, 9, 20648–20661. https://ieeexplore.ieee.org/document/9328247
This package reconstructs patient-specific vascular networks based on patient priors obtained through MR angiograms. It accepts clinical image in the .nii.gz format and outputs a network configuration file (.npy) as well as the generated mesh (.stl). It also supports 3-D rendering of the generated network with Autodesk CAD (package Fusion 360).
The package consists of a complete reconstruction pipeline that includes image processing, root and leaf node selection, multi-tree global constructive optimization, network model generation (3-D rendering), and network model analysis. Most of the scripts are written in python, with the optimization algorithm implemented in C.
To run the reconstruction process, the following additional software/packages are required:
- vesseg
- bin2mesh
- vmtk
- binvox
- thinvox
Then, follow the instructions in VesselGen/pipeline.txt step by step.