This is a patient-specific 2D Cine-MRI simulator capable to generate a simulated ground truth contour of the desired organs using pre-treatment images. The input images are a 4DCT scan and an 3D MR with organ contours. The video simulator has the following input parameters: video time, frames per second, breathing cycle time, breathing amplitude, and additive noise.
The figure below illustrates the video simulation process. The process comprises two stages: breathing modeling and video synthesis. The breathing model is a pre-processing stage and uses full 3D information to consider out-of-plane motion in the 2D Cine-MRI. The video synthesis stage can be run several times changing the simulation parameters to create different variants and motion conditions.
Python:
- yaml
- numpy
- matplotlib
- pydicom
- skimage
- SimpleITK >= 2.0
External:
- ANTs webpage
Note: You require to have ANTs installed in your system with a working binary command 'antsRegistration'.
To run a simulation you require a set of images. A sample image folder is provided in 'images/sample'. Create a cinemr with simulation using the command line as:
python simulation.py -v images/patient/ model/patient/ out/ parameters.yaml
To cite when using this toolbox, please reference, as appropriate:
@inproceedings{tascon2021cine,
title={Cine-MRI Simulation to Evaluate Tumor Tracking},
author={Tasc{'o}n-Vidarte, Jos{'e} D and Wahlstedt, Isak and Jomier, Julien
and Erleben, Kenny and Vogelius, Ivan R and Darkner, Sune},
booktitle={International Workshop on Simulation and Synthesis in Medical Imaging},
pages={131--141},
year={2021},
organization={Springer}
}
Tascón-Vidarte J.D., Wahlstedt I., Jomier J., Erleben K., Vogelius I.R. and Darkner S. "Cine-MRI simulation to evaluate tumor tracking" International Workshop on Simulation and Synthesis in Medical Imaging. Springer, Cham, 2021.