We use DICOM2FEM application ([1]) for semi-automatic generation of the finite element mesh of the human liver from computed tomography scans. The input is a CT data set stored in a serie of DICOM files, the pydicom package is used to import the volumetric data, see [2]. The segmentation method is based on the graph cut algorithm (pygco library, [3]) and machine learning algorithms from scikit-learn. The Taubin smoothing algorithm and Marching cubes algorithm are implemented in order to produce more realistic geometrical models of body parts.
FE model of the human liver including portal and hepatic vessel trees:
The generated FE mesh is imported into SfePy (Simple Finite Elements in Python, see [4]) and numerical simulations are performed in order to get the pressure and flow distribution in the liver tissue. The mathematical model of tissue perfusion is based on hierarchical and multicompartment models, cf. [5, 6].
Concentration of contrast fluid in liver parenchyma:
[1] http://sfepy.org/dicom2fem
[2] http://code.google.com/p/pydicom
[3] https://github.com/amueller/gco_python
[4] http://sfepy.org
[5] C. Michler et al. A computationally efficient framework for the simulation of cardiac perfusion using a multi-compartment Darcy porous-media flow model. In International Journal For Numerical Methods In Biomedical Engineering, Vol. 29, 2013. http://dx.doi.org/10.1002/cnm.2520
[6] E. Rohan, V. Lukeš, A. Jonášová and O. Bublík. Towards microstructure based tissue perfusion reconstruction from CT using multiscale modeling. In Proceedings of the 10th World Congress on Comput. Mechanics, Sao Paulo, 2012.
