SimVascular: An Open Source Pipeline for Cardiovascular Simulation

Abstract

Patient-specific cardiovascular simulation has become a paradigm in cardiovascular research and is emerging as a powerful tool in basic, translational and clinical research. In this paper we discuss the recent development of a fully open-source SimVascular software package, which provides a complete pipeline from medical image data segmentation to patient-specific blood flow simulation and analysis. This package serves as a research tool for cardiovascular modeling and simulation, and has contributed to numerous advances in personalized medicine, surgical planning and medical device design. The SimVascular software has recently been refactored and expanded to enhance functionality, usability, efficiency and accuracy of image-based patient-specific modeling tools. Moreover, SimVascular previously required several licensed components that hindered new user adoption and code management and our recent developments have replaced these commercial components to create a fully open source pipeline. These developments foster advances in cardiovascular modeling research, increased collaboration, standardization of methods, and a growing developer community.

Keywords: Hemodynamics; Image-based CFD; Open-source; Patient-specific modeling.

FIGURE 1.

Inheritance diagram of cvRepositoryData. Derived classes in aqua are open source while derived classes in gray are commercial and optional.

FIGURE 2.

The SimVascular pipeline leads the user from visualization of image data through to completion of blood flow simulations. Steps 2–4 correspond to the lofted 2D segmentation process. Adapted from Ref. .

FIGURE 3.

The SimVascular pipeline is mirrored in the main work tabs of the GUI (enclosed in red box). Paths → Segmentation → Model → Meshing → Simulations.

FIGURE 4.

Creation of a vascular geometry using the lofted 2D segmentation approach involves moving a cross-sectional image window along each vessel path (a) to create a series of segmentations (b) that are lofted to form each vessel (c). A solid model is generated by the union of individual vessel models (d).

FIGURE 5.

Left: A slice along the vessel pathline is segmented using level set segmentation techniques. Right: The same slice is segmented using threshold techniques.

FIGURE 6.

A geometry imported into SimVascular and prepared for meshing using the PolyData solid model package. (1) The imported geometry (2) Extra and undesired portions of the geometry are removed and holes are filled (3) The geometry is smoothed, decimated, and subdivided.

FIGURE 7.

A variety of meshing options are available in SimVascular. (a) Uniformly prescribed element size on mesh (b) Boundary layer mesh (c) Mesh with spherical refinement (d) Radius-based mesh.

FIGURE 8.

On the left, “open-loop” boundary conditions are prescribed on a model of an aorta (from Ref. 29). RCR circuits are applied to represent the downstream vasculature. On the right, “closed-loop” boundary conditions are applied to a Hemi-Fontan model (from Ref. 25).

FIGURE 9.

In vivo validation of SimVascular’s finite element flow solver for aortic coarctation. (left) Comparison of fluctuation intensity (TKE) fields from PCMRI and from SimVascular (CFD) during systole. (right, top) Percentage of the descending aorta (boxed region) with fluctuation intensity above various thresholds at systole. (right, bottom) Integral of the fluctuation intensity field over the descending aorta (boxed region) vs. time. Figures adapted from Ref. .

FIGURE 10.

A sampling of the wide variety of model categories and simulation results available online in the vascular model repository at http://www.vascularmodel.com.

FIGURE 11.

The vascular mode repository combines the results of over 100 studies of varying image data, model complexity, and simulation type.

FIGURE 12.

The image volume and the constructed model for the aortic and femoral arteries of a 21 year old female (left two panels). Representative simulation results of the time-averaged pressure field and oscillatory shear index (OSI) field are shown (right two panels).

FIGURE 13.

Time averaged wall shear stress and average pressure over one cardiac cycle for the pulmonary arteries of a 67 year old woman.

FIGURE 14.

Model of the aorta, coronary arteries, and a bypass graft constructed from CT. Velocity magnitude volume render during end diastole (left) and wall displacements during peak systole calculated from FSI simulation (right). Adapted from Ref. .