Principal Investigator John Williams
Discrete Element Modeling (DEM) has become an important tool in investigating a variety of discontinuous systems, including pharmaceutical powder processing, granular flow, and coupled solidfluid flow for geotechnical systems. However, little application of DEM has been seen in biological systems modeling. This poster will offer a few areas in which DEM can be applied as well as an overview of the computational capabilities of the method in general and the authors' applications specifically (see also http://portals.mit.edu/portalfactory -> DEM).
Studying the macroscopic process of some vascular diseases is well-suited for coupled DEM simulation approaches. For instance, thrombus formation is an important contributing factor to many cardiovascular diseases. Thrombi are typically initiated by damage to the endothelium layer of the vascular wall, which exposes collagen and insoluble vFW (von Willebrand Factor). When the vFW is activated, platelets transiently bind to the vFW based on proximity to the vFW as well as other state factors, such as the relative velocity of the platelet to the endothelium. Adhesion of platelets is facilitated through the joint action of GPIIb/IIIa, vFW, and Fibrinogen. This model of platelet adhesion can be captured well using DEM, taking advantage of the method's suitability for describing behavior based on spatial proximity-based interaction and the ability to couple fluid behavior (for both laminar and turbulent flow regimes) with granular solid transport. Studies with DEM in 2D have shown promising results at helping to model the macroscopic behavior of platelets (see Longest, 2003; Yano, 2003; and Miyazaki, 2003). Similar processes of platelet transport are important in the formation of stenotic lesions and mural thrombi.
The framework of DEM is flexible and allows the researcher to adjust the numbers and resolution of the parameters of interest. For instance, simple geometries such as spheres can be used; however, if it is discovered that particle shape has an effect on the behavior of interest, the particle shape can be approximated by ellipsoids or by arbitrarily-shaped polyhedra. If particle deformability is of interest, DEM can be coupled with FEM to capture individual grain deformation. DEM can even be altered such that the contact model gives each grain its own behavior (which may even be memory based) for applications such as protein binding and disease migration between spatially disjoint cells. Relaxing the view of DEM as a basic physical model, the underlying algorithms can be adapted to problems such as disease transmission between individuals or populations.
Modeling of systems as discontinua can yield important insight into the underlying behavior of the system, and DEM can provide a tool for probing this behavior.