Here we seek to improve the accuracy of hemodynamic modeling of coiled cerebral aneurysms. This goal is significant due to the prevalence of cerebral aneurysms, their dismal prognosis when ruptured, and treatment failure rates (resulting in aneurysm recurrence and risk of either brain hemorrhage or need for retreatment) of up to 25%. Hemodynamic forces are thought to influence aneurysm treatment outcomes, but the standard method of computational fluid dynamics (CFD) modeling of such forces within coiled aneurysms (termed the ?porous medium technique?) is error-prone. Improving the accuracy of CFD modeling of coiled aneurysms will strengthen the predictive value of patient-specific CFD, which could improve aneurysm treatment efficacy and reduce death and disability, as well as health care costs associated with multiple hospitalizations. This project builds on our ongoing NIH-funded expertise at creating CFD models of brain aneurysms, and a partnership with the European Synchrotron Research Facility, to develop an improved method of CFD modeling of coiled cerebral aneurysms that can be applied in a clinical setting. First, we will create high-fidelity 3D-printed aneurysm models based on patient-specific aneurysm anatomy, and place the same commercially- available aneurysm coils used in actual patient treatment into each model. These coiled aneurysm models will be scanned at 12 m resolution using synchrotron x-ray microtomography, providing detailed 3D images of the complex coil geometry. These images will be incorporated into CFD models of clinically relevant hemodynamic variables, and will be considered a reference standard to which other modeling techniques are compared. Then, we will create a new set of CFD models of the same aneurysms, using the standard porous medium technique to represent the coil mass. This technique simplifies the complex coil geometry into a material of uniform porosity, which our preliminary analysis suggests is a source of significant error in the calculation of hemodynamic variables. We will quantify this error by comparing these CFD models to the reference standard CFD models created using microtomography. Then, we will employ the homogenization of multiple scale expansions technique, in which the complex structure of the coil mass is represented by macroscopic equations that better approximate permeability. We will develop a set of corrective factors (a ?coil modeling toolkit?) that can be used in future CFD models of coiled aneurysms with better accuracy than the standard porous medium technique. Finally, we will determine the improved accuracy of this technique by using the coil modeling toolkit to create CFD models of a new set of aneurysms, for which 3D-printing and microtomography are not required. We will compare these results to the reference standard (both using CFD and using in vitro flow measurements through 3D-printed models) and quantify the improvement in accuracy gained using the coil modeling toolkit. This improved accuracy will strengthen the clinical impact of CFD studies of aneurysm treatment.
The proposed study investigates how to improve the accuracy of computer modeling of coiled cerebral aneurysms. The typical method of creating computer models of coils within an aneurysm (called the porous medium method) relies on assumptions of how blood flow interacts with the coils. To determine how accurate this technique is, we will compare it to a reference standard of high-resolution scans of 3D-printed models of coiled cerebral aneurysms (containing actual aneurysm coils) created using synchrotron microtomography. Then, we will create an improved, novel method of approximating aneurysm coils using the homogenization of multiple scale expansions, and evaluate the new method?s accuracy; this method will improve our ability to measure clinically meaningful blood flow changes in patients with coiled aneurysms in the future.
|Chivukula, Venkat Keshav; Levitt, Michael R; Clark, Alicia et al. (2018) Reconstructing patient-specific cerebral aneurysm vasculature for in vitro investigations and treatment efficacy assessments. J Clin Neurosci :|