By Haithem Babiker, David Frakes and Fernando Gonzalez
Over the past two decades, computer simulation has become an increasingly important tool in scientific research and engineering design. In the aerospace industry for example, simulation has become the standard tool for predicting performance of a newly designed part, well before the part is physically built. In clinical medicine, computer simulation is emerging as a viable option for surgical training and planning. For example, neurosurgical simulators now allow a user to navigate a catheter through challenging patient-specific anatomy and to deploy different endovascular devices including coils, stents, and flow diverters. These new simulators represent an important breakthrough in advancing surgical training toward life-like realism, and medical device companies have also begun to use them for demonstrating the deployment of emerging endovascular devices.
Although endovascular treatment simulation has great potential to further enhance the state-of-the-art in surgical training, the greatest potential for impact lies in applying simulation to surgical planning. In that role, simulation capabilities must extend to answer important clinical questions on a patient-specific basis. Specifically, endovascular simulations must transition into predictive tools that can be used during preoperative planning to test different treatment alternatives and determine the best option for a particular patient. Aneurysm treatment exemplifies the potential impact of such tools. The growing arsenal of methods and devices for treating cerebral aneurysms, which includes stents, flow diverters, and coils, as well as a wide range of deployment strategies, calls for additional measures in preoperative planning that go beyond the capabilities of current qualitative conventions and non-interactive imaging technologies.
It would be exceedingly useful in clinical practice to have a virtual tool that could help answer questions that arise prior to treatment such as: How will additional coils, a particular coil design, or a stent affect outcome? Which device, or combination of devices, will be most effective? Extensive clinical data suggest that optimising device selection and deployment may be a highly patient-specific problem. Determining the ideal treatment regime, on a case-by-case basis, will require improved endovascular simulations that can accurately model the complexities of endovascular device geometries, their deployments, and the resulting haemodynamic outcomes.
Next generation endovascular treatment simulations are the primary focus of research at the Arizona State University Image Processing Applications Laboratory (IPALab). Specifically, IPALab is working toward advanced computational tools that can realistically simulate the mechanics of endovascular device deployment and the subsequent haemodynamic outcomes.
One example is a novel computational approach to virtual coil embolisation that uses a finite element method to simulate the complex structural dynamics of embolic coils during deployment. Each embolic coil is modelled according to a specific set of design specifications and structural properties defined by the coil manufacturer. During simulation, multiple embolic coils are deployed into a computational model of a cerebral aneurysm via virtual microcatheter. This approach produces realistic coil configurations that closely match actual physical deployments of coils in in-vitro aneurysm models, as shown in Figure 1. The computational models of the deployed coil configurations and aneurysm are then used in computational fluid dynamics (CFD) simulations to predict post-treatment blood flow within the aneurysm, as shown in Figure 2.
Using IPALab’s computational approach, the effects of different embolic coil designs and deployment strategies on aneurysm filling and blood flow can be quantified. Further, simulations can be used to understand the baseline effectiveness of coil embolisation for different types aneurysmal geometries.
Another project at the lab focuses on the computational modelling of stents and flow diverters, both as standalone devices and as structural support for embolic coils. The approach for deploying these devices is similar to that developed for coils. First, a virtual catheter is advanced to the target site using a static finite element method. Next, the stent or flow diverter is slowly unsheathed and allowed to expand across the aneurysm using a dynamic finite element method. CFD is then used to simulate fluid dynamics in and around the treated aneurysm.
Haithem Babiker is with the School of Biological and Health Systems Engineering, Arizona State University, Tempe, USA
David Frakes is with the School of Biological and Health Systems Engineering, Arizona State University, Tempe and AZ School of Electrical, Computer, and Energy Engineering, Arizona State University, Tempe, USA
L Fernando Gonzalez is with the Department of Neurological Surgery, Jefferson Medical College, Philadelphia, PA School of Biological and Health Systems Engineering, Arizona State University, Tempe, USA
