Greg Byrne

Computational fluid dynamic simulations using patient-specific vascular geometries provide a powerful tool for understanding, treating, and preventing a variety of cerebrovascular diseases. The underlying goal of our research is to use such patient-specific modeling to connect hemodynamic variables with clinical events, thereby gaining a better understanding of intracranial aneurysm development and the mechanisms that may be responsible for their rupture. Hemodynamic variables have the potential to provide important quantitative information that can be used to improve clinical risk assessments, which, at the moment, rely largely on the use of a few image-based metrics derived from the aneurysm geometry (size, aspect ratio, etc).

Our study used 2 methods of analysis designed to provide an objective and quantitative characterization for the large-scale spatiotemporal properties of intracranial blood flows. These methods provide 2 quantities that allow us to connect specific large-scale hemodynamic structure to rupture. The first method of analysis targeted coherent swirling flow structures (vortices) that are commonly formed within the aneurysm sac as a result of recirculating blood flow. We identified these vortices by constructing 1-dimensional vortex axes known as “vortex core lines,” around which the flow rotates. These core lines provide a useful tool for visualizing the skeletal structure of the blood flow and, by estimating their total length, they also provide a simple quantitative measure to characterize the spatial flow complexity. The second method of analysis used proper orthogonal decomposition to separate the spatial and temporal dynamics of the blood flow into a weighted sum of the dominant (most energetic) spatial flow structures. Using this decomposition, we were able to estimate an entropy measure for the flow over the cardiac cycle by computing how the energy is distributed among the dominant spatial flow structures. This provides a simple quantitative measure to characterize temporal flow stability.

Using the quantities derived from our 2 methods of analysis, we classified hemodynamics according to spatial complexity and temporal stability in a database of 210 patient-specific aneurysm geometries with associated clinical histories. Our results show that complex, unstable blood flow dynamics characterized by longer core line length and higher entropy could induce biologic processes that predispose an aneurysm for rupture. These conclusions agree with previous results from a 2005 study by Cebral et al that used a well-defined qualitative approach for flow classification of the same database.

The techniques used in our study are now used routinely by our group to visualize flow structures and address ongoing research questions regarding blood flows in cerebral aneurysms. For example, one of the unexpected results from our work was the discovery of complex swirling flow structures in which a vortex becomes embedded within a larger vortex flowing in the opposite direction. The resulting bubble structure is known to be associated with a vortex breakdown process that to the best of our knowledge has not been investigated in the context of biofluids. Understanding the circumstances under which these embedded vortices form, and the role that they may play in aneurysm growth and rupture, remains unexplored and is the subject of future work.

Read this article at AJNR.org …