Multi-scale Modeling in Biomechanics & Biomedicine

A major emphasis of my research has been on development of computational models for physiological phenomena in patient-specific settings, with specific focus on cardiovascular biomechanics and diseases. Currently, cardiovascular diseases like stroke and heart-attack comprise one of the leading causes of death and disability worldwide. The underlying complex physical and biomechanical phenomena governing these disease occurrences are still not fully understood. High-performance computations have gained prominence as a non-invasive probe to understand this complexity. A key contribution (and ongoing focus area) of my research has been on combining image-based modeling, computational fluid dynamics, and discrete particle based methods to create novel computational models for cardiovascular diseases. The long-term goals are to employ these models to not only understand disease mechanics and etiology, but also to create patient-specific tools that inform treatment planning and biomedical device design.  
Mechanics and Etiology of Stroke and Embolisms:
Fragmented pieces of clot or calcified aggregates can get released in the bloodstream, and travel along blood vessels to cause a blockage (occlusion) of a vessel supplying a vital organ. The fragments are called “Emboli”, and an occlusion in the brain causes a stroke. We have established computational models that enables a clear understanding of the interaction of a freely moving embolus with complex hemodynamics characteristic of large arteries. Using this tool, we have characterized the distribution statistics of emboli to the brain and other vital organs, and isolated the effect of factors like embolus size and composition on its distribution to these organs. Specifically, for stroke, we have been successful in using this tool to address the complex source-destination relationship for embolic occlusion – enabling the differentiation between stroke risks due to emboli originating in the heart vs the large arteries. This knowledge is crucial for long-term treatment of patients, but was thus far an unresolved challenge.  
Multi-scale Thrombosis Models:
The pathological clotting of blood (thrombosis) is a major cause of various cardiovascular diseases including heart attack and stroke. The underlying phenomena governing thrombosis spans a broad range of length and time-scales. This multi-scale nature makes it challenging to understand macroscopic thrombus formation. We are developing a novel meso-scopic, particle-based modeling framework that enables bridging information across the various scales efficiently, while being efficient at capturing events like aggregation and fragmentation (which are difficult to handle using continuum methods). We have employed this technique, coupled with CFD, to specifically investigate the interaction of unsteady, pulsatile, hemodynamics with an arbitrarily shaped clot. We have looked at not only macroscopic clot-hemodynamics interactions, but also intra-thrombus micro-scale flows. 
Image-based CFD Modeling of Cardiovascular Fluid Mechanics:
The unsteady, pulsatile, viscous flow of blood across arbitrarily shaped arteries in real anatomical models leads to complex, nearly chaotic, flow structures - especially in the larger arteries. Aside from the specific disease scenarios described above, an over-arching focus of my work has been on understanding the nature of flow and transport in such complex large artery hemodynamics using image-based CFD as a non-invasive probe. Simulations using this tool provides a rich variety of flow and loading information. I have extensively looked at hemodynamics of the aortic arch, and have also used these simulation tools to study the flow in human Circle of Willis (ring like network connecting the six major cerebral arteries). Specifically, for the Circle, I am interested in evaluating the role of anatomical variations of incomplete circle topologies on flow, and employing information from various imaging modalities to derive accurate flow velocity information across the cerebral arteries.