Generation, computational biofluid mechanics, and visualization of complex blood flow dynamics

Pablo Huang Zhang

doi:10.17918/nbwp-et44

The advent of new non-invasive imaging modalities (i.e. 4D MRI, 3D Echocardiography) in recent years have facilitated the study and growing recognition that some of the blood flow in the cardiovascular system is naturally spiral and three-dimensional. The helical organization of the myocardial fibers, the heart's torsional contraction dynamics, aortic valve structure, the out-of-plane geometry of the aorta and tortuosity of vessels all contribute to the generation of spiral patterns of blood flow. In nature, many forms of fluid transport (e.g. whirlpool, cyclones) demonstrate high efficiency, flow entrainment, and stability due to their spirality. Flow in the cardiovascular system may also benefit from similar self-stabilizing impulsion dynamics. Although spiral blood flow structures have been observed in the aorta and other large arteries, many questions remain unanswered regarding its influence on normative cardiovascular physiology and pathophysiology. The research work herein aims to study spiral flow dynamics and to understand its specific characteristics, especially those in athero-susceptible regions. Computational fluid dynamics (CFD) was used to study the modulation of spiral flow and its impact in idealized vascular phantoms (Aim 1) and realistic vascular geometries, namely the aortic arch with an anastomosed cannula, representative of the outflow graft of a mechanical circulatory support device (Aim 2). Aim 1 served as a test platform for studying spiral flow characteristics. Aim 2 provided an example of the translational applicability of spiral flow. Benchtop flow circuits were used to validate key aspects of the in-silico simulations. This research work brought together computational fluid dynamics with 3D vascular printing and benchtop mock circulatory flow loop visualization and analysis methodologies. The ability of spiral flow to clear and reduce the size of recirculation zones in a set of idealized vascular phantoms was demonstrated in Aim 1. The phantoms tested were angled conduits with 45°, 90°, and 135° turns and idealized asymmetric and axisymmetric stenosis models. A spiral flow inducer was utilized to enable in-silico to in-vitro comparisons, while standalone phantoms were used to test the impact of spiral flow modulation. In the vascular phantoms coupled to spiral flow inducer models, the recirculation zones at the corners of the angled conduits and the flow separation post-coarctation in stenosis models demonstrated a marked decrease in size of regions of low velocities (< 5 cm/s) and low wall shear stress (WSS < 3 dyn/cm2). In these idealized stenosis models, spiral flow diminished the post-coarctation reattachment length by up to 45%. These results were validated using a benchtop experimental setup incorporating ultrasound velocity visualization and 3D printed parts. Simulations with standalone vascular phantoms demonstrated that the recirculation zones decreased as the helical content increased. In the 45° angled conduit, there was a 6.3-fold decrease in volume of thresholded low velocities and 1.8-fold decrease in low WSS areas when comparing straight flow to spiral flow with lamda=L* (wavelength, lamda, equal to the characteristic length, L*, defined for each phantom). For the 90° case there was a 2.6-fold decrease in low velocities and a 5.4-fold reduction in low WSS between straight and lamda=L*. In the 135° case, there was a 4.2-fold decrease in low velocity volumes, and a 3.2-fold decrease in low WSS areas between straight and lamda=L*. Compared to straight flow, spiral flow exhibited a swirling, non-colliding, and enhanced washout dynamics in the vascular phantoms tested. CFD simulations in Aim 2 demonstrated that the out-of-plane geometry of the aorta itself induced spiral forms of blood flow. In patient-specific aorta models, each model had a distinct flow signature that was affected depending on the spiral inlet conditions, which impacted regions of low velocities and areas of low WSS. Results from an aortic model with a virtually anastomosed cannula, representative of a mechanical circulatory support device outflow cannula, demonstrated a 1.5-fold decrease of low WSS (< 3 dyn/cm2) regions with increasing spiral flow content, but also showed a 1.4-fold average increase in areas of high WSS (> 80 dyn/cm2) for the highest helical content compared to straight flow. The highest WSS at the fluid impact site in the inner curvature of the aorta belonged to the test case with the highest helical content, reaching 150 dyn/cm2. Cannula angle variation dictated the impact site of the outflow jet. In all cases, counter-clockwise spiral flow decreased regions of low velocity (up to 1.2-fold reduction) and resultant areas of low WSS (up to 1.1-fold decrease). The cannula angled down case provided the best decrease in low WSS areas, however, it had the largest area of high WSS. With clockwise and counter-clockwise spiral flow, the areas of high WSS in the cannula angled down case decreased by 10.5% and 29.8% respectively. Spiral flow has been shown to improve washout of hard-to-reach recirculation zones, reducing regions of low velocity and decreasing areas of low WSS. In particular, the clinical translation may prove to be impactful in blood recirculating devices, helping improve near-wall transport and flow dynamics, diminishing jetting and fluid collisions, mitigating device-related adverse events, and encouraging athero-protective conditions. The findings of this research are expected to inform the next generation engineering designs of vascular/endovascular prosthesis, stents, cardiac valves, and mechanical circulatory support devices.

Generation, computational biofluid mechanics, and visualization of complex blood flow dynamics

Files and links (1)

Abstract

Metrics

Details

Generation, computational biofluid mechanics, and visualization of complex blood flow dynamics

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media