Session L07: Medical Devices II

In-vitro investigation of the influence of aneurysmal sac morphology on the hemodynamics of cerebral aneurysms treated with flow diverting stents

Flow diverting stents (FDS) are a common cerebral aneurysm treatment. These high-porosity meshes aim at reducing flow into the aneurysm, promoting re-endothelization, and leading to the formation of a stable thrombus filling the aneurysmal sac. 25% of procedures fail, requiring follow-up and re-treatment. We investigate the effect of the aneurysm geometry on the efficacy of FDS at reducing flow into the aneurysm, via Stereo Particle Image Velocimetry (2D-3C PIV) measurements in 12 idealized aneurysm flow phantoms. Parent vessel curvature, neck size and sac aspect ratio were varied. These models were set up in a flow loop, before and after treatment, at physiological flow rates, corresponding to different Dean and Womersley numbers. A single counter-rotating vortex, whose circulation correlates with the Dean number, was observed for all untreated aneurysms. Aspect ratio variations yielded different post-treatment flow patterns: inversion of vorticity during the cycle, counter-, co-rotating vortex. Depending on the shape, the post-treatment circulation ranged from 65% to only a few % of the pre-treatment value. This study highlights the influence of some parameters: high impact of the morphology and negligible for Wo, helping to predict better the risk of treatment failure.   

Flow Field Dynamics of a Wide-Neck Intracranial Aneurysm With and Without the Presence of Double-Wall Vascular Stent

An experimental study is performed to investigate the flow dynamics of a wide-neck intracranial aneurysm with and without the presence of a double-wall vascular stent. A laboratory model of a lateral side-wall aneurysm 3 cm in length and with a 10 mm wide-neck was fabricated using clear plexiglass. The aneurysm model was placed in a circuit driven by a peristaltic pump that delivered water cycling at the rate of one pulse per second. The fluid flow rate was measured using a Honeywell vortex shedding flow meter and was kept constant at 200 mL/min. The flow field was captured by means of stereoscopic Particle Image Velocimetry. The illumination is provided by a dual-oscillator Nd:YAG laser with 30 mJ/pulse at 532nm wavelength with a maximum repetition rate of 10kHz. A light sheet of 1 mm thickness is created by suitable combination of spherical and cylindrical lenses. The images are recorded by two CMOS cameras (Phantom VEO 410L) with a resolution 1280(H) x 800(V) pixels of a size of 20 microns. A total of 2500 image pairs per second were acquired. The results show that the double-wall vascular stent greatly reduced flow recirculation and wall shear stress. The study provides a basis for the future feasibility of double-wall vascular stents in preventing aneurysm rupture.   

Computational assessment of the hemodynamics and platelet activation potential of a Trileaflet Mechanical Heart Valve and their comparison with a Bioprosthetic and a Bileaflet Mechanical Heart Valve

Albeit thousands of patients undergo prosthetic heart valve implantations each year, a coagulation-free, durable, and biocompatible prosthetic valve remains elusive. Novel designs of prosthetic heart valves are therefore required to be analyzed to ensure a physiological function of the heart post the implantation of the prosthesis. In this research, the hemodynamics and platelet activation potential of a Trileaflet Mechanical Heart Valve, named as Triflo Valve, have been investigated and compared against existing designs of a Bioprosthetic Heart Valve (BHV) and a Bileaflet Mechanical Heart Valve (BMHV). The flow field of the Triflo Valve and the BHV is dominated by a central jet being void of eddies, with a triangular cross-section, which resembles physiological flow, contrary to the BMHV having two strong lateral jets. The Total Platelet Activation (TPA) in the Triflo Valve was observed to be 14% higher than BHV whereas that for the BMHV was 73% higher. The Triflo Valve exhibits around 52% lesser TPA in comparison to BMHV and demonstrates promising features for its consideration towards aortic valve replacements.   

Author not Attending

The ongoing epidemic of advanced heart failure in the United States has caused a sharp rise in the utilization of heart pumps, known as left ventricular assist devices (LVADs), over the last decade. Despite technological improvements in the current generation of LVADs, LVAD-supported patients remain prone to complications resulting from enormously altered hemodynamics, which could contribute to a stroke or mortality while on device support. The HeartMate III is the latest generation centrifugal continuous LVAD which initiates an artificial pulse to aid in improving blood circulation by pumping blood from the left ventricle to the aorta. Studies, to date, on the HeartMate III are very limited and explore the altered hemodynamics effect on blood flow only at a specific RPM and flow rate and using typically wall shear stress (WSS). This study entails the evaluation of not only WSS for various conditions, but also identifies the contribution of temporal variation of RPM on WSS and WSS-related parameters using high-fidelity simulations. Harmonic analysis will be conducted to understand the contribution of various frequencies generated by the impeller rotation, interaction of flow around impeller's blades, as well as temporal variation of RPM on WSS and WSS-related parameters.   

Flow-induced stress distribution on von Willebrand factor protein molecules in turbulent flow

Under the effects of non-physiological hemodynamic stresses, von Willebrand factor (vWF) protein molecules that circulate witin the blood are known to unravel causing loss of efficacy.^1,2 In this study, the distributions of shear and extensional stress along the trajectories of passive particles that serve as surrogates of vWF molecules were calculated. Poiseuille and Poiseuille-Couette flows in conditions that correspond to blood pumps and transitional flow in ventricular assist devices, respectively, were simulated with the use of Lagrangian methods and direct numerical simulations. It was found that the distribution of hydrodynamic stress was a function of the location where the vWF particles were released into the channel and of the exposure time. Apart from the distribution of stresses, the history of the stresses was also computed allowing the evaluation of how many particles experience stresses above the critcal value for protein damage.  This type of work can contribute to the design of medical devices.

References

1. Springer, T.A., Blood. 124, 1412  (2014).

2. Pham et al., Scientific Reports 12, 171 (2022)

    

Experimental Quantification of Hemolysis Through ECMO System Oxygenator

Extracorporeal Membrane Oxygenation (ECMO) is the final attempt to save a patients’ life. Before 2020, the administration of ECMO was primarily used for the treatment of Acute Respiratory Distress Syndrome or other severe cardio-pulmonary conditions, but has recently been used to treat COVID-19 patients. It has been argued that hemolysis is an important issue when using this system, however the details are unknown. The aim of this study is to quantify hemolysis induction through the oxygenation module. An experimental setup was created by scaling up the geometry of the membrane to conduct particle image velocimetry on the system. Newtonian and non-Newtonian fluids were used, both of which were Reynolds number-matched to the clinical system. Two-dimensional vector fields were obtained for several planes of the device over a range of flow rates. Upon calculation of the viscous shear stress from the velocity fields, hemolysis induction predictions were calculated considering a model for blood damage. For the Newtonian experiments, an early onset of hemolysis was detected only after considering long exposure times. Results from the non-Newtonian experiments also lead to small levels of induced hemolysis suggesting that the oxygenator module causes very little to no blood damage.

    

A cylindrical-layer domain scaling approach to modeling urea clearance in a hemodialyzer

A hemodialyzer (artificial kidney) typically consists of a cylindrical housing containing about 10,000 parallel hollow fibers that carry blood along the axis of the housing, while the dialyzer fluid (dialysate) flows around the fibers in the opposite direction. Uremic toxins are cleared from the blood by diffusing across the fibers' semipermeable membranes into the dialysate. It is too computationally costly to simulate the flow in the entire dialyzer in three dimensions. As a result, most simulations either extrapolate single-fiber results to the entire flow domain or use a porous medium approach, and miss the interfiber flow details. Here, we employed a domain scaling approach to model urea clearance in hemodialyzer models of increasing complexity. We developed 1-, 7-, 19-, 37-, 61-, 91- and 127-fiber 3D models of a commercial hemodialyzer. We performed simulations using clinically-relevant blood (100-500 mL/min) and dialysate flow rates (200-800 mL/min) and found the urea clearance increased as the dialysate flow rate increased at a constant blood flow rate, similar to known clinical results. This suggests that the cylindrical-layer domain scaling approach may capture sufficient fiber-fiber interactions to accurately model the steady flow physics in a full-scale hemodialyzer.   

Toward breath sensors that are self-powered by design

Piezoelectric materials are widely used to generate electric charge from mechanical deformation or vice versa. These strategies are increasingly common in implantable medical devices, where sensing must be done on small scales. In the case of a flow rate sensor, a sensor's energy harvesting rate could be mapped to that flow rate, making it "Self-Powered by Design (SPD)". Prior fluids-based SPD work has focused on turbulence-driven resonance and has been largely empirical. Here we explore sub-resonance SPD sensing via a case study of human breathing. We present a model of self-powered piezoelectric sensing/harvesting and validate that model against experimental results. We evaluate how sensor size scaling with Reynolds number impacts its performance as we approach implantable scale, at which we are conducting in-vivo testing in a rabbit model and have detected signs of a methylcholine-induced asthma attack from within the airway. We offer a form of SPD sensing that scales down to micro- or nano-scales, where flows are locally laminar and wake-driven resonance is not an option, and a model-based roadmap for future SPD sensing solutions, including detecting broadband flow information.   

Use of drops and jets for front-of-the-eye drug delivery

In this talk, we will present results from experimental work on fluid delivery to the eye, targeted toward ophthalmic drug delivery. It is known that front-of-the-eye (FOTE) delivery using standard "dropper bottles" are very inefficient due to incorrect usage, the blink reflex, and fast drainage via the nasolacrimal duct. To circumvent these issues, it has been proposed that smaller, single droplets of polymeric solution (to increase viscosity and adhesion) may improve outcomes in terms of bioavailability. Therefore, we have explored the dynamics of droplet impact on acrylic eyeball targets, which can be dry or pre-wetted with a solution of artificial tears. Here, we will summarize those findings, along with an interesting observation of non-coalescence. We also investigate the use of discrete pulsed jets as a novel method of delivery.   

Rutherford device to measure deformability of red blood cells

Red blood cells (RBCs) lose their deformability in response to many environmental and physiological cues. We present a simple microfluidic device, consisting of a channel that opens up to a funnel behind a  single obstacle. By tracking the trajectories of single RBCs after they pass the obstacle we are able  to measure their Young’s modulus.  Softer RBCs and stiffer RBCs emerge along different streamlines in our device. We calibrate the trajectories of healthy and artificially stiffened RBCs in this microfluidic device against the Young’s moduli obtained from AFM measurements of these RBCs. Further, we demonstrated that healthy and malaria-infected RBCs follow different trajectories in our device.