Optimization of the IntraVascular Oxygenator Catheter Using Angular Oscillation

Abstract

We demonstrate a methodology which both improves oxygen transport and reduces or eliminates bubble formation in a novel hyperbaric membrane oxygenator catheter model system. Angular oscillations were introduced to a bundle of hollow fiber membranes (HFMs) supplied with hyperbaric 100% oxygen at average gauge pressures up to 0.35 barg. Oscillating bundles enabled delivery of an oxygen flux of up to 400 mL min^-1 m^-2 in an aqueous solution, a doubling over a previous non-oscillating setup. Similarly, the addition of angular oscillations facilitated a five-fold reduction in pressure to achieve similar oxygen flux. The increased angular speed of oscillation improved flux, while the addition of angular micro-oscillation variations resulted in flux reductions of 7-20% compared to continuous macro-oscillation only, depending on mixing conditions. However, semi-quantitative visual observation demonstrated that angular oscillations reduced or eliminated the instance of oxygen bubble formation on the HFMs. The modeled mass transfer coefficients indicated a quasi linear relationship between rotational velocity and flux, suggesting that faster oscillation speeds could further improve oxygen mass transport allowing for HFM bundles to maintain high oxygen fluxes while eliminating bubble formation. This encourages further development of our compact oxygenating catheter that could be used intravascularly.

Keywords: Acute respiratory distress syndrome; Artificial lung; Bubbleless aeration; Hypoxic respiratory failure; Intravascular oxygenator catheter.

Figure 1:

Angular position over time for an example mixing regime that uses 45 degree macro steps (α) separated every 1 second with 22.5 degree micro oscillations (β) at a speed of 3200 deg s⁻¹

Figure 2:

Process schematic of bench top system using water and blood mimic. The ECMO membrane lung is used to maintain temperature at 37 °C and to remove oxygen before the system inlet.

Figure 3:

Number of bubbles counted on fiber C with (blue) and without (green) micro-oscillations. Mix 1 and 5 had angular velocities of 3200 deg/sec and mix 3 and 7 had 720 deg/sec angular velocity. All data collected on a single fiber loop in a liquid flow rate of 0.5 L min⁻¹. The shaded areas are visual guides to illustrate that addition of micro-oscillations reduced bubble formation.

Figure 5:

Flux of fiber bundle C (32 total fibers, 14.6 cm each) at various mix regimes at high (0.345 barg) and low (0.147 barg) pressures under two liquid flow conditions (0.5 and 2.0 LPM).

Figure 6:

Dissolved oxygen model data vs. actual for the system outlet with fiber bundle C (A) and fiber bundle A (B) at 2 LPM. Fiber C was used to train the model and fiber A used to validate the model. Mixing with micro oscillations (5, 6 and 7) were notably underpredicted in all but one case.

Figure 7:

Fitted mass transfer coefficients (k) vs. angular speed of mixing for macro and macro + micro oscillation mix methods. The mass transfer coefficients with micro-oscillation were different than those without micro-oscillation based on model convergence during experimental data fitting.