CO₂ removal via membrane oxygenators during lung protective ventilation has become a reliable clinical technique. For further optimization of oxygenators, accurate prediction of the CO₂ removal rate is necessary. It can either be determined by measuring the CO₂ content in the exhaust gas of the oxygenator (sweep flow-based) or using blood gas analyzer data and a CO₂ solubility model (blood-based). In this study, we determined the CO₂ removal rate of a prototype oxygenator utilizing both methods in in vitro trials with bovine and in vivo trials with porcine blood. While the sweep flow-based method is reliably accurate, the blood-based method depends on the accuracy of the solubility model. In this work, we quantified performances of four different solubility models by calculating the deviation of the CO₂ removal rates determined by both methods. Obtained data suggest that the simplest model (Loeppky) performs better than the more complex ones (May, Siggaard-Anderson, and Zierenberg). The models of May, Siggaard-Anderson, and Zierenberg show a significantly better performance for in vitro bovine blood data than for in vivo porcine blood data. Furthermore, the suitability of the Loeppky model parameters for bovine blood (in vitro) and porcine blood (in vivo) is evaluated.

Computational fluid dynamics (CFD) provides a flexible tool for investigation of separation processes within membrane hollow fiber modules. By enabling a three-dimensional and time dependent description of the corresponding transport phenomena, very detailed information about mass transfer or geometrical influences can be provided. The high level of detail comes with high computational costs, especially since species transport simulations must discretize and resolve steep gradients in the concentration polarization layer at the membrane. In contrast, flow simulations are not required to resolve these gradients. Hence, there is a large gap in the scale and complexity of computationally feasible geometries when comparing flow and species transport simulations. A method, which tries to cover the mentioned gap, is presented in the present article. It allows upscaling of the findings of species transport simulations, conducted for reduced geometries, on the geometrical scales of flow simulations. Consequently, total transmembrane transport of complete modules can be numerically predicted. The upscaling method does not require any empirical correlation to incorporate geometrical characteristics but solely depends on results acquired by CFD flow simulations. In the scope of this research, the proposed method is explained, conducted, and validated. This is done by the example of CO₂ removal in a prototype hollow fiber membrane oxygenator.

Extracorporeal membrane oxygenators are essential medical devices for the treatment of patients with respiratory failure. A promising approach to improve oxygenator performance is the use of microstructured hollow fiber membranes that increase the available gas exchange surface area. However, by altering the traditional circular fiber shape, the risk of low flow, stagnating zones that obstruct mass transfer and encourage thrombus formation, may increase. Finding an optimal fiber shape is therefore a significant task. In this study, experimentally validated computational fluid dynamics simulations were used to investigate transverse flow within fiber packings of circular and microstructured fiber geometries. A numerical model was applied to calculate the local Sherwood number on the membrane surface, allowing for qualitative comparison of gas exchange capacities in low-velocity areas caused by the microstructured geometries. These adverse flow structures lead to a tradeoff between increased surface area and mass transfer. Based on our simulations, we suggest an optimal fiber shape for further investigations that increases potential mass transfer by up to 48% in comparison to the traditional, circular hollow fiber shape.

Blood pumps have found applications in heart support devices, oxygenators, and dialysis systems, among others. Often, there is no room for sensors, or the sensors are simply unreliable when long-term operation is required. However, control systems rely on those hard-to-measure parameters, such as blood flow rate and pressure difference, thus their estimation takes a central role in the development process of such medical devices. The viscosity of the blood not only influences the estimation of those parameters but is often a parameter that is of great interest to both doctors and engineers. In this work, estimation methods for blood flow rate, pressure difference, and viscosity are presented using Gaussian process regression models. Different water–glycerol mixtures were used to model blood. Data was collected from a custom-built blood pump, designed for intracorporeal oxygenators in an in vitro test circuit. The estimation was performed from motor current and motor speed measurements and its accuracy was measured for: blood flow rate r² = 0.98, root mean squared error (RMSE) = 46 mL^.min⁻¹; pressure difference r² = 0.98, RMSE = 8.7 mmHg; and viscosity r² = 0.98, RMSE = 0.049 mPa^.s. The results suggest that the presented methods can be used to accurately predict blood flow rate, pressure, and viscosity online.

CO₂ removal via membrane oxygenators has become an important and reliable clinical technique. Nevertheless, oxygenators must be further optimized to increase CO₂ removal performance and to reduce severe side effects. Here, in vitro tests with water can significantly reduce costs and effort during development. However, they must be able to reasonably represent the CO₂ removal performance observed with blood. In this study, the deviation between the CO₂ removal rate determined in vivo with porcine blood from that determined in vitro with water is quantified. The magnitude of this deviation (approx. 10%) is consistent with results reported in the literature. To better understand the remaining difference in CO₂ removal rate and in order to assess the application limits of in vitro water tests, CFD simulations were conducted. They allow to quantify and investigate the influences of the differing fluid properties of blood and water on the CO₂ removal rate. The CFD results indicate that the main CO₂ transport resistance, the diffusional boundary layer, behaves generally differently in blood and water. Hence, studies of the CO₂ boundary layer should be preferably conducted with blood. In contrast, water tests can be considered suitable for reliable determination of the total CO₂ removal performance of oxygenators.

Membrane oxygenators are devices that benefit from automatic control. This is especially true for implantable membrane oxygenators—a class of wearable rehabilitation devices that show high potential for fast recovery after lung injury. We present a performance comparison for reference tracking of carbon dioxide partial pressure between three control algorithms—a classical proportional-integral (PI) controller, a modern non-linear model predictive controller, and a novel deep reinforcement learning controller. The results are based on simulation studies of an improved compartmental model of a membrane oxygenator. The compartmental model of the oxygenator was improved by decoupling the oxygen kinetics from the system and only using the oxygen saturation as an input to the model. Both the gas flow rate and blood flow rate were used as the manipulated variable of the controllers. All three controllers were able to track references satisfactorily, based on several performance metrics. The PI controller had the fastest response, with an average rise time and settling time of 1.18 s and 2.24 s and the lowest root mean squared error of 1.06 mmHg. The NMPC controller showed the lowest steady state error of 0.17 mmHg and reached the reference signal with less than 2% error in 90% of the cases within 15 s. The PI and RL reached the reference with less than 2% error in 84% and 50% of the cases, respectively, and showed a steady state error of 0.29 mmHg and 0.5 mmHg.