Ki Moo Lim

Quantitative Analysis of Pulsatile Flow Contribution to Ultrafiltration

We evaluated the quantitative contribution of pulsatile flow to ultrafiltration (UF) in terms of fluid power, membrane stretch, and reduction of membrane layering. An in vitro comparison of the UF rate using pulsatile and roller pumps was performed with distilled water and bovine whole blood. The mean transmembrane pressure (TMPm) and UF rate were higher with the pulsatile pump for the same mean flow rate: 6.6 mm Hg and 21.1 mL/min higher on average for distilled water and 34.2 mm Hg and 31.4 mL/min higher on average for blood. The average UF rate was 8.4 mL/min higher with the pulsatile pump for the same TMPm with bovine blood. However, the relationship between the UF rate and the TMPm was independent of the flow configuration for distilled water. We showed that the higher UF rate in the pulsatile pump is mainly due to greater fluid power and reduction of membrane layering, while the membrane stretch was not an important factor.

Comparison of a Pulsatile Blood Pump and a Peristaltic Roller Pump During Hemoperfusion Treatment in a Canine Model of Paraquat Poisoning

This study examined the treatment efficacy and the damage to the blood during hemoperfusion for treating paraquat poisoning using two blood pump mechanisms. Paraquat-poisoned animal models were prepared. A conventional hemodialysis machine, AK90, with a peristaltic roller pump and a cardiopulmonary support system, T-PLS, with a pulsatile blood pump were used during the animal experiments. A total of 12 dogs were treated with hemoperfusion using a charcoal column. Six dogs were treated with hemoperfusion and T-PLS, and the other six were treated with AK90. A paraquat dose of 30 mg/kg was administrated by an intravenous injection. Both pumps maintained blood flow rates of 125 mL/min measured by an ultrasonic flowmeter. For anticoagulation, heparin was administrated by an initial bolus (250 IU/kg) and a continuous injection (100 IU/kg/h). During the experiments, T-PLS and AK90 showed a similar toxin removal efficacy. Both devices decreased the plasma paraquat concentration to 10% of the initial dose within 4-h hemoperfusion. The two pumps showed similar hemolysis properties with acceptable levels. Although T-PLS was developed as a cardiopulmonary bypass system, it can also be used as a hemoperfusion treatment device.

Numerical Simulation of the Effect of Sodium Profile on Cardiovascular Response to Hemodialysis

Purpose We developed a numerical model that predicts cardiovascular system response to hemodialysis, focusing on the effect of sodium profile during treatment. Materials and Methods The model consists of a 2-compartment solute kinetics model, 3-compartment body fluid model, and 12-lumped-parameter representation of the cardiovascular circulation model connected to set-point models of the arterial baroreflexes. The solute kinetics model includes the dynamics of solutes in the intracellular and extracellular pools and a fluid balance model for the intracellular, interstitial, and plasma volumes. Perturbation due to hemodialysis treatment induces a pressure change in the blood vessels and the arterial baroreceptors then trigger control mechanisms (autoregulation system). These in turn alter heart rate, systemic arterial resistance, and cardiac contractility. The model parameters are based largely on the reported values. Results We present the results obtained by numerical simulations of cardiovascular response during hemodialysis with 3 different dialysate sodium concentration profiles. In each case, dialysate sodium concentration profile was first calculated using an inverse algorithm according to plasma sodium concentration profiles, and then the percentage changes in each compartment pressure, heart rate, and systolic ventricular compliance and systemic arterial resistance during hemodialysis were determined. A plasma concentration with an upward convex curve profile produced a cardiovascular response more stable than linear or downward convex curves. Conclusion By conducting numerical tests of dialysis/cardivascular models for various treatment profiles and creating a database from the results, it should be possible to estimate an optimal sodium profile for each patient.

Parallel Computing Performance of a 3D Cardiac Tissue Model

We developed a three dimensional cardiac tissue model based on human cardiac cell and mono-domain approximation for action potential propagation. The human myocyte model proposed by ten Tusscher et al. (TNNP model) (2004) for cell electrophysiology and a mono-domain method for electric wave propagation are used to simulate the cardiac tissue propagation mechanism using a finite element method. To delineate the parallel computing performance of the cardiac tissue simulation, we tested a 3D cardiac tissue model based on a monodomain method. In the 3D cardiac tissue with three cell layers, we generated a reentrant wave using S1-S2 protocol.