Design and Development of a Computational Tool for a Dialyzer by Using Computational Fluid Dynamic (CFD) Model

Abstract

In order to reduce the hemodialysis cost and duration, an investigation of the effect of dialyzer design and process variables on the solute clearance rate is required. It is not easy to translate the in vivo transfer process with in vitro experiments, as it involves a high cost to produce various designs and membranes for the dialyzer. The primary objective of this study was the design and development of a computational tool for a dialyzer by using a computational fluid dynamic (CFD) model. Due to their complexity, only researchers with expertise in computational analysis can use dialyzer models. Therefore, COMSOL Inc. (Stockholm, Sweden) has made an application on membrane dialysis to study the impact of different design and process parameters on dialyzed liquid concentration. Still, membrane mathematical modeling is not considered in this application. This void hinders an investigation of the impact of membrane characteristics on the solute clearance rate. This study has developed a stand-alone computational tool in COMSOL Multiphysics 5.4 to fill this void. A review of the literature conducted shows that there are no suitable stand-alone computational tools for kidney dialysis. Very little work has been undertaken to validate the stand-alone computational tool. Medical staff in the hospitals require a computational tool that can be installed quickly and provide results with limited knowledge of dialysis. This work aims to construct a user-friendly computational tool to solve this problem. The development of a user-friendly stand-alone computational tool for the dialyzer is described thoroughly. This application simulates a mathematical model with the Finite Element Method using the COMSOL Multiphysics solver. The software tool is converted to a stand-alone version with the COMSOL compiler. The stand-alone computational tool provides the clearance rate of six different toxins and module packing density. Compared with the previous application, the stand-alone computational tool of membrane dialysis enables the user to investigate the impact of membrane characteristics and process parameters on the clearance rate of different solutes. The results are also inconsistent with the literature data, and the differences ranges are 0.09–6.35% and 0.22–2.63% for urea clearance rate and glucose clearance rate, respectively. Statistical analysis of the results is presented as mean with 95% confidence intervals (CIs) and p values 0.9472 and 0.833 of the urea and glucose clearance rates, respectively.