Computational Analysis of a Membranebased Separation Process in an Artificial Kidney

Abstract

Patients with kidney failure must undergo long, painful, costly weekly hemodialysis sessions until a donor is found. Reducing the process's time and cost means improving each hemodialysis session's efficiency. Mathematical models are tools used to evaluate the relationship between parameters. Previous models used complex mathematical equations that needed an understanding of various subjects. A user-friendly app is developed using MATLAB R2022 that evaluates parameters, defines their relationship, and makes it convenient to be used by both healthcare professionals and researchers in engineering. The model evaluated the effect of blood flow rate on clearance and predicted that this model had less percentage difference with experimentally measured clearance than previous in-silico models. Also, as the value of the dialysate flow rate increases from 200 mL/min to 1000 mL/min, the percentage difference between the current and previous models reduces at first but then increases at a constant rate. According to the single pool model, drop-in Urea concentration with time is rapid initially but becomes minute towards the end of the session. At a dialysate flow rate of 400 mL/min, the percentage difference between the clearance of the co and counter-current dialyzer is maximum. It starts to decrease as the dialysate flow rate increases. With increasing length and radius of hollow fibers, clearance increases, and the percentage difference between current and previous models' predicted clearance reduces. A rise in residual renal clearance by a value of 0.5 mL/min doubles the standard KT/V, and a similar effect can be seen with an increasing number of weekly dialysis sessions.