Identification and Control for Automated Regulation of Hemodynamic Variables During Hemodialysis

This paper proposes a novel model-based control methodology for a computer-controlled hemodialysis system, designed to maintain the hemodynamic stability of end-stage renal failure patients undergoing fluid removal during hemodialysis. The first objective of this paper is to introduce a linear parameter varying system to model the hemodynamic response of patients during hemodialysis. Ultrafiltration rate (UFR) and dialysate sodium concentration (DSC) are imposed as the inputs, and the model computes the relative blood volume (RBV), percentage change in heart rate (Delta HR), and systolic blood pressure (SBP) during the course of hemodialysis. The model parameters were estimated based on data collected from 12 patients undergoing 4 profiled hemodialysis sessions. The modeling results demonstrated that the proposed model could be useful for estimating the individual patient's hemodynamic behavior during hemodialysis. Based on the model, the second objective is to implement a computer-controlled hemodialysis system for the regulation of RBV and HR during hemodialysis while maintaining SBP within stable range. The proposed controller is based on a model predictive control approach utilizing pre-defined constraints on the control inputs (UFR and DSC) as well as the output (SBP). The designed control system was experimentally verified on four patients. The results demonstrated that the proposed computer-controlled hemodialysis system regulated the RBV and HR of the patients according to individual reference profiles with an average mean square error of 0.24% and 2.6%, respectively, and thus can be potentially useful for ensuring the stability of patients undergoing hemodialysis by avoiding sudden changes in hemodynamic variables.