Extreme membrane position detection and pump output estimation in pulsatile VADs

Abstract

One of the disadvantages of pulsatile ventricular assist devices (VADs) is the lack of feedback information about the flow and the range of diaphragm movement. This influences the proper selection of control parameters and control automation. Objective: The aim of the present study was to develop algorithms for complete emptying (CE) and complete filling (CF) detection as well as stroke volume estimation. Methods: The developed methods do not require additional measurements, as they are based on the data from the VAD control unit and the modeled value of the flow in the VAD blood chamber Siewnicka and Janiszowski (2018) [1]. The experiments were performed on a measurement stand of a hybrid simulator of the cardiovascular system with a connected POLVAD-MEV device. The algorithms were developed and then tested for a wide range of device operation conditions. Results: The CF detection algorithm had one missed detection out of 39 measurements. The CE algorithm was free of errors for all investigated cases. The pump output algorithm was approximately 90% accurate in relation to the direct measurement. Conclusion and significance: The application of these methods to pulsatile VADs may be very convenient for the medical operators of these devices and could result in a considerable improvement in the field of treatment safety. An abstract of this work was reported in Siewnicka et al., 2017 [2].