Flow Rate Estimation of a Centrifugal Blood Pump Using the Balance of Control Currents Through the Electromagnets in the Magnetic Bearing

Shuya Shida, Toru Masuzawa, Masahiro Osa, Yutaka Suzuki
Vol. 12 (2023) p. 244-252

Implantable ventricular assist devices (iVADs) are commonly used to treat patients with severe heart failure. To ensure safe and reliable operation of iVADs, it is crucial to develop a method for estimating the flow rate (Q) of the centrifugal blood pump used in iVAD without relying on a flowmeter. To address this need, we previously developed a flow rate estimation method (FEM-r) that focuses on the correlation between Q and the levitation position (r) of an impeller suspended by a magnetic bearing in a centrifugal blood pump. The FEM-r maintains a high level of estimation accuracy even when changes in blood viscosity are expected during treatment. This advantageous characteristic distinguishes FEM-r from the conventional method (FEM-IMT) that employs motor current as an estimation index. However, a drawback of FEM-r is the requirement for an additional displacement sensor to measure r. In this study, we propose a solution to this issue by leveraging the correlation between r and the balance of currents (IEMs) flowing through the electromagnets in a magnetic bearing. Thus, we propose an estimation method called FEM-IEM, which utilizes the balance of IEMs as a flow rate estimation index. The balance of IEMs can be obtained using the minimum number of sensors required for iVAD control. First, we developed an estimation equation for FEM-IEM and determined its estimation coefficients by multiple regression analysis based on the data obtained from experiments using a mock loop. The root mean square error (RMSE) and determination coefficient (R2) of the multiple regression analysis were 0.293 L/min and 0.965, respectively. The RMSE between the estimated and measured Q values were 0.284, 0.347, and 1.069 L/min for FEM-IEM, FEM-r, and FEM-IMT, respectively, indicating that the accuracy of FEM-IEM was comparable to that of FEM-r and significantly higher than that of FEM-IMT. Thus, the simple FEM-IEM proposed in this study demonstrates excellent performance for clinical use, allowing accurate flow rate estimation of a centrifugal blood pump without requiring additional position sensor, viscosity measurement, or compensation method, even when the blood viscosity fluctuates.