Accuracy of the HVAD Pump Flow Estimation Algorithm

Algorithms are an important feature for assessment of ventricular assist device (VAD) performance and patient management. Flow estimation is one controller algorithm currently implemented in the HVAD Pump System. This parameter estimates the amount of blood flow passing through the pump and is calculated using three inputs: pump rotational speed, electrical current, and hematocrit. The methodology for flow estimation calculation is discussed subsequently. Both in vitro and in vivo studies were conducted to assess algorithm accuracy. During in vitro testing, three pumps were tested in four water-glycerol solutions at 37º C with viscosities equivalent to patient hematocrits of 20%, 30%, 40%, and 50%. These studies demonstrated a linear correlation between measured pump flow and estimated pump flow of > 0.94 for all test conditions. In-vivo studies (n = 9) were conducted in an acute ovine model where a reference flow probe was placed on the VAD outflow graft. In these experiments, estimated pump flow (mean, minimum, and maximum) demonstrated a linear correlation with measured pump flow. The best-fit linear equation for the data is y = 0.96x+0.54, which had a correlation of 0.92. Additionally, waveform fidelity was high (R2 > 0.96) in normal (i.e. non-suction) cases where flow pulsatility was > 2 L/min