Development of a suction detection system for a motorized pulsatile blood pump

A computational model has been developed to study the effects of left ventricular assist devices (LVADs) on the cardiovascular system during a ventricular collapse. The model consists of a toroidal pulsatile blood pump and a closed loop circulatory system. Together, they predict the pump's motor current traces that reflect ventricular suck-down and provide insights into torque magnitudes that the pump experiences. In addition, the model investigates likeliness of a suction event and predicts reasonable outcomes for a few test cases. Ventricular collapse was modeled with the help of a mock circulatory loop consisting of a artificial left ventricle and centrifugal continuous flow pump. This study also investigates different suction detection schemes and proposes the most suitable suction detection algorithm for the TORVAD pump, toroidal left ventricular assist device. Model predictions were further compared against the data sampled during in vivo animal trials with the TORVAD system. The two sets of results are in good accordance.