Control of implantable axial blood pumps based on physiological demand

Axial blood pumps have entered the arena of prolonged clinical support very successfully. The pump performance and sensitivity to delta pressure make them particularly suited for VAD support. However, in contrast to pulsatile pumps they do not offer an inherent Starling-like mechanism to adapt pumping performance to venous return and physiological demand. Within a joint project of the University of Vienna and MicroMed, an algorithm for physiological pump control was developed. The physician defines "Desired flow levels" at rest and exercise (see figure 1), depending on heart rate. The pump flow signal is analyzed continuously to determine if such pump flow can be achieved in consideration of potential suction and other constraints. Based on various experiences with LVAD patients 7 different algorithms were developed for automatic detection of ventricular suction. These algorithms analyze the flow pattern in time domain for the presence of distinct suction indicators. For selection and numerical optimization of the algorithms a data base of classified data records from more than 100 patients were collected