Shape optimization of a trumpet-tipped LVAD inflow cannula to reduce blood damage using a genetic algorithm

Use of computational fluid dynamics (CFD) in the field of blood-contacting medical device design and analysis has been growing in recent years. For example, the U.S. Food and Drug Administration (FDA) Center of Devices and Radiological Health (CDRH) has accelerated interest in industry and academia with nozzle and blood pump benchmarks to uncover best practices and to hopefully elevate the status of CFD to be applied as a safety analysis tool for medical devices. One area, not discussed as often as the pure simulation is the design optimization of hemodynamic devices. A systematic shape “optimization” should be distinguished from a simple “design improvement” by performing many flow field computations and design iterations to improve performance. In this paper, the shape optimization of a trumpet-tipped inflow cannula is presented using a single-objective genetic algorithm (GA) to minimize the blood damage. Many varying accounts in the literature have pointed to the advantages of the trumpet-tipped left ventricular assist device (LVAD) cannula for low blood damage and uniform velocity distribution with little to no backflow when compared to other shapes such as blunt, beveled and caged cannulas.