Computational Fluid Dynamics in Cardiovascular Devices

Abstract

Computational fluid dynamics has become integral to the design and regulatory evaluation of cardiovascular devices, from prosthetic heart valves to ventricular assist devices, yet the fidelity and clinical relevance of CFD simulations vary widely across device categories and modelling approaches. We assessed 226 CFD-for-cardiovascular-device programmes conducted across biomedical engineering centres in Italy, Germany, and Estonia between 2014 and 2024, spanning five device categories: bioprosthetic heart valves, mechanical heart valves, ventricular assist devices, coronary stents, and aortic endografts. A CFD Cardiovascular Device Maturity Index (CCDMI) was constructed from five sub-scores -- haemodynamic prediction accuracy, thrombogenicity modelling fidelity, patient-specific geometry incorporation, experimental validation completeness, and regulatory submission integration -- with weights from regression against progression to regulatory acceptance or design-freeze decisions. CCDMI correlated with regulatory progression at r = +0.84 and discriminated advancing from stalled programmes with an AUC of 0.884. Bioprosthetic heart valve programmes scored highest (mean CCDMI 0.826), while aortic endografts trailed at 0.596. Only 36.3 percent of programmes exceeded the 0.75 CCDMI threshold. Haemodynamic prediction accuracy carried the largest regression weight (beta = +0.280), followed by experimental validation completeness (beta = +0.228)