Framework for Patient-Specific Hemodynamic Modeling of Left Ventricle Assist Device (LVAD) Patients

We present a high-fidelity, patient-specific computational framework for simulating cardiovascular hemodynamics in left ventricular assist device (LVAD) patients, with the ability to evaluate the impact of mitral and aortic valve repair. The method couples three-dimensional computational fluid dynamics (CFD) with a zero-dimensional lumped-parameter network representing the full cardiovascular system. Valve dynamics are governed by pressure–flow physics with patient-specific regurgitant areas, enabling the investigation of valve insufficiency and surgical repair. The model incorporates dynamic contrast-enhanced CT imaging, echocardiography, and catheter-based measurements. We demonstrate the utility of this framework using a clinical case involving a patient on long-term LVAD support with mitral regurgitation and aortic insufficiency. The coupled model resolves detailed pressure and velocity fields in the left heart, LVAD cannulae, and aortic root, while maintaining system-wide physiological consistency. Importantly, the framework also quantifies right ventricular loading and function, allowing assessment of how left-sided valve interventions propagate through the cardiopulmonary circulation. Using this pipeline, we simulate mitral and aortic valve repair by modulating valve closure mechanics and regurgitant area, revealing changes in both local and global hemodynamics. This modeling platform provides a powerful means to study the interplay between mechanical circulatory support, native valve physiology, and right heart function. It enables rigorous, patient-specific evaluation of surgical interventions in LVAD patients and delivers quantitative insight into clinically important metrics—such as aortic pulsatility, RV afterload, and chamber-level flow patterns—that are challenging to capture with current imaging modalities.