Multiphysics computational models for cardiac flow and virtual cardiography.

A multiphysics simulation approach is developed for predicting cardiac flows as well as for conducting virtual echocardiography (ECHO) and phonocardiography (PC) of those flows. Intraventricular blood flow in pathological heart conditions is simulated by solving the three-dimensional incompressible Navier-Stokes equations with an immersed boundary method, and using this computational hemodynamic data, echocardiographic and phonocardiographic signals are synthesized by separate simulations that model the physics of ultrasound wave scattering and flow-induced sound, respectively. For virtual ECHO, a Doppler ultrasound image is reproduced through Lagrangian particle tracking of blood cell particles and application of sound wave scattering theory. For virtual PC, the generation and propagation of blood flow-induced sounds ('hemoacoustics') is directly simulated by a computational acoustics model. The virtual ECHO is applied to reproduce a color M-mode Doppler image for the left ventricle as well as continuous Doppler image for the outflow tract of the left ventricle, which can be verified directly against clinically acquired data. The potential of the virtual PC approach for providing new insights between disease and heart sounds is demonstrated by applying it to modeling systolic murmurs caused by hypertrophic cardiomyopathy.