We are developing computational tools to perform virtual surgeries under physiological conditions with patient-specific anatomies. Virtual surgery can revolutionize the biomedical device (BMD) design and implantation by: enabling the optimization of BMD on the specific patient's anatomy and flow conditions, which has already been shown to significantly affect the hemodynamics; and facilitating surgical planning to select the best location/orientation for BMD implantation. As shown by past research the difference in the implantation, for example, of a bileaflet mechanical heart valve (BMHV) can affect the performance and hemodynamics of the valve. We have developed a powerful CFD tool that can simulate the blood flow through biomedical devices with moving boundaries and the fluid-structure interaction (FSI) under physiologic conditions. This tool has been tested in different applications with complex anatomic configurations, such as aneurysm hemodynamics, Fontan surgeries, and BMHV flows. The FSI simulations of a BMHV flow was validated against in vitro experiments and could capture all the flow features with great accuracy. We have recently carried out FSI simulations of a BMHV implanted in a anatomically realistic aorta in which the left ventricle (LV) was replaced by a pulsatile inflow waveform. In this work, we propose to extend our method to simulate the flow through a BMHV driven by the actual LV motion. The anatomy of the heart is obtained from MRI scan of unhealthy volunteer (St. Jude Hospital) and the left ventricle and the aorta geometry are extracted. The first step is to simulate the flow through a stationary LV with an implanted BMHV to test the capabilities of FSI-CFD tool in real anatomical setting. The second step is to impose theoretical kinematics for LV motion and test the interaction of the BMHV with the flow. The last step is to extract the real patient-specific LV wall motions from MRI data and specify it into the simulation. The shear stress and other parameters will be calculated in the flow field and the cardiovascular walls to identify the locations with high chance of blood cell damage. This work is supported by NIH Grant RO1-HL-07262 and the Minnesota Supercomputing Institute.