A heart valve serves like a one-way sign on the road; it keeps blood flowing in the correct direction. Many valvular diseases prevent the valves from carrying out their function, which can be devastating. When patients have diseased heart valves, they typically receive a prosthetic heart valve replacement. Between 170,000 and 250,000 prosthetic valve replacements are performed in the world every year. Given how many patients receive a heart valve replacement, and how important these surgeries are for patient heart health, it is critical to understand how heart valve replacement can influence movement of the heart’s ventricles and blood flow. This study describes a novel mathematical model to simulate these interactions on a computer following a prosthetic valve replacement.

​

What did these researchers do?

Inspired by computer modeling work related to aircraft engines, these researchers developed a unique and efficient mathematical model to describe blood flow patterns after aortic and mitral valve replacements. Here, “aortic” and “mitral” refer to two of the four valves in the heart. The aortic valve regulates blood flow from the heart to the body, and the mitral valve controls blood flow between two of the heart’s chambers (the left atrium and left ventricle). In this model, it was assumed that the exact same type of heart valve replacement was used for both valves. This model captured realistic blood flow patterns during each heartbeat, as well as the motion that the left ventricle and aortic/mitral valves perform with each heartbeat. Ultimately, the model showed that blood flow patterns are significantly affected by valve replacements and found that the replacement resulted in an increase in the workload needed to push blood out of the ventricle. While this increased workload ensures that enough blood gets to the body, it puts additional stress on the heart and can therefore negatively affect patient outcomes. This suggests that replacing two different valves with the same prosthetic valve replacement is not ideal, and that valve-specific prostheses (e.g. aortic-specific valve replacements) should be used instead.

Why is this important?

Throughout each heartbeat, the ventricles, heart valves, and blood interact in complex ways. With each heartbeat, the ventricle contracts to squeeze blood out through the heart valve and into the human body. This “contraction” causes a phenomenon known as “deformation,” which describes a change in the shape of the ventricle, and during contraction and relaxation, there are also significant changes in the pressure that exists within the ventricles. To make matters even more complex, the large amount of ventricular deformation that occurs during each heartbeat makes the computational model extremely time-consuming and data-intensive. Compared to other clinical evaluations like ultrasound to evaluate heart function, this novel simulation provides a framework to describe all qualities of blood flow in three dimensions and during all times in a heartbeat. Importantly, this simulation can also be altered to fit different patient dimensions, test different prosthetic valve designs, and test the function of valve replacements implanted in several possible positions (for example, aortic versus mitral). Ultimately, this model has the potential to simulate the surgical procedure before the actual surgery takes place, allowing the surgeons to evaluate the optimal procedure, which may drastically improve the outcomes for patients after surgery.

​

How did the researchers do this?

To simplify the complex interactions between the heart’s ventricle, valves, and subsequent blood flow, each parameter was modeled separately. The ventricular motion during each heartbeat was simulated based on the deformation that occurs with each contraction. Next, the deformation of the valves during each heartbeat was modeled. Finally, calculations for how the motion of the valve leaflets would affect blood flow patterns were performed. Combining these individual components, the research team’s cumulative model was able to simulate an entire heartbeat in 38 to 42 hours using 1252 computer nodes, each with two Intel 12-core CPUs (you can imagine 2500 computers doing this stimulation in parallel). The model was then validated by comparing the characteristics of simulated blood flow over four consecutive heartbeats with the characteristics of real clinical data, finding that the simulation closely describes real blood flow patterns.

Computer-simulated flow in the left ventricle of the heart