Can a linear electrical analog model of a mechanical valve predict flow by using a pressure gradient?

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Medical Device Simulation

Background:

• Predicting prosthetic heart valve hemodynamics is crucial for device design and clinical assessment.
• Previous models successfully simulated biological aortic valve function using electrical analog parameters.

Purpose of the Study:

• To evaluate if a developed electrical analog model could accurately predict flow dynamics in mechanical aortic valves.
• To assess the efficacy of a linear model with a nonlinear filter for hemodynamic prediction.

Main Methods:

• An electrical analog model with compliance, resistance, and inertance parameters was employed.
• Simulated pressures and flow were generated using a pulse duplicator system across varied heart rates and stroke volumes.
• Model parameters were fitted using least-squares, and predictions were refined with a nonlinear filter.

Main Results:

• The linear model and nonlinear filter demonstrated variable accuracy, with mean flow errors ranging from 3% to 90%.
• Prediction accuracy was better at normal and high flow rates but significantly poorer at low flow rates.
• The nonlinear filter did not fully resolve discrepancies across all flow ranges, particularly at higher harmonics.

Conclusions:

• The applied linear electrical analog model and nonlinear filter have limited predictive capability for mechanical aortic valve flow.
• Further research into alternative nonlinear modeling approaches is recommended for improved hemodynamic simulation.