ADEPT: A Noninvasive Method for Determining Elastic Parameters of Valve Tissue

Area of Science:

• Biomedical Engineering
• Computational Biology
• Medical Imaging

Background:

• Accurate prediction of heart valve repair requires patient-specific mechanical parameters.
• Current noninvasive methods are insufficient for determining these in vivo parameters.
• Computer simulations of virtual interventions can optimize surgical outcomes.

Purpose of the Study:

• To introduce ADEPT (A noninvasive method for Determining Elastic Parameters of valve Tissue).
• To demonstrate ADEPT's application in estimating tricuspid valve mechanical properties.
• To enhance the accuracy of computational models for heart valve repair.

Main Methods:

• 3D echocardiogram image registration to track valve displacements.
• Physics-informed neural networks to estimate nonlinear mechanical properties.
• Development of a patient-specific computational model using estimated parameters.

Main Results:

• ADEPT successfully estimated patient-specific elastic parameters of the tricuspid valve.
• The simulated model with ADEPT parameters achieved a mean symmetric distance < 1 mm.
• The patient-specific model demonstrated double the accuracy compared to models with generic parameters.

Conclusions:

• ADEPT provides a novel noninvasive approach for determining in vivo valve tissue mechanical properties.
• Accurate patient-specific parameters significantly improve the predictive power of computational models for valve repair.
• This method holds potential for optimizing virtual interventions and clinical decision-making in valvular heart disease.