Advancing aortic stenosis assessment: Validation of fluid-structure interaction models against 4D flow MRI data

Area of Science:

• Cardiovascular Mechanics
• Biomedical Engineering
• Computational Biology

Background:

• In vitro validation of computational models for the aortic valve (AV) is common, but systematic in vivo validations are scarce.
• Accurate computational models are crucial for understanding aortic stenosis (AS) and guiding treatment decisions.

Purpose of the Study:

• To systematically validate patient-specific computational fluid-structure interaction (FSI) models of the AV using in vivo data.
• To demonstrate the utility of validated in silico models for enhancing the analysis of in vivo data and assessing critical physiological quantities.

Main Methods:

• Developed patient-specific FSI models of the AV for five patients using computed tomography and 4D flow magnetic resonance imaging data.
• Employed a patient-specific calibration strategy to determine the AV Young's modulus.
• Validated the computational models against in vivo 4D flow measurements.

Main Results:

• The validated computational models accurately replicated in vivo flow dynamics.
• Demonstrated the models' capability to serve as cross-checks for in vivo data, reducing noise and identifying erratic behavior.
• Computed blood residence time, a critical quantity for assessing thrombus formation probability.

Conclusions:

• Validated patient-specific AV computational models offer a robust framework for improving AS diagnosis and treatment.
• In silico models integrated with in vivo data enhance precision in assessing AS and personalizing treatment strategies.
• This approach facilitates the design of next-generation AV bioprostheses and advances the clinical integration of computational technologies.