Biomechanically Informed Image Registration for Patient-Specific Aortic Valve Strain Analysis

Area of Science:

• Cardiovascular Biomechanics
• Medical Imaging Analysis
• Computational Fluid Dynamics

Background:

• Aortic valve (AV) biomechanics are crucial for cardiac function.
• Pathological AVs, like bicuspid valves, exhibit altered biomechanics, accelerating disease.
• Current imaging and computational methods struggle to accurately capture complex AV motion and deformation.

Purpose of the Study:

• To develop an advanced framework for precise patient-specific AV biomechanical assessment.
• To overcome limitations in tracking rapid valve motion and complex deformations.
• To improve characterization of leaflet loading and strain in various AV conditions.

Main Methods:

• Developed a framework combining image registration with the finite element method (FEM).
• Utilized patient-specific valve geometries from 4D echocardiography and CT scans.
• Simulated AV closure to generate intermediate deformation states for enhanced leaflet tracking.

Main Results:

• FEM-augmented registration improved AV tracking accuracy by 40% compared to direct registration.
• Enabled bounded-uncertainty strain estimation by aligning simulated and imaging data.
• Quantified leaflet strains (areal, Green-Lagrange, deviatoric) in trileaflet, bicuspid, and pediatric valves.

Conclusions:

• The FEM-augmented registration framework significantly enhances geometric tracking of the aortic valve.
• Provides bounded-uncertainty leaflet strain estimates crucial for individualized intervention planning.
• Offers potential for generating training data for advanced machine learning methods in cardiac imaging.