Virtual Shape-Editing of Patient-Specific Vascular Models Using Regularized Kelvinlets.

Area of Science:

• Biomedical Engineering
• Computational Science
• Medical Imaging

Background:

• Cardiovascular diseases and interventions alter vascular shape and hemodynamics.
• Computational modeling of patient-specific vasculature aids hemodynamic analysis.
• Existing shape modification methods are manual and anatomically limited.

Purpose of the Study:

• To develop robust, physics-based techniques for modifying vascular models.
• To enable automatic generation of smooth deformations across diverse vascular anatomies.
• To facilitate the creation of patient-specific vascular geometries for research.

Main Methods:

• Adapted Regularized Kelvinlets, analytical solutions to linear elastostatics, for elastic shape-editing.
• Developed three methods to artificially create aneurysms, stenoses, and tortuosity.
• Applied methods to multiple patient-specific vascular models.

Main Results:

• Demonstrated successful generation of geometric changes (aneurysms, stenoses, tortuosity) across various vascular anatomies.
• Showcased the ability to create variations in shape and size of induced geometric changes.
• Validated the robustness of Kelvinlet-based deformers on diverse models.

Conclusions:

• Kelvinlet-based deformers enable robust shape editing of vascular models irrespective of anatomical location.
• Parametric control over the size of geometric changes is achievable.
• Facilitates virtual surgery simulations, hemodynamic impact studies, and synthetic data generation.