Patient-Specific Cardio-Respiratory Model for Optimization of Cardiac Radioablation

Area of Science:

• Medical physics
• Cardiac electrophysiology
• Radiotherapy

Background:

• Stereotactic Arrhythmia Radioablation (STAR) shows promise for refractory ventricular tachycardia.
• Cardiac and respiratory motion significantly challenge STAR's precision.
• Existing techniques aim to mitigate motion-induced treatment inaccuracies.

Purpose of the Study:

• To create a patient-specific dynamic model using cardiac and respiratory dynamic CT scans.
• To enable simulation of STAR treatments for evaluating motion management strategies.
• To assess the impact of cardiorespiratory dynamics on treatment precision.

Main Methods:

• Deep learning-based segmentation for cardiac structure geometry extraction.
• Deformable and rigid image registrations to assess cardiac and respiratory displacements.
• Integration of the dynamic model with dose maps for localized dose evaluation.
• Reproducibility assessment and physical phantom validation of treatment simulations.

Main Results:

• The developed dynamic model successfully simulated STAR treatments.
• Cardiorespiratory dynamics were found to have a potentially significant and patient-specific impact.
• The model's reproducibility was validated against expert references and phantom data.
• Illustrative cases on nine patients highlighted the importance of individual motion assessment.

Conclusions:

• Patient-specific dynamic modeling is crucial for evaluating motion management in STAR.
• Understanding cardiorespiratory dynamics is essential for optimizing STAR precision.
• This approach facilitates the development and validation of advanced motion management techniques.