Ultrafast treatment plan optimization for volumetric modulated arc therapy (VMAT)

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Optimization

Background:

• Volumetric Modulated Arc Therapy (VMAT) requires efficient treatment plan optimization.
• Existing methods may face challenges in balancing plan quality and computational time.

Discussion:

• The study introduces a novel aperture-based algorithm for VMAT optimization using a column generation approach.
• The algorithm formulates the problem as a large-scale convex programming problem, generating apertures sequentially.
• It incorporates a cost function that ensures desired dose distribution and smooth dose rate variations, considering MLC constraints.

Key Insights:

• The developed algorithm achieves high-quality VMAT plan optimization with remarkable efficiency.
• Tested on clinical cases (prostate, head-and-neck), the algorithm generated deliverable plans rapidly.
• Optimization times were significantly reduced, especially with GPU acceleration (18-31 seconds).

Outlook:

• This efficient aperture-based approach holds promise for improving VMAT treatment planning workflows.
• Further validation and implementation in clinical settings could enhance radiotherapy delivery.
• Potential for broader application in complex radiation therapy optimization problems.