Speed versus accuracy in a fast convolution photon dose calculation for conformal radiotherapy.

Area of Science:

• Medical Physics
• Radiotherapy
• Computational Dosimetry

Background:

• Accurate dose calculation is crucial for effective radiotherapy treatment planning.
• Existing methods may present compromises between computational speed and accuracy.
• Treatment planning optimization requires fast and flexible algorithms with minimal beam data.

Purpose of the Study:

• To describe and evaluate a convolution dose calculation algorithm for megavoltage photon beams.
• To assess the trade-off between speed and accuracy for treatment planning optimization.
• To validate the algorithm's performance against a commercial treatment planning system.

Main Methods:

• Developed a convolution algorithm using a tabular beam model and discrete scatter kernel.
• Fitted beam parameters to measured or calculated dose distributions.
• Applied the algorithm to pelvic and thoracic conformal plans.
• Compared results with the Pinnacle3 treatment planning system.

Main Results:

• The algorithm achieved calculation times of approximately 4 seconds for a 100 x 100 mm field.
• Dose-volume histogram agreement with Pinnacle3 was within 5% dose or 8% volume.
• Optimized grid resolution and kernel sampling improved speed without significant accuracy loss.

Conclusions:

• The developed convolution dose calculation algorithm is suitable for radiotherapy treatment planning optimization.
• It provides a balance of speed, flexibility, and accuracy.
• The method requires minimal beam data, enhancing its practical applicability.