A technique for generating phase-space-based Monte Carlo beamlets in radiotherapy applications.

Area of Science:

• Medical Physics
• Computational Biology
• Radiotherapy Physics

Background:

• Monte Carlo (MC) methods are increasingly used for accurate dose calculations in radiotherapy.
• Current MC beamlet generation relies on particle source models (PSM), which may limit accuracy.
• MC beamlets are crucial optimization entities in advanced treatment planning.

Purpose of the Study:

• To implement and evaluate a phase-space-based approach for MC beamlet generation.
• To improve the accuracy of beamlet dose distributions compared to PSM methods.
• To provide a method for generating absolute dose distributions for IMRT optimization.

Main Methods:

• Utilized a standard BEAMnrc phase space as the input.
• Developed an efficient two-pass sorting algorithm to divide phase space into beamlets.
• Labeled particles using inheritable particle history for accurate tracking.
• Transported individual beamlets into patient CT data to generate doselets.
• Described methods for doselet normalization and conversion to Gy for IMRT.

Main Results:

• The phase-space-based method demonstrated potential for greater accuracy in beamlet dose distributions.
• An efficient sorting algorithm processed five million particles in under 8 seconds on a single CPU core.
• The approach allows for separate transport of beamlets, generating distinct dose distributions (doselets).

Conclusions:

• The phase-space-based approach offers a more accurate method for MC beamlet generation.
• This technique facilitates the creation of precise dose distributions for radiotherapy treatment planning.
• The described methods support the integration of MC beamlets into IMRT optimization workflows.