A convolution-superposition fluence model for the Siemens HD120 multi leaf collimator with application to a 3D VMAT...

Area of Science:

• Medical Physics
• Radiotherapy Physics
• Computational Dosimetry

Background:

• Accurate dose calculation is crucial for radiotherapy treatment planning.
• Monte Carlo simulations, while accurate, are computationally intensive.
• Developing faster, in-house dose calculation software is desirable for research and clinical needs.

Purpose of the Study:

• To create a rapid fluence model for the Siemens HD120 multi leaf collimator (MLC) using convolution-superposition.
• To develop a 3D Volumetric Modulated Arc Therapy (VMAT) dose engine utilizing this MLC fluence model.
• To provide an efficient alternative to time-consuming Monte Carlo methods for in-house software development.

Main Methods:

• Commissioned MLC transmission, interleaf leakage, and tongue-and-groove maps using EPID-acquired images.
• Integrated these maps and a head-scatter model into a convolution-superposition algorithm for fluence modeling.
• Developed a 3D VMAT dose engine incorporating pre-computed dose kernels and a curvature-correction map.

Main Results:

• Achieved excellent agreement between simulated and measured EPID responses during commissioning.
• Demonstrated high pass rates (98.9%) for 2D gamma analysis on sweeping-window fields.
• Showed excellent agreement (98.7% pass rate) between the developed dose engine and Eclipse (Acuros) for VMAT patient plans.
• Reported fast calculation times: 13.6 seconds/MLC field and 1.4 minutes/arc for 3D VMAT dose.

Conclusions:

• A fast, convolution-superposition-based fluence model for the Siemens HD120 MLC has been successfully commissioned.
• The developed 3D VMAT dose engine demonstrates accuracy comparable to commercial software.
• This work facilitates the development of efficient in-house dose calculation tools for radiotherapy research and clinical applications.