WE-E-BRB-09: A GPU-Based Monte Carlo QA Tool for IMRT and VMAT.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Imaging

Background:

• Quality assurance (QA) in radiation therapy is crucial for patient safety and treatment efficacy.
• Traditional QA methods can be time-consuming and may not fully capture complex treatment delivery variations.
• The integration of patient-specific data and advanced computational techniques is essential for robust QA.

Purpose of the Study:

• To develop a novel, GPU-accelerated Monte Carlo (MC) 3D dosimetry tool for quality assurance (QA).
• To utilize patient geometry and actual treatment delivery information for enhanced QA accuracy.
• To create an efficient and clinically applicable QA solution for IMRT and VMAT.

Main Methods:

• Generating fluence maps from treatment plans and performing secondary dose calculations (SDC) using a GPU-based MC engine (gDPM).
• Extracting delivered fluence maps from machine log files and performing delivered dose calculations (DDC) for error detection.
• Comparing SDC, DDC, and planned dose (PD) to assess data transfer and machine delivery accuracy.
• Developing a web application for clinical integration and testing on six patients (4 VMAT, 2 IMRT).

Main Results:

• All patient comparisons (SDC, DDC, PD) demonstrated gamma passing rates exceeding 95% within the 20% isodose line.
• Dose distributions from SDC, DDC, and PD were found to be highly consistent.
• The entire QA process for typical IMRT or VMAT cases was completed in under one minute.

Conclusions:

• A GPU-based MC dosimetry QA tool has been successfully developed.
• The tool provides efficient and user-friendly QA for IMRT and VMAT treatments.
• This technology enhances the accuracy and speed of radiation therapy QA.