A model-based 3D patient-specific pre-treatment QA method for VMAT using the EPID.

Area of Science:

• Medical Physics
• Radiation Oncology
• Radiotherapy Quality Assurance

Background:

• Pre-treatment quality assurance (QA) is crucial for radiotherapy accuracy.
• Epidermal electronic portal imaging devices (EPID) offer potential for in-vivo dosimetry and QA.
• Model-based dose reconstruction from EPID images can provide 3D dose estimates.

Purpose of the Study:

• To develop and validate a model-based, 3D patient dose reconstruction method using EPID images for pre-treatment QA.
• To assess the method's sensitivity to potential Multileaf Collimator (MLC) delivery errors.
• To compare EPID-reconstructed doses with Treatment Planning System (TPS) predictions and ion chamber measurements.

Main Methods:

• A back-projection dose reconstruction model processed cine-mode EPID images to estimate 3D dose on CT data.
• Validation involved 24 Stereotactic Body Radiation Therapy-Volumetric Modulated Arc Therapy (SBRT-VMAT) plans.
• Comparisons included ion chamber measurements, TPS predictions (AAA, AcurosXB), and dose distributions analyzed using gamma (γ) tests (3%/3mm, 2%/2mm).

Main Results:

• EPID dose reconstruction showed good agreement with ion chamber measurements (average difference < 1.34%).
• All plans passed EPID vs. AcurosXB and EPID vs. forward model γ-comparisons.
• The method successfully detected MLC errors of ~1.0mm magnitude using a 3%/3mm γ-test, demonstrating sensitivity to delivery inaccuracies.

Conclusions:

• The developed model-based 3D EPID dose reconstruction is a viable tool for pre-treatment QA in SBRT-VMAT.
• The method effectively identifies MLC delivery errors, contributing to improved patient safety.
• Further investigation with stricter criteria (2%/2mm γ-test) revealed enhanced detection capabilities for MLC errors.