A method to reconstruct and apply 3D primary fluence for treatment delivery verification

Area of Science:

• Medical Physics
• Radiation Oncology
• Radiotherapy Quality Assurance

Background:

• Current radiotherapy quality assurance (QA) methods for Intensity-Modulated Radiation Therapy (IMRT) and Volumetric Modulated Arc Therapy (VMAT) may not detect all treatment delivery errors.
• There is a need for efficient and comprehensive verification methods that can identify subtle discrepancies in radiation delivery.

Purpose of the Study:

• To report a novel method for IMRT and VMAT treatment delivery verification using 3D volumetric primary beam fluences.
• To demonstrate the efficiency of 3D beam fluence reconstruction.
• To show that 3D fluence calculation (3DFC) QA detects more errors than existing methods, especially for VMAT.
• To validate that 3DFC QA results correlate with physical phantom measurements.

Main Methods:

• Reconstructed 3D volumetric primary fluences by forward-projecting beam parameters from DICOM plans and delivery log files using first-order ray tracing.
• Compared reconstructed 3D fluences from plan data versus delivery log data to verify treatment delivery.
• Assessed passing rates using voxel intensity differences and 3D gamma analysis.
• Investigated QA sensitivity and correlations with physical phantom measurements using 20 clinical plans with introduced errors.

Main Results:

• The 3DFC QA method detected common delivery errors (gantry angle, MU, jaw, collimator, MLC position) at less than standard tolerances.
• 3DFC QA demonstrated higher sensitivity in detecting errors compared to measurement-based QA methods.
• Significant correlations were found between 3DFC QA passing rates and physical phantom measurement-based QA passing rates.

Conclusions:

• The 3DFC method offers a less demanding, more comprehensive measurement-less verification alternative to dose recalculations or log file parameter checks.
• 3DFC QA complements physical phantom measurements and can be used for verification when measurements are unavailable.
• The method has been successfully implemented for pretreatment QA and daily automated log file verification, effectively identifying data transfer and delivery errors.