Statistical variability and confidence intervals for planar dose QA pass rates.

Area of Science:

• Medical Physics
• Radiation Oncology
• Quality Assurance

Background:

• Pretreatment quality assurance (QA) for intensity-modulated photon fields commonly uses pass rates based on percent difference (%Diff) and distance-to-agreement (DTA).
• Low areal density of detectors in dosimeters leads to statistical variability in pass rates, dependent on detector sampling geometry.
• Understanding this variability is crucial for accurate dose comparison.

Purpose of the Study:

• Analyze the statistics of common pass rate calculation methods for radiation therapy QA.
• Propose methods for establishing confidence intervals for pass rates obtained with low-density detector arrays.
• Improve the reliability of dose comparison metrics in intensity-modulated radiation therapy.

Main Methods:

• Acquired dose planes for prostate and head/neck intensity-modulated fields using diode arrays and EPID.
• Simulated low-density measurements by downsampling high-density EPID data.
• Varied detector grid positions and densities (1-2 detectors/cm²) to assess pass rate variability.
• Calculated pass rates using %Diff/DTA composite analysis and gamma evaluation with local and global normalization.

Main Results:

• Gamma analysis yielded 2%-5% higher pass rates than %Diff/DTA composite analysis.
• Local normalization resulted in 2%-12% lower pass rates compared to global maximum normalization.
• Pass rate distributions for low-density arrays can be predicted using binomial distribution to establish confidence intervals.

Conclusions:

• Choice of calculation metric and parameters significantly impacts dose plane QA pass rates.
• Pass rates from low-density arrays have statistical uncertainty that can be modeled with confidence intervals.
• Report pass rates with a complete description of the calculation method and associated confidence intervals.