A comparison of four indices for combining distance and dose differences

Area of Science:

• Medical Physics
• Radiation Oncology
• Dosimetry

Background:

• Comparing radiation dose distributions is crucial for treatment plan quality assurance.
• Existing methods combine dose difference and distance to agreement into single measures.
• These measures are often presented as pass/fail or numeric indices.

Purpose of the Study:

• To compare various indices for evaluating one-dimensional (1D) and three-dimensional (3D) dose distributions.
• To determine the most appropriate index for different clinical and measurement scenarios.
• To analyze the derivation of numeric indices from pass/fail metrics.

Main Methods:

• Comparison of gamma, kappa, International Commission on Radiation Units & Measurements (ICRU) Report 83, and box indices.
• Evaluation using 1D distributions with varying dose gradients and a 3D clinical treatment plan.
• Analysis of the impact of dose and distance offsets on index performance.

Main Results:

• The ICRU Report 83 index shows discontinuities unless specific gradient cutoffs are applied, then matching kappa index results.
• All indices yield similar outcomes at very high or low dose gradients.
• For moderate gradients, the box index is least strict, followed by gamma, then kappa.

Conclusions:

• The kappa index is recommended when clinical tolerances exceed measurement uncertainties, using clinical tolerances for setting parameters.
• The box index is preferred when measurement system uncertainties are dominant, aiding in identifying delivery errors.
• The choice of index depends on the relative magnitudes of clinical tolerances and measurement uncertainties.