Methods, software and datasets to verify DVH calculations against analytical values: Twenty years late(r)

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Imaging

Background:

• Accurate dose-volume histogram (DVH) calculation is critical for radiation therapy treatment planning.
• Existing DVH algorithms may have limitations in accuracy and consistency.
• A standardized, independent method is needed to evaluate DVH calculation accuracy.

Purpose of the Study:

• To develop and validate a software-independent standard for evaluating dose-volume histogram (DVH) calculation accuracy.
• To identify limitations and failure modes in DVH calculation algorithms.
• To provide a benchmark for assessing the precision of commercial DVH software.

Main Methods:

• Designed geometrically simple objects (sphere, cylinder, cone) with varying orientations and spatial resolutions.
• Generated synthetic dose grids with linear 1D dose gradients.
• Calculated ground truth DVH curves analytically.
• Tested two commercial DVH algorithms (Pinnacle, PlanIQ) against analytical standards using custom analysis software.

Main Results:

• PlanIQ showed significantly fewer deviations from analytical values compared to Pinnacle across various tests.
• Pinnacle exhibited higher error rates and variability, attributed to insufficient 3D grid supersampling and inconsistent 'end-capping' implementation.
• PlanIQ's deviations were mainly associated with modeling complex shapes (rotated cylinder) with coarse axial spacing.

Conclusions:

• The developed method provides a reliable, software-independent standard for DVH accuracy assessment.
• The tests effectively probe DVH algorithm design, measure accuracy, and identify failure modes.
• The findings highlight differences in accuracy between commercial DVH software, emphasizing the need for careful algorithm selection and validation.