Yet anOther Dose Algorithm (YODA) for independent computations of dose and dose changes due to anatomical changes

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Science

Background:

• Accurate dose calculation is crucial for proton therapy.
• Anatomical changes during treatment necessitate adaptive planning.
• Current methods may lack system independence or efficiency in assessing anatomical variations.

Purpose of the Study:

• To evaluate a novel physics-based, deterministic algorithm for treatment planning system-independent dose calculations.
• To assess the algorithm's capability in computing dosimetric differences due to anatomical changes.
• To validate the algorithm's accuracy against established methods like Monte Carlo simulations.

Main Methods:

• A semi-numerical approach solving partial differential equations for proton phase-space density.
• Gaussian beam splitting for lateral heterogeneity correction.
• Application of adjoint theory to approximate dose changes from anatomical variations.
• Benchmarking against Monte Carlo simulations (TOPAS) using 3D gamma index analysis.

Main Results:

• The algorithm achieved a 94.55% passing rate (1 mm, 1%, 10% cut-off) for a lung test case compared to Monte Carlo simulations.
• Dosimetric difference calculations showed accuracy within 5.7% for non-robustly optimized plans and 4.8% for robustly optimized plans on repeat CTs.
• The algorithm demonstrated accuracy in both single and multi-spot proton irradiations.

Conclusions:

• The Yet anOther Dose Algorithm (YODA) provides accurate dose computations comparable to TOPAS.
• YODA can effectively compute dosimetric changes resulting from anatomical variations with acceptable errors.
• This algorithm facilitates patient-specific quality assurance for online adaptive proton therapy.