Scenario-free robust optimization algorithm for IMRT and IMPT treatment planning.

Area of Science:

• Medical Physics
• Computational Biology
• Radiotherapy

Background:

• Robust treatment planning algorithms for intensity modulated proton therapy (IMPT) and intensity modulated radiation therapy (IMRT) aim to reduce dose uncertainties by incorporating error scenarios.
• Traditional methods face computational challenges due to the curse of dimensionality, making them potentially prohibitive for complex treatment plans.

Purpose of the Study:

• To propose a scenario-free probabilistic robust optimization algorithm that addresses the runtime and memory limitations of conventional robustness techniques.
• To develop an efficient computational method for radiation therapy treatment planning.

Main Methods:

• The scenario-free approach optimizes cost functions based on expected dose distributions and total variance, utilizing precomputed influence matrices.
• This method avoids storing individual error scenarios, reducing computational and memory burdens.
• The algorithm was implemented in matRad and benchmarked against traditional robust and margin-based methods for photon and proton plans.

Main Results:

• The scenario-free algorithm achieved comparable plan quality to traditional robust methods while reducing dose standard deviation in specific structures.
• It demonstrated substantial computational time savings (5-600x faster) compared to traditional robust approaches.
• Runtime and memory usage were independent of the number of error scenarios, similar to non-robust methods.

Conclusions:

• A novel scenario-free optimization approach using precomputed probabilistic quantities was successfully developed and validated.
• This method offers significant computational advantages, making it suitable for complex 3D and 4D robust optimization with numerous error scenarios or CT phases.
• The approach preserves compatibility with advanced uncertainty modeling while meeting dose and robustness requirements.