Effective optimization strategy for large optimization volume object, remaining volume at risk (RVR):α-value...

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Intensity-modulated radiation therapy (IMRT) optimization for large or complex volumes like the remaining volume at risk (RVR) is challenging due to patient-specific shapes.
• Generalized Equivalent Uniform Dose (gEUD) offers a method for dose-volume constraint control, but the optimal use of its parameter 'a'-values is not well-defined.

Purpose of the Study:

• To elucidate the characteristics of 'a'-values in gEUD optimization.
• To establish effective strategies for utilizing 'a'-values in IMRT planning, particularly for the RVR.
• To enable less empirical and more precise dose-volume control in IMRT.

Main Methods:

• The study analyzed the gEUD curve as a function of the 'a'-value, a weighted generalized mean, using differential geometry to understand its deformation during optimization.
• The 'a'-value roles were classified based on the gradient and curvature of the gEUD curve.
• A strategy employing multiple 'a'-values was developed and tested for RVR optimization in head and neck cancer patients undergoing IMRT.

Main Results:

• Lower 'a'-values (1-3) were effective for the entire dose-volume range, while higher 'a'-values (12-20) specifically targeted the high-dose range.
• Intermediate 'a'-values provided effective dose reduction across a broad range.
• Multiple gEUD-based optimization demonstrated significant superiority over standard exponential constraints, improving D2% by up to 25.9% and V10Gy by 8.1%.

Conclusions:

• Appropriate selection and application of 'a'-values are crucial for effective gEUD optimization.
• A fixed, multiple 'a'-value strategy provides favorable dose-volume optimization for challenging RVR regions in IMRT.
• This approach reduces reliance on planner-specific empirical adjustments for complex treatment planning scenarios.