Optimization of equivalent uniform dose using the L-curve criterion

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Equivalent Uniform Dose (EUD) optimization in intensity-modulated radiation therapy (IMRT) aims to standardize radiobiological effects across treatment plans.
• A key challenge is the underdetermined nature of spatial dose distributions in EUD-based inverse planning, potentially increasing dose nonuniformity.
• This issue is compounded by the ill-posedness of the underlying integral equation.

Purpose of the Study:

• To develop a method for optimizing EUD in IMRT inverse planning while controlling dose nonuniformity.
• To implement an objective function incorporating beam intensity smoothness to mitigate underdetermination.

Main Methods:

• Incorporated a beam intensity smoothness criterion into an EUD-based objective function.
• Utilized a variational regularization technique, analogous to methods used in dose-based least-squares optimization.
• Employed the L-curve criterion to determine the optimal regularization parameter.

Main Results:

• The proposed approach effectively reduced dose nonuniformity in EUD-based inverse planning.
• Variational regularization, coupled with the L-curve criterion, demonstrated significant improvements in plan quality.
• The method ensures that optimized EUD aligns with prescribed values while managing dose variations.

Conclusions:

• Variational regularization is a viable technique to address dose nonuniformity in EUD-based IMRT inverse planning.
• This method offers a way to balance EUD optimization with the need for clinically acceptable dose distributions.
• The findings suggest improved consistency and reliability of radiotherapy treatment plans.