Controlling the dose distribution with gEUD-type constraints within the convex radiotherapy optimization framework.

Area of Science:

• Medical Physics
• Computational Oncology
• Radiotherapy Treatment Planning

Background:

• Intensity modulated radiation therapy (IMRT) has advanced cancer treatment.
• Treatment planning optimization is computationally complex due to non-convex models.
• The potential for improved treatment plans with existing technology remains an open question.

Purpose of the Study:

• To investigate the integration of dose-volume histogram (DVH) prescriptions into convex optimization for inverse radiotherapy planning.
• To assess the efficacy of generalized moment constraints (GMCs) in approximating DVHs.
• To establish an analytic relationship between DVHs and generalized equivalent uniform dose (gEUD) values.

Main Methods:

• Developed a convex optimization framework incorporating DVH constraints using generalized moment constraints (GMCs).
• Established an analytic relationship between DVHs and a sequence of gEUD values.
• Conducted computational studies, focusing on the rectum DVH, to evaluate the proposed approach.

Main Results:

• The proposed convex optimization approach effectively approximates DVH prescriptions.
• An analytic relationship between DVHs and gEUD values was established.
• The method demonstrated promising results in computational studies, particularly for the rectum.

Conclusions:

• The novel convex optimization approach offers a feasible method for improving radiotherapy treatment planning.
• This approach simplifies the complex problem of DVH optimization, moving away from computationally intensive mixed-integer models.
• The method is expected to be implementable on conventional treatment planning systems, enhancing accessibility to advanced radiotherapy techniques.