An automated planning optimization framework for cervical intensity-modulated radiation therapy using voxel dose...

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Knowledge-based planning in radiation therapy often struggles with prediction uncertainties.
• Managing these uncertainties is crucial for optimizing treatment plans and patient outcomes.

Purpose of the Study:

• To develop an automated framework that integrates voxel-dose prediction with adaptive optimization.
• To overcome limitations in knowledge-based planning by addressing prediction uncertainties.

Main Methods:

• An automated framework was developed, combining voxel-dose prediction with adaptive optimization.
• A fuzzy-inference system was employed to dynamically adjust objectives, correcting clinical guidance violations.
• The framework was tested on 27 cervical intensity-modulated radiation therapy (IMRT) cases.

Main Results:

• Significantly improved sparing of organs of interest (OOIs) was achieved.
• Rectal volume receiving 45 Gy was reduced by 5.5 ± 2.2% (p < 0.005).
• Target coverage comparable to clinical plans was maintained.

Conclusions:

• The developed framework successfully generates high-quality IMRT plans.
• Adaptive optimization effectively addresses prediction uncertainties in radiation therapy planning.
• This approach enhances organ sparing while ensuring adequate tumor coverage.