Automatic treatment plan re-optimization for adaptive radiotherapy guided with the initial plan DVHs.

Area of Science:

• Radiation Oncology
• Medical Physics
• Computational Biology

Background:

• Adaptive radiation therapy (ART) aims to minimize normal tissue toxicity and enhance tumor control by adapting treatments to patient anatomy.
• Manual re-planning for ART is time-consuming and impractical for clinical use, necessitating automated solutions.
• Existing ART methods can leverage prior treatment plan information, like dose-volume histograms (DVHs), to guide re-planning.

Purpose of the Study:

• To develop an automated re-planning algorithm for ART that generates plans with DVH curves similar to or better than the original clinical plan.
• To create an efficient and effective re-planning algorithm for the clinical realization of ART.

Main Methods:

• An iterative algorithm combining fluence map optimization and adaptive voxel weighting factors based on DVH deviations.
• The inner loop optimizes dose using a quadratic objective function, while the outer loop adjusts voxel weights to match original DVH curves.
• The algorithm was implemented on a GPU for high efficiency and tested on three head-and-neck cancer intensity-modulated radiation therapy (IMRT) cases.

Main Results:

• The automated re-planning algorithm achieved DVH curves superior to the original plan for most structures after 30 iterations.
• The re-optimization process was highly efficient, completing in approximately 30 seconds using an in-house optimization engine.
• The algorithm demonstrated feasibility in head-and-neck cancer IMRT cases using treatment planning CT scans.

Conclusions:

• The developed automated re-planning algorithm is efficient and effective for ART.
• This approach can generate high-quality treatment plans with improved DVH metrics compared to original plans.
• The algorithm shows promise for clinical implementation, facilitating faster and more accurate adaptive treatments.