Fluence map optimization (FMO) with dose-volume constraints in IMRT using the geometric distance sorting method.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Optimization

Background:

• Inverse planning in Intensity-Modulated Radiation Therapy (IMRT) requires optimizing fluence maps to meet dose-volume constraints.
• Traditional methods for handling these constraints can be inefficient, particularly for complex treatment geometries.

Purpose of the Study:

• To develop and evaluate a new heuristic algorithm for fluence map optimization using a geometric distance sorting technique.
• To improve the efficiency and stability of inverse planning by better managing dose-volume constraints.

Main Methods:

• An iterative approach starting with a basic optimization model, gradually incorporating dose-volume constraints.
• Utilizing an interior point method to solve quadratic programming problems in each iteration.
• Employing a novel geometric distance sorting technique to select voxels for constraint addition, minimizing objective function increases.
• Implementing a smart constraint adding/deleting strategy for stable convergence.

Main Results:

• The proposed geometric distance sorting method demonstrated superior performance in guiding constraint selection compared to traditional dose sorting techniques.
• The algorithm showed particular effectiveness for cases with non-convex target regions.
• Experimental results on head-neck, prostate, lung, and oropharyngeal cancer cases confirmed the algorithm's efficiency and stability.

Conclusions:

• The new heuristic algorithm offers a more efficient and robust approach to fluence map optimization with dose-volume constraints in IMRT.
• The geometric distance sorting technique represents a significant advancement over traditional methods, enhancing inverse planning for complex anatomical targets.