A new sparse optimization scheme for simultaneous beam angle and fluence map optimization in radiotherapy planning.

Area of Science:

• Medical Physics
• Computational Optimization
• Radiation Oncology

Background:

• Beam Angle Optimization (BAO) is crucial for Intensity-Modulated Radiation Therapy (IMRT) planning.
• Existing L1-minimization methods for BAO yield suboptimal treatment plans.
• Gradient-based methods offer improvements but may have limitations.

Purpose of the Study:

• To introduce a novel sparse optimization framework for BAO in IMRT.
• To address inconsistencies in L1-minimization approaches by incorporating group variable selection.
• To enhance both the quality of IMRT plans and computational efficiency.

Main Methods:

• Proposed a new sparse optimization framework using the group-folded concave penalty (gFCP).
• Employed a modified gradient-based method to solve the gFCP formulation.
• Evaluated the gFCP scheme against L1-minimization and Gradient Norm Method (GNM) on prostate, head-and-neck, and liver IMRT cases.

Main Results:

• The gFCP-based scheme outperformed L1-minimization in plan quality across all tested IMRT cases.
• gFCP demonstrated comparable computation times to L1-minimization.
• Compared to GNM, gFCP improved both plan quality and computational efficiency.

Conclusions:

• The proposed gFCP-based sparse optimization framework offers a promising advancement for BAO in IMRT.
• This method has the potential to reduce treatment planning time.
• The gFCP approach can lead to superior IMRT plan quality.