An adaptive spot placement method on Cartesian grid for pencil beam scanning proton therapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Pencil beam scanning proton radiotherapy (RT) enables precise dose delivery, sparing healthy tissues.
• Current non-adaptive sampling (NS) methods use a fixed grid, potentially leading to suboptimal treatment plans.
• NS can be inefficient and compromise plan quality due to constant spot density regardless of tumor geometry.

Purpose of the Study:

• To develop and evaluate an adaptive sampling (AS) spot placement method for proton RT.
• To improve target dose conformality and delivery efficiency in proton therapy.
• To investigate AS's performance across different minimum monitor unit (MMU) constraints.

Main Methods:

• Developed an adaptive sampling (AS) spot placement strategy on a Cartesian grid.
• AS adjusts spot density based on tumor target geometry and uncertainties.
• AS utilizes finer spot grids at target boundaries and coarser grids in the interior.

Main Results:

• AS achieved comparable plan quality to NS for regular MMU, using significantly fewer spots (~10%).
• AS substantially improved plan quality over NS for large MMU.
• With similar spot counts, AS consistently yielded better plan quality than NS for both regular and large MMU.

Conclusions:

• AS offers a superior approach to spot placement in pencil beam scanning proton RT.
• The AS method enhances treatment plan quality and delivery efficiency.
• AS provides a robust and adaptable strategy for proton therapy planning.