Feasibility of case-based beam generation for robotic radiosurgery

Area of Science:

• Medical Physics
• Radiation Oncology
• Robotics in Medicine

Background:

• Robotic radiosurgery offers precise tumor targeting by leveraging a robotic arm's flexibility.
• Optimizing treatment beam placement in robotic radiosurgery presents a significant planning challenge.
• Existing methods often rely on heuristics with randomly selected candidate beams.

Purpose of the Study:

• To develop a more efficient method for selecting treatment beams in robotic radiosurgery.
• To reduce the computational time required for treatment planning.
• To maintain or improve the quality of radiation dose distribution.

Main Methods:

• A case-based reasoning approach was developed to generate candidate beams.
• Similarity metrics were defined based on anatomical structures and dose distributions.
• A subset of treatment beams from previous plans was adapted for new cases.
• The inverse planning problem was solved using the generated candidate beams.

Main Results:

• The novel approach yielded comparable plan quality to conventional methods.
• Fewer candidate beams were required, reducing the search space.
• Treatment planning time was reduced by up to 47% (from ~19 min to <11 min).
• Dose homogeneity and critical structure sparing were maintained in prostate cancer cases.

Conclusions:

• Case-based beam generation is feasible for robotic radiosurgery.
• This approach can substantially decrease planning time for similar clinical cases.
• High plan quality can be preserved while improving efficiency.