A linear programming approach to inverse planning in Gamma Knife radiosurgery

Area of Science:

• Medical Physics
• Radiosurgery Technology
• Computational Optimization

Background:

• Leksell Gamma Knife enables precise dose delivery in stereotactic radiosurgery.
• Previous inverse planning approaches had limitations and lacked advanced capabilities.

Purpose of the Study:

• To introduce a novel inverse planning approach for Leksell Gamma Knife.
• To address shortcomings of prior methods and enable new treatment planning functionalities.

Main Methods:

• Utilized sector-duration optimization with fixed isocenters via linear programming.
• Investigated beam-on time penalization for plan quality trade-offs.
• Implemented dualization and representative subsampling to reduce problem size and solution time.

Main Results:

• Beam-on time was reduced by 2-3 times compared to standard methods.
• Dualization and subsampling yielded 5-20 times faster optimization.
• New plans showed comparable coverage, better selectivity, and reduced beam-on time versus 75 clinical plans.
• Improved gradient index was achieved in 44% of cases.
• Optimization times ranged from 2.3 to 26 seconds (median 5.7 seconds).

Conclusions:

• A combination of techniques enables clinically feasible sector-duration optimization.
• The new approach enhances treatment planning efficiency and plan quality for Gamma Knife radiosurgery.