Click here to


Are you sure ?

Yes, do it No, cancel

Treatment Time Cutter Ants: A Novel Swarm-Intelligence Approach to Reduce Trimmer Sequencing Times

B Smith1*, D Hyer2 , R Flynn2 , P Hill3 , W Culberson1 , (1) Department of Medical Physics, School of Medicine and Public Health, University of Wisconsin, Madison, WI (2) Department of Radiation Oncology, University of Iowa, Iowa City, Iowa (3) Department of Human Oncology, School of Medicine and Public Health, University of Wisconsin, Madison, WI


(Tuesday, 7/16/2019) 7:30 AM - 9:30 AM

Room: Stars at Night Ballroom 1

Purpose: The dynamic collimation system (DCS) is an energy-layer specific collimator which utilizes four trimmer blades to intercept a scanned proton beam as it nears the target periphery. While the DCS has been shown to improve target dose conformity, additional treatment time is needed since the trimmers move slower than the scanned proton beam. The purpose of this work was to develop an algorithm to determine the optimal path for the trimmers to take during beam delivery to minimize the trimmer sequencing time.

Methods: It was assumed that the beam scanning time was negligible compared to a corresponding trimmer movement. Following this assumption, ant colony optimization (ACO) methods were used to determine the optimal trimmer sequencing route for three intracranial treatments. Analogous to a swarm of ants searching for food, each history, or ant, during ACO augments the next iteration based on likelihood of success. The optimization considered acceleration and a maximum operational trimmer speed of 2.50 m/s. Once a trimmer path was selected using ACO, the trimmer sequencing time was evaluated and if more than 65 seconds, beam spots were forced to share a common set of trimmers until the time goal was achieved.

Results: Trimmer sequencing times of 60 seconds were obtained for all treatment plans with variable levels of spot grouping. For each of the treatment plans, ACO treatment plans were nearly identical to their idealized counterpart plan achieving the same target coverage and had a negligible impact on the target conformity. Beam spots had to share common sets of trimmer positions to meet the time goal. Without grouping, treatment delivery times ranged between 121.6 to 283.4 seconds, and randomly chosen sequence routes without ACO could result in longer delivery times.

Conclusion: Swarm-intelligence ACO methods can rapidly and successfully minimize the trimmer sequencing time of DCS treatments.

Funding Support, Disclosures, and Conflict of Interest: Mr. Blake Smith and Dr. Wesley Culberson have nothing to disclose. Dr. Daniel Hyer, Dr. Ryan Flynn, and Dr. Patrick Hill are co-inventors on a patent that has been licensed to IBA.


Collimation, Protons, Optimization


TH- External Beam- Particle therapy: Proton therapy - dose optimization

Contact Email