Heuristic optimization of the scanning path of particle therapy beams

Abstract

Purpose: Quasi-discrete scannig is a delivery strategy for proton and ion beam therapy in which the beam is turned off when a slice is finished and a new energy must be set, but not during the scanning between consecutive spots. Different scanning paths lead to different dose distributions due to the contribution of the unintended transit dose between spots. In this work an algorithm to optimize the scanning path for quasi-discrete scanned beams is presented.Methods: The classical simulated annealing algorithm is used. It is a heuristic algorithm frequently used in combinatorial optimization problems, which allows to obtain nearly optimal solutions in acceptable running times.Results: A study focused on the best choice of operational parameters on which the algorithm performance depends is presented. The convergence properties of the algorithm have been further improved by using the next-neighbour algorithm to generate the starting paths. Conclusions: The algorithm implemented for this work can optimize efficiently the scanning path of quasi-discrete scanned particle beams. Optimized scanning paths decrease the transit dose and lead to better dose distributions.

Scan path optimization with/without clustering for active beam delivery in charged particle therapy

“…Shorter paths were obtained and consequently the treatment time was reduced. In a study by Pardo et al the delivered dose distribution using both algorithms (SA and zigzag algorithm) was quantified: results showed that a reduction in the transit particles was achieved for the paths obtained with the SA algorithm [8]. The main goal of this work is to compare the clinical/economical implications of using different scan path optimization strategies with what is currently being used (scan paths provided by Syngo RT Planning) in real treatment plans.…”

“…However, this leads to a sub-optimal use of the equipment, resulting in (i) a waste of particles spilled at each synchrotron pulse for non optimized paths and (ii) an increase in the treatment time due to the delay caused by the beam deflection activation. These reasons and the fact that shorter paths can reduce transit particles [8] motivated the search of competitive methods for the optimization of the scan path, aiming at cost-effective treatment delivery in quasidiscrete scanning.…”

“…transit particles) are assigned to the destination spot. However, by doing this, different paths may lead to different dose distributions due to different transit doses [8]. In order to overcome this issue a solution relying on magnetic beam deflection outside of the extraction line using a dedicated device (beam chopper) is applied after the scan paths have been determined.…”

“…The importance of scan path optimization in quasidiscrete scanning was recently discussed in two different studies [3,8]. In both studies a simulated annealing (SA) algorithm was used to calculate the scanning paths.…”

“…The broader interested in quasidiscrete scanning is mainly due to the faster repainting that it offers, important in the presence of moving targets [5e7]. Due to the fact that the irradiation beam is never turned off within a single slice, the particles delivered in between irradiation spots (transit particles) may lead to changes in the delivered dose distribution [8]. Moreover, the scan path also plays an important role in the presence of organs at risk located in vacancies within the overall spot distribution and in presence of moving targets, where fast repainting may be required to mitigate motion effects [3,5e8].…”