Radiotherapy Planning Using an Improved Search Strategy in Particle Swarm Optimization

Abstract

Objective Evolutionary stochastic global optimization algorithms are widely used in large-scale, non-convex problems. However, enhancing the search efficiency and repeatability of these techniques often requires well-customized approaches. This study investigates one such approach. Methods We use particle swarm optimization (PSO) algorithm to solve a 4-dimensional radiation therapy (RT) inverse planning problem, where the key idea is to use respiratory motion as an additional degree of freedom in lung cancer RT. The primary goal is to administer a lethal dose to the tumor target while sparing surrounding healthy tissue. Our iteratively adjusts radiation fluence-weights for all beam apertures across all respiratory phases. We implement three PSO-based approaches: conventionally-used unconstrained, hard-constrained and our proposed virtual search. As proof of concept, five lung cancer patient cases are optimized over ten runs using each PSO approach. For comparison, a dynamically penalized likelihood (DPL) algorithm- a popular RT optimization technique is also implemented and used. Results The proposed technique significantly improves the robustness to random initialization while requiring fewer iteration cycles to converge across all cases. DPL manages to find the global optimum in 2 out of 5 RT cases over significantly more iterations. Conclusion The proposed virtual search approach boosts the swarm search efficiency and, consequently, improves the optimization convergence rate and robustness for PSO. Significance RT planning is a large-scale, non-convex optimization problem, where finding optimal solutions in a clinically practical time is critical. Our proposed approach can potentially improve the optimization efficiency in similar time-sensitive problems.

Inversed-Planned Respiratory Phase Gating in Lung Conformal Radiation Therapy

“…We used 20 particles and 30 iterations (10 iterations per OS) for all optimization runs and employed the same objective function as in [23]. In each iteration cycle, the total dose ( D Total ) was calculated as…”

“…We used the objective function introduced by Modiri et al in [23], where dose-volume constraints were included, and therefore, the problem was non-convex in nature. PSO’s global search and parallelizability were its key features for being chosen as our solver.…”

“…PSO’s global search and parallelizability were its key features for being chosen as our solver. PSO has been successfully used in radiotherapy inverse planning of non-convex solution space in previous studies such as [23]. Our MATLAB prototype took ≤1.5 hrs to optimize each case; however, this process time may potentially be significantly reduced using compiled language platforms.…”

“…We use an inverse planning approach that employs a metaheuristic global optimization algorithm, particle swarm optimization (PSO [22]), to determine the optimal MUs per beam and identify an optimal gating phase for each beam. Similar approach has previously been demonstrated by Modiri et al for 4D treatment planning in [23]. For evaluation, we compare the resulting optimal plans with their equivalent single-phase EE-gated plans in terms of dosimetric quality and beam-on time.…”

Individualized estimates of overall survival in radiation therapy plan optimization — A concept study

“…We used inverse planning and particle swarm optimization (PSO), a highly parallelized, metaheuristic, global optimization algorithm, to minimize our modeled overall risk. PSO has been customized and successfully used in various other RT inverse planning studies with nonconvex solution spaces . We allowed the optimization algorithm to choose among a large set of gantry angles that included, but was not limited to, the beam angles routinely used in clinical practice.…”