Research On Optimization Of Intensity Modulated Radiotherapy System Based On Swarm Intelligence Algorithm

Intensity Modulated Radiation Therapy(IMRT)is the most common radiotherapy technology for precise radiotherapy.It is a reverse planning process.Its purpose is to make the radiation high-dose area and the tumor target area highly conformal in the three-dimensional direction of the treatment plan,and to make the high dose concentrated in the tumor target area as much as possible,while the surrounding normal tissues and organs are exposed to the least dose.Therapy Planning System(TPS)is a software system that enables the design and evaluation of IMRT radiotherapy plans.It uses computer technology to automate the design of radiotherapy plans,and it is easy to get the best radiation therapy plan,which effectively improves the patient's tumor control probability and survival rate.In this paper,based on the TPS system,the finite pencil beam convolution model based on photon beam is used to determine the optimization objective function,and experimental research is carried out on various phantoms and cases.The dose calculation model used in this paper is a finite pencil beam convolution model,so the beam weight optimization is essentially to optimize the pencil beam weight.How to automatically select the field parameters in radiotherapy has caused a lot of attention.The direction of the field and its weight optimization are two essential aspects.This paper starts from the direction of the field and the optimization of the pencil beam weight,and uses the group intelligence algorithm to improve the optimized performance.Both the artificial bee colony algorithm and the genetic algorithm belong to the swarm intelligence algorithm.They are a new type of global optimization search technology.They solve problems by learning from nature,learn from the intelligent behavior of bee colonies and the evolutionary mechanisms of biological inheritance.n particular,they do not require the objective function to have constraints such as continuity,linear conditions,and derivative existence,and also have the characteristic of parallel computing,which makes it perform well in terms of speed and performance.The main results and innovations of this paper include:(1)In the problem of optimization of the direction of the field,this paper proposes artificial bee colony algorithm to optimize the direction of the field,combined with the prior knowledge of the field direction found by the predecessors,set constraints on the direction of the field,narrow the search space,and the initial solution is distributed more evenly on the search space.Finally,the advantage of the local search ability of the artificial bee colony algorithm is used to find a better solution in the local details,which accelerates the convergence speed of the optimization problem.(2)On the problem of pencil beam weight optimization,the size of the pencil beam weight is large,which affects the convergence speed.This paper proposes the artificial bee colony algorithm combined with genetic algorithm to optimize the weight of each finite pencil beam.By combining the strong local search ability of the artificial bee colony algorithm with the strong global search ability of the genetic algorithm,the two algorithms complement each other's deficiencies,improve the convergence speed of the optimization,and get the optimal setting of the pencil beam weights faster.