A Study Of Multi-objective Optimization Algorithms In IMRT Inverse Planning Based On NSGA-Ⅲ Algorithm

Radiotherapy Radiation therapy is one of the three conventional means of treating cancer patients by using the energy of emitted rays to irradiate the patient’s target area in order to kill cancer cells and cure the patient.Radiation therapy is a selective local treatment that concentrates irradiation on specific target areas of patients,but other organs near the target area of the lesion,such as the spinal cord and heart,are exposed to certain radiation.To achieve effective killing of cancer cells,it is necessary to have a sufficiently high dose and uniform distribution to achieve maximum control of the tumor while minimizing radiation damage to the surrounding normal tissues and organs.Therefore,in order to achieve the requirement of precise radiation therapy,the research in this paper focuses on two aspects of clinical patient positioning error prediction and intensity-modulated radiation treatment planning optimization.In order to select a better radiation treatment plan from the clinical perspective,the patient’s positional error target area and the outward expansion boundary of PTV should be considered when formulating the radiation treatment plan.In this paper,we collected clinical positional error data sets of 50 lung cancer patients treated with IMRT,and used GMM model to predict the distribution pattern of positional error,and obtained the distribution pattern and probability prediction of positional error after analyzing the parameters,and also calculated the extrapolation boundary of PTV to provide reference for IMRT radiation treatment planning.The goal of the physiatrist in developing a radiotherapy plan is to try to select the best radiotherapy regimen in order to administer a high dose of radiation in the target area and to cause maximum damage to normal tissues and vital organs within the target area.Therefore,the radiotherapy plan must be optimized when selecting the radiotherapy plan.The algorithm used for optimization of the current domestic IMRT inverse planning system is multi-objective in nature,and the theory of the algorithm used for optimization needs to be updated.In this paper,we propose an IMRT inverse planning optimization method based on the NSGA-III algorithm to provide a new algorithmic reference for multi-objective optimization of the intensity modulated radiotherapy inverse planning.Since there are many parameters to be optimized in IMRT,it is difficult to optimize all of them by computer,and we usually select some of them for optimization.In this paper,the subfield intensity is used as the decision variable for optimization,and the simulation model of IMRT irradiation is established in MATLAB platform by using the arithmetic simulation method,sampling the PTV,OAR(Organization At Risk)and NT(Normal Tissue)of the planned target area,and using the point source model for each sampling point for dose calculation The new dose distribution matrix was obtained and compared with the clinical prescription dose,and the optimized plan was validated to meet the clinical requirements.