A Research On The Influence Of Accelerator Delivery Errors On Volumetric Modulated Arc Therapy

Objective:The purpose of this study was to investigate the dosimetric,2D dose verification,3D anatomical dose verification effect of monitor unit(MU)errors,gantry rotation angle errors,collimator position errors,and multi-leaf collimator(MLC)position errors for nasopharyngeal carcinoma(NPC)and cervical cancer volumetric modulated arc therapy(VMAT),and study the their correlation,Sensitivity and consistency.Materials and Methods:10 patients with NPC and 10 patients with cervical carcinoma were selected respectively.And to simulate VMAT delivery errors,extract the.Trail file form the Pinnacle~3 V9.2 planning system and use the program written by Matlab to read and modify each control point parameter,such as the actual operation errors of each parameter of accelerator,0.5°,1°and 2°gantry rotation angle errors,1%and 2%MU errors,1mm,2mm and 3mm collimator close,open and shift errors,and0.5mm,1mm and 2mm MLC close,open and shift errors.As a result,each clinical plans produced 24 modify plans.By comparing the clinical plans dose and modify plans dose base on TPS calculate,to evaluate the dosimetry effect of DVH-based dosimetry and dose verification at Gamma index were 2%/1mm,2%/2mm,3%/2mm and 3%/3mm.In addition,Spearman's rank correlation coefficient,receiver operating characteristic curve,and Kappa calculation method were used to calculate the correlation,sensitivity and consistency of dosimetry,2D%GP and 3D%GP,respectively.Results:In this paper,it was discovered that systematic MU errors,systematic accelerator gantry rotation angle errors,systematic collimator position errors and systematic MLC shift errors lead to the maximum changes of the PTV dose limit was less than 2.12%,the mean 2D%GP and 3D%GP were>95%criterion for a Gamma index of 3%/2 mm,and there were no strong correlation and consistency with DVH-based(|r|<0.8,kappa<0.8),no strong sensitivity to dose error(AUC<0.9).Moreover,for MLC close and open errors,the correlation and consistency between the global 3D%GP based on the anatomical structure,the 2D GP%based on the center plane,the 2D%GP based on the face of the detectors and the dose limit of target were lower in turn,and the sensitivity to the dose difference is also reduced in turn.For NPC,the correlation and consistency of%GP and DVH-based change were stronger with Gamma index was 2%/2mm.In addition,there were no significant correlation and consistency between the anatomical based 3D%GP and the 2D%GP(|r|<0.8,Kappa<0.8),and medium or strong correlation and consistency for 3D%GP based on specific structure(0.66<r<0.96,Kappa value range was 0.610~0.987,P<0.001).And there were strong correlation and consistency between the 2D GP%based on the center plane and the 2D%GP based on the face of the detectors(r>0.8,Kappa>0.8,P<0.001).Conclusion:No significant effect was found for MU errors,accelerator gantry rotation angle errors,collimator position errors and MLC shift errors for NPC and cervical cancer VMAT patients.Compared to the commonly used 2D%GP based on the face of the detectors,the 2D%GP based on the center plane is more reflective of clinically relevant dosimetric,Moreover,the 3D anatomical dose verification method is superior in clinic-related quality assurance.Therefore,for VMAT QA,medical institutions should choose the appropriate Gamma analysis method and Gamma index according to the existing equipment.