论文部分内容阅读
Purpose:To improve an in-house fully automatic treatment planning system for radiotherapy of all kinds of cancer.Methods:The in-house autoplan system is a patent expert system designed to be embedded in a treatment planning system that has integrated experts' clinical and physics experience,and novel algorithms.For every new patient,the autoplan system could fully automatically generate plans that meet or exceed the institution's dose-volume constraints by one button click in the TPS.Previously,several kinds of cases including lung cancer,esophagus cancer,prostate cancer,base of tongue cancer can be applied to the autoplan system.More than 100 clinical plans of different kinds of cases including nasopharynx cancer,oralpharynx cancer,hypopharynx cancer,larynx cancer,tonsil cancer,parotid cancer,head and neck NK/T cell lymphoma,sinonasal carcinoma,lymphoma,thymoma,mesothelioma,breast cancer,cervical cancer,endometrial cancer,rectum cancer,bladder cancer,sarcoma,malignant melanoma,and different kinds of cancer with bone metastasis from 6 institutions were redesigned manually to summarize general optimization methods for every kinds of case.The general optimization method can nearly achieve the best OAR and normal tissue protection without too much target coverage compromise.The general optimization methods were then integrated in the autoplan system.Autoplans of these cases designed by the autoplan system were compared with these clinical plans by medical physicists and physicians.Results:IMRT and VMAT Plans of these kinds of cases can be generated automatically within 0.5 to 3 hours(1h on average)by one button click in the TPS.Every autoplans meet or exceed the institution's dose-volume constraints.Dose of all the OARs and normal tissue of autoplans is significant lower or not higher than that of clinical plans with similar target dose coverage.All the autoplans were preferred.CONCLUSIONS:The in-house fully automatic treatment planning system was improved for radiotherapy of all kinds of cancer.Plan quality was significant improved and human resources were greatly saved.