YAN Huawei,ZHANG Ji,LIN Zhixi, et al. The role of machine learning in dose verification of intensity modulated radiotherapy planning[J]. JOURNAL OF WEZHOU MEDICAL UNIVERSITY, 2023, 53(2): 137-141.
Abstract：Objective: To explore the role of machine learning in dose verification of intensity modulated radiotherapy planning. Methods: A total of 141 patients who received double-arc volumetric modulated arc therapy (VMAT) treatment in the First Affiliated Hospital of Wenzhou Medical University from March 2019 to August 2020 were enrolled, extracting 13 complexity parameters of intensity modulation plan, and gamma pass rate (GPR) under different conditions were collected. The data was randomly divided into the training set and the test set at a ratio of 7:3. The Pearson correlation analysis and LASSO were used to screen parameters. The models based on the machine learning method of SVM were used to predict values and classification of GPR, respectively. The root mean square error (RMSE) and mean square error (MAE) were used to evaluate the accuracy of model numerical prediction and the AUC to evaluate the accuracy of model classification. Result: In the numerical prediction of GPR, under the respective condition of 3% mm, 3% mm and 2% mm, the RMSE in the test set was 2.22, 3.51, 4.59, and MAE 1.56, 2.68, 3.67 accordingly. In the GPR classification prediction, under the respective condition of 3%/3 mm, 3%/2 mm, and 2%/2 mm, the AUC result of the test set was 0.79, 0.78, 0.77 accordingly. Conclusion: The dose verification of IMRT/VMAT plan based on machine learning is of clinical application value, which provides a new perspective for quality assurance.
