Search

Purpose: We sought to develop machine learning models to predict the results of patient‐specific quality assurance (QA) for volumetric modulated arc therapy (VMAT), which were represented by several dose‐evaluation metrics—including the gamma passing rates (GPRs)—and criteria based on the radiomic features of 3D dose distribution in a phantom. Methods: A total of 4,250 radiomic features of 3D dose distribution in a cylindrical dummy phantom for 140 arcs from 106 clinical VMAT plans were extracted. We obtained the following dose‐evaluation metrics: GPRs with global and local normalization, the dose difference (DD) in 1% and 2% passing rates (DD1% and DD2%) for 10% and 50% dose threshold, and the distance‐to‐agreement in 1‐mm and 2‐mm passing rates (DTA1 mm and DTA2 mm) for 0.5%/mm and 1.0%.mm dose gradient threshold determined by measurement using a diode array in patient‐specific QA. The machine learning regression models for predicting the values of the dose‐evaluation metrics using the radiomic features were developed based on the elastic net (EN) and extra trees (ET) models. The feature selection and tuning of hyperparameters were performed with nested cross‐validation in which four‐fold cross‐validation is used within the inner loop, and the performance of each model was evaluated in terms of the root mean square error (RMSE), the mean absolute error (MAE), and Spearman's rank correlation coefficient. Results: The RMSE and MAE for the developed machine learning models ranged from <1% to nearly <10% depending on the dose‐evaluation metric, the criteria, and dose and dose gradient thresholds used for both machine learning models. It was advantageous to focus on high dose region for predicating global GPR, DDs, and DTAs. For certain metrics and criteria, it was possible to create models applicable for patients' heterogeneity by training only with dose distributions in phantom. Conclusions: The developed machine learning models showed high performance for predicting dose‐evaluation metrics especially for high dose region depending on the metric and criteria. Our results demonstrate that the radiomic features of dose distribution can be considered good indicators of the plan complexity and useful in predicting measured dose evaluation metrics. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this study was to create and evaluate deep learning‐based models to detect and classify errors of multi‐leaf collimator (MLC) modeling parameters in volumetric modulated radiation therapy (VMAT), namely the transmission factor (TF) and the dosimetric leaf gap (DLG). Methods: A total of 33 clinical VMAT plans for prostate and head‐and‐neck cancer were used, assuming a cylindrical and homogeneous phantom, and error plans were created by altering the original value of the TF and the DLG by ± 10, 20, and 30% in the treatment planning system (TPS). The Gaussian filters of σ=0.5$\sigma = 0.5$ and 1.0 were applied to the planar dose maps of the error‐free plan to mimic the measurement dose map, and thus dose difference maps between the error‐free and error plans were obtained. We evaluated 3 deep learning‐based models, created to perform the following detections/classifications: (1) error‐free versus TF error, (2) error‐free versus DLG error, and (3) TF versus DLG error. Models to classify the sign of the errors were also created and evaluated. A gamma analysis was performed for comparison. Results: The detection and classification of TF and DLG error were feasible for σ=0.5$\sigma = 0.5$; however, a considerable reduction of accuracy was observed for σ=1.0$\sigma = 1.0$ depending on the magnitude of error and treatment site. The sign of errors was detectable by the specifically trained models for σ=0.5$\sigma = 0.5$ and 1.0. The gamma analysis could not detect errors. Conclusions: We demonstrated that the deep learning‐based models could feasibly detect and classify TF and DLG errors in VMAT dose distributions, depending on the magnitude of the error, treatment site, and the degree of mimicked measurement doses. [ABSTRACT FROM AUTHOR]

• Selection criteria for circular cone and MLC for CyberKnife SRS were evaluated. • A mathematical score was adapted to assess the quality of the dose distribution. • MLC is recommended when the tumor radius exceeds 7.6 mm in metastatic brain tumors. • In the benign brain tumors, MLC is recommended when the tumor radius exceeds 17.5 mm. • Collimator selection based on these criteria improves plan quality of CyberKnife SRS. This study aimed to determine the selection criteria for circular collimator (CC)- and multileaf collimator (MLC)-based stereotactic radiosurgery (SRS)/stereotactic radiotherapy (SRT) plans for brain metastases (BM) and benign intracranial disease (BID) in terms of geometric parameters using CyberKnife (CK). Forty-eight and eighty-five patients with BM and BID, respectively, were included. Two plans using CC and MLC were created for each case. Six dosimetric parameters and mathematical scores (MS) were extracted from each plan to assess plan quality. Two geometric parameters in BM—equivalent radius (rGTV) and sphericity index (SI) of the gross tumor volume—and three in BID—rGTV, SI, and the distance between the GTV and organ at risk (dOAR)—were calculated. Their effect on the superiority of CC– or MLC-based plans in terms of dosimetric parameters and MS was evaluated using multiple regression analysis. The rGTV was associated with improved dosimetric parameters of MLC-based plans, especially the GTV conformity in BM and BID cases (β: 0.70 and 0.51) and the OAR sparing in BM cases (β: 0.82), where β represents the regression coefficient. Based on the MS, where the weights for the GTV coverage and OAR sparing were equal, the thresholds at which the MLC-based plans become comparable or superior to the CC-based plans in BM and BID were rGTV > 7.6 and >17.5 mm, respectively. Meanwhile, SI and dOAR were weakly correlated (β ≤ 0.30). In SRS/SRT plans for BM and BID cases using CyberKnife, geometric parameters, especially rGTV, must be considered when selecting CC or MLC. [ABSTRACT FROM AUTHOR]

There are no quantitative selection criteria for identifying high-grade glioma (HGG) patients who are suited for volumetric-modulated arc therapy (VMAT). This study aimed to develop selection criteria that can be used for the selection of the optimal treatment modality in HGG. We analyzed 20 patients with HGG treated by 3D conformal radiotherapy (3DCRT). First, VMAT plans were created for each patient retrospectively. For each plan, the normal tissue complication probability (NTCP) for normal brain was calculated. We then divided the patients based on the NTCP s of the 3DCRT plans for normal brain, using the threshold of 5%. We compared the NTCP s of the two plans and the gross tumor volumes (GTV s) of the two groups. For the GTV s, we used receiver operating characteristic curves to identify the cut-off value for predicting NTCP < 5%. We determined the respective correlations between the GTV and the GTV's largest cross-sectional diameter and largest cross-sectional area. In the NTCP ≥ 5% group, the NTCP s for the VMAT plans were significantly lower than those for the 3DCRT plans (P = 0.0011). The NTCP ≥ 5% group's GTV was significantly larger than that of the NTCP < 5% group (P = 0.0016), and the cut-off value of the GTV was 130.5 cm3. The GTV was strongly correlated with the GTV 's largest cross-sectional diameter (R 2 = 0.82) and largest cross-sectional area (R 2 = 0.94), which produced the cut-off values of 7.5 cm and 41 cm2, respectively. It was concluded that VMAT is more appropriate than 3DCRT in cases in which the GTV is ≥130.5 cm3. [ABSTRACT FROM AUTHOR]