The purpose of this study was to develop a predictive model for patient‐specific VMAT QA results using multileaf collimator (MLC) effect and texture analysis. The MLC speed, acceleration and texture analysis features were extracted from 106 VMAT plans as predictors. Gamma passing rate (GPR) was collected as a response class with gamma criteria of 2%/2 mm and 3%/2 mm. The model was trained using two machine learning methods: AdaBoost classification and bagged regression trees model. GPR was classified into the “PASS” and “FAIL” for the classification model using the institutional warning level. The accuracy of the model was assessed using sensitivity and specificity. In addition, the accuracy of the regression model was determined using the difference between predicted and measured GPR. For the AdaBoost classification model, the sensitivity/specificity was 94.12%/100% and 63.63%/53.13% at gamma criteria of 2%/2 mm and 3%/2 mm, respectively. For the bagged regression trees model, the sensitivity/specificity was 94.12%/91.89% and 61.18%/68.75% at gamma criteria of 2%/2 mm and 3%/2 mm, respectively. The root mean square error (RMSE) of difference between predicted and measured GPR was found at 2.44 and 1.22 for gamma criteria of 2%/2 mm and 3%/2 mm, respectively. The promising result was found at tighter gamma criteria 2%/2 mm with 94.12% sensitivity (both bagged regression trees and AdaBoost classification model) and 100% specificity (AdaBoost classification model).
The purpose of this study was to compare three computed tomography (CT) images under different conditions—average intensity projection (AIP), free breathing (FB), mid‐ventilation (MidV)—used for radiotherapy contouring and planning in lung cancer patients. Two image sets derived from four‐dimensional CT (4DCT) acquisition (AIP and MidV) and three‐dimensional CT with FB were generated and used to plan for 29 lung cancer patients. Organs at risk (OARs) were delineated for each image. AIP images were calculated with 3D conformal radiotherapy (3DCRT) and intensity‐modulated radiation therapy (IMRT). Planning with the same target coverage was applied to the FB and MidV image sets. Plans with small and large tumors were compared regarding OAR volumes, geometrical center differences in OARs, and dosimetric indices. A gamma index analysis was also performed to compare dose distributions. There were no significant differences (P > 0.05) in OAR volumes, the geometrical center differences, maximum and mean doses of the OARs between both tumor sizes. For 3DCRT, the gamma analysis results indicated an acceptable dose distribution agreement of 95% with 2%/2 mm criteria. Although, the gamma index results show distinct contrast of dose distribution outside the planning target volume (PTV) in IMRT, but within the PTV, it was acceptable. All three images could be used for OAR delineation and dose calculation in lung cancer. AIP image sets seemed to be suitable for dose calculation while patient movement between series acquisition of FB images should be considered when defining target volumes on 4DCT images.
The study’s purpose was to develop and validate Electronic Portal Imaging Device (EPID)-based dosimetry for Stereotactic Radiosurgery (SRS) and Stereotactic Radiation Therapy (SRT) patient-specific Quality Assurance (QA). The co-operation between extended Source-to-Imager Distance (SID) to reduce the saturation effect and simplify the EPID-based dosimetry model was used to perform patient-specific QA in SRS and SRT plans. The four parameters were included for converting the image to dose at depth 10 cm; dose-response linearity with MU, beam profile correction, collimator scatter and water kernel. The model accuracy was validated with 10 SRS/SRT plans. The traditional diode arrays with MapCHECK were also used to perform patient-specific QA for assuring model accuracy. The 150 cm-SID was found a possibility to reduce the saturation effect. The result of model accuracy was found good agreement between our EPID-based dosimetry and TPS calculation with GPR more than 98% for gamma criteria of 3%/3 mm, more than 95% for gamma criteria of 2%/2 mm, and the results related to the measurement with MapCHECK. This study demonstrated the method to perform SRT and SRT patient-specific QA using EPID-based dosimetry in the FFF beam by co-operating between the extended SID that can reduce the saturation effect and estimate the planar dose distribution with the in-house model.
The purpose of this study was to develop Electronic Portal Imaging Devices (EPID)-based dosimetry for Flattening-Filter-Free (FFF) beam verification. All radiation measurements were performed with source to imager distances (SID) of 150 cm to reduce saturation effect. EPID images were converted to radiation absorbed dose with our algorithm including four parameters: linearity of dose response with Monitor Unit (MU), beam profile correction, collimator scatter, and scatter kernel. The Calibration Units (CU) of image were scaled to dose (Gy) by using linearity of dose response with MU. Off-axis response differences between EPID and water were reduced with beam profile correction. Scatter kernel was applied to EPID images to reduce the residual error. The algorithm accuracy was validated with 12 arcs of Volumetric Modulated Arc Therapy (VMAT) plans by using gamma analysis comparing between EPID-based dosimetry and a plane dose distribution of Treatment Planning Systems (TPS). Gamma Passing Rates (GPR) were used to determine the dose agreements with criteria of 2%, 2 mm and 3%, 3 mm. The mean of GPR was 97.91%, and 99.62% for criteria of 2%,2 mm and, 3%, 3 mm, respectively. Our EPID-based dosimetry showed good agreement with plane dose distribution in water. These results indicated that our EPID-based dosimetry can perform FFF-beam verification.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.