Explore the vast range of titles available.

In order to deliver accurate and safe treatment to cancer patients in radiation therapy using advanced techniques such as intensity modulated radiation therapy (IMRT) and volumetric‐arc radiation therapy (VMAT), patient specific quality assurance (QA) should be performed before treatment. IMRT/VMAT dose measurements in a phantom using various devices have been clinically adopted as standard method for QA. This approach allows the verification of the accuracy of the dose calculation, data transfer, and the delivery system. However, patient‐specific QA procedures are expensive and require significant time and effort by the physicists. Over the past 5 years, machine learning (ML) and deep learning (DL) algorithms for predictions of IMRT/VMAT QA outcome have been investigated. Various ML and DL models have shown promising prediction accuracy and a high potential as time‐efficient virtual QA tool. In this paper, we review the ML and DL based models that were developed for patient specific IMRT and VMAT QA outcome predictions from algorithmic and clinical applicability perspectives. We focus on comparing the algorithms, the dataset sizes, the input parameters and features, the QA outcome prediction approaches, the validation, the performance, the clinical applicability, and the potential clinical impact. In addition, we discuss the present challenges as well as the future directions in the implementation of these models. To the best of our knowledge, this is the first review on the application of ML and DL based models in IMRT/VMAT QA predictions.

Background: Quality assurance (QA) programs are designed to improve the quality and safety of radiation treatments, including patient-specific QA (PSQA). The objective of this study was to investigate the conditions in which pretreatment PSQA is performed, to evaluate the root cause of the implementation of more complex techniques, and to identify areas for potential improvement. Materials/Methods: The Octavius 4D (O4D) system accuracy was evaluated using an O4D homogeneous phantom for different field sizes. Tests of the system response to dose linearity, field sizes, and PDD differences were performed against calculated doses for a 6 MV photon beam. The pretreatment verification of 40 VMAT plans was performed using the PTW VeriSoft software (version 8.0.1) for local and global 3D gamma analysis. The reconstructed 3D dose was compared to the calculated dose using 2%/2 mm and 3%/3 mm, 20% of the low-dose threshold, and 95% of the gamma passing rate (%GP) tolerance level. The sensitivity of the O4D system in detecting VMAT delivery and setup errors has been investigated by measuring the variation in %GP values before and after the simulated errors. Results: The O4D system reported good agreement for linearity, field size, and PDD differences with TPS dose, being within ±2% tolerance. The output factors were consistent between the ionization chamber and the O4D detector down to a 4 × 4 cm2 field size with a maximum deviation less than 1%. The introduction of deliberate errors caused a decrease in %GP values. In most scenarios, the %GP value of the simulated errors was detected with 2%/2 mm. Conclusion: The results indicate that the O4D system is sensitive enough to detect delivery and setup errors with the restrictive global criterion of 2%/2 mm for routine pretreatment verification.

Purpose This study aimed to improve the safety and accuracy of radiotherapy by establishing tolerance (TL) and action (AL) limits for the gamma index in patient‐specific quality assurance (PSQA) for intensity‐modulated radiation therapy (IMRT) and volumetric‐modulated arc therapy (VMAT) using SunCHECK software, as per AAPM TG‐218 report recommendations. Methods The study included 125 patients divided into six groups by treatment regions (H&N, thoracic and pelvic) and techniques (VMAT, IMRT). SunCHECK was used to calculate the gamma passing rate (%GP) and dose error (%DE) for each patient, for the planning target volume and organs at risk (OARs). The TL and AL were then determined for each group according to TG‐218 recommendations. We conducted a comprehensive analysis to compare %DE among different groups and examined the relationship between %GP and %DE. Results The TL and AL of all groups were more stringent than the common standard as defined by the TG218 report. The TL and AL values of the groups differed significantly, and the values for the thoracic groups were lower for both VMAT and IMRT. The %DE of the parameters D95%, D90%, and Dmean in the planning target volume, and Dmean and Dmax in OARs were significantly different. The dose deviation of VMAT was larger than IMRT, especially in the thoracic group. A %GP and %DE correlation analysis showed a strong correlation for the planning target volume, but a weak correlation for the OARs. Additionally, a significant correlation existed between %GP of SunCHECK and Delta4. Conclusion The study established TL and AL values tailored to various anatomical regions and treatment techniques at our institution. Establishing PSQA workflows for VMAT and IMRT offers valuable clinical insights and guidance. We also suggest developing a standard combining clinically relevant metrics with %GP to evaluate PSQA results comprehensively.

Objective Measurement‐based patient specific quality assurance (PSQA) is an increasingly debated topic among medical physicists. Developments like online adaptive radiotherapy and same‐day stereotactic treatments limit the time to do measurement‐based PSQA. Herein, we develop a predictive machine learning model to supplement PSQA by predicting the gamma passing rate (GPR) per stereotactic arc. This streamlines PSQA, providing planners the insight to replan potentially sub‐optimal plans, to mitigate machine time inefficiencies. Methods 122 patients that had previously received HyperArc stereotactic radiosurgery/radiotherapy on a TrueBeam LINAC (Millenium 120 MLCs, 6MV‐FFF) were used to generate a long short‐term memory (LSTM) recurrent neural network to predict the GPR for a 2%/2 mm criteria. GPRs were discretized into three classes: Ideal (≥95%), Investigate [85%–95%), and Replan (<85%). In total, 468 VMAT arcs were used for this model with a class distribution of 370 (Ideal), 65 (Investigate), and 33 (Replan). To counteract the imbalanced data, the minority classes were over‐sampled using synthetic minority over‐sampling technique to generate a balanced dataset. The LSTM model was trained in Python with an 80‐20 training‐testing stratified split. Individual class sensitivity and specificity were recorded following a one versus all method. The final model was deployed clinically through Eclipse Scripting. Results The model demonstrated the following (sensitivity, specificity) for the testing data: Ideal (78.4%, 87.2%), Investigate (75.7%, 89.9%), and Replan (93.2%, 96.6%). The primary focus of this model is to identify failing beams and allow the planner to address this prior to running the PSQA, as such the Replan class was the most important for evaluation. A sensitivity of 93.2% indicates that the model will identify 93.2% of HyperArc plans that need to be replanned with a very high certainty due to the 96.6% specificity. Conclusions The predictive GPR model developed within this research enables HyperArc planners to immediately assess the GPR for each stereotactic arc and preemptively replan potentially failing arcs to optimize the PSQA machine time.

Purpose This study aims to develop an algorithm to predict gamma passing rate (GPR) in the volumetric‐modulated arc therapy (VMAT) technique. Materials and methods A total of 118 clinical VMAT plans, including 28 mediastina, 25 head and neck, 40 brains intensity‐modulated radiosurgery, and 25 prostate cases, were created in RayStation treatment planning system for Edge and TrueBeam linacs. In‐house scripts were developed to compute Modulation indices such as plan‐averaged beam area (PA), plan‐averaged beam irregularity (PI), total monitor unit (MU), leaf travel/arc length, mean dose rate variation, and mean gantry speed variation. Pretreatment verifications were performed on ArcCHECK phantom with SNC software. GPR was calculated with 3%/2 mm and 10% threshold. The dataset was randomly split into a training (70%) and a test (30%) dataset. A random forest regression (RFR) model and support vector regression (SVR) with linear kernel were trained to predict GPR using the complexity metrics as input. The prediction performance was evaluated by calculating the mean absolute error (MAE), R2, and root mean square error (RMSE). Results RMSEs at γ 3%/2 mm for RFR and SVR were 1.407 ± 0.103 and 1.447 ± 0.121, respectively. MAE was 1.14 ± 0.084 for RFR and 1.101 ± 0.09 for SVR. R2 was equal to 0.703 ± 0.027 and 0.689 ± 0.053 for RFR and SVR, respectively. GPR of 3%/2 mm with a 10% threshold can be predicted with an error smaller than 3% for 94% of plans using RFR and SVR models. The most important metrics that had the greatest impact on how accurately GPR can be predicted were determined to be the PA, PI, and total MU. Conclusion In terms of its prediction values and errors, SVR (linear) appeared to be comparable with RFR for this dataset. Based on our results, the PA, PI, and total MU calculations may be useful in guiding VMAT plan evaluation and ultimately reducing uncertainties in planning and radiation delivery.

Background In vivo transit dosimetry using an electronic portal imaging device (EPID‐IVTD) is an important tool for verifying the accuracy of radiation therapy treatments. Despite its potential, the implementation of EPID‐IVTD in breast intensity modulated radiation therapy (IMRT) treatments has not yet been standardized, limiting its clinical adoption. A standardized EPID‐IVTD method could enhance treatment accuracy and improve patient safety in routine clinical practice. Purpose This study aims to develop a method for EPID‐IVTD for whole breast IMRT treatment. Methods Gamma passing rates (GPRs) analysis was the basis of the work conducted on a dataset of 50 patients. The first phase of the work focused on the identification of the reference fraction. In the second phase a method for performing EPID‐IVTD was implemented. Lower‐tolerance and ‐action limits (l‐TL and l‐AL), as introduced by AAPM TG 218, were employed to determine the reference fraction and used as alert and alarm thresholds, respectively, in EPID‐IVTD monitoring. Results The first treatment fraction demonstrated the best dosimetric agreement with the theoretical plan and was therefore used as the reference in the second phase of the study. EPID‐IVTD results showed that 75% of the GPRs ranged from 97.5% to 99.9%, 93.83% were above the l‐TL, 4.31% fell between l‐TL and l‐AL, and 1.86% were below l‐AL. Conclusions A method for the implementation of an effective EPID‐IVTD in whole breast IMRT treatment was developed and is now routinely applied at our center, enabling efficient monitoring in clinical practice.

Background Volumetric modulated arc therapy (VMAT) requires an accurate patient‐specific quality assurance (PSQA) program. In clinical practice, this is usually performed using the γ‐index and the two‐dimensional gamma passing rate (2D %GP). A three‐dimensional (3D) index incorporating the patient anatomy could be more useful for the 3D dose distribution verification. Purpose The current study demonstrates a thorough investigation of VMAT PSQA treatment plans by examining the correlation between 3D Gamma passing rate (%GP) and 2D %GP. The aim was to establish the tolerance limits (TL) and action limits (AL) that could be adopted in clinical practice. Materials and methods PSQA was performed for 67 head and neck (H&N) and 69 prostate treatment plans, using an appropriate phantom and the γ‐index method. The 3%/2  mm acceptance criterion was used. Treatment plans’ 2D% GP and 3D %GP values were collected and correlated with individual 3D %GP values of planning target volume (PTV) and organs at risk (OARs). Institutional TL and AL of both 2D %GP and 3D %GP were established using 30 prostate and 30 H&N treatment plans, as per recommendations proposed by AAPM TG‐218. Results A moderate correlation was observed between 2D %GP and 3D %GP of the treatment plans. Τhe correlations demonstrated a stronger association for the 3D %GP than for the 2D %GP with respect to the 3D %GP of the individual structures considered. The TL for the 3D %GP (both plan and individual) and the plans’ 2D %GP were generally more stringent, while the AL showed a wider range compared to the corresponding limits suggested by the TG‐218 protocol for plan 2D %GP. Conclusions Institution‐specific 3D %GP as well as TL and AL for treatment plan, PTV and OARs could be incorporated in the PSQA procedure in synergy with the 2D evaluation, as they can provide a more‐in‐depth‐view of the treatment quality.

Background We investigate the gamma passing rate (GPR) consistency when applying different types of gamma analyses, linacs, and dosimeters for volumetric modulated arc therapy (VMAT). Methods A total of 240 VMAT plans for various treatment sites, which were generated with Trilogy (140 plans) and TrueBeam STx (100 plans), were retrospectively selected. For each VMAT plan, planar dose distributions were measured with both MapCHECK2 and ArcCHECK dosimeters. During the planar dose distribution measurements, the actual multileaf collimator (MLC) positions, gantry angles, and delivered monitor units were recorded and compared to the values in the original VMAT plans to calculate mechanical errors. For each VMAT plan, both the global and local gamma analyses were performed with 3%/3 mm, 2%/2 mm, 2%/1 mm, 1%/2 mm, and 1%/1 mm. The Pearson correlation coefficients ( r ) were calculated 1) between the global and the local GPRs, 2) between GPRs with the MapCHECK2 and the ArcCHECK dosimeters, 3) and between GPRs and the mechanical errors during the VMAT delivery. Results For the MapCHECK2 measurements, strong correlations between the global and local GPRs were observed only with 1%/2 mm and 1%/1 mm ( r  > 0.8 with p  < 0.001), while weak or no correlations were observed for the ArcCHECK measurement. Between the MapCHECK2 and ArcCHECK measurements, the global GPRs showed no correlations (all with p  > 0.05), while the local GPRs showed moderate correlations only with 2%/1 mm and 1%/1 mm for TrueBeam STx ( r  > 0.5 with p  < 0.001). Both the global and local GPRs always showed weak or no correlations with the MLC positional errors except for the GPRs of MapCHECK2 with 1%/2 mm and 1%/1 mm for TrueBeam STx and the GPR of ArcCHECK with 1%/2 mm for Trilogy ( r  < − 0.5 with p  < 0.001). Conclusions The GPRs varied according to the types of gamma analyses, dosimeters, and linacs. Therefore, each institution should carefully establish their own gamma analysis protocol by determining the type of gamma index analysis and the gamma criterion with their own linac and their own dosimeter.

Purpose This study investigates the feasibility of utilizing deep learning models to robustly predict patient‐specific quality assurance (PSQA) outcomes in fixed field intensity‐modulated radiation therapy (FF‐IMRT) plans on the Halcyon linear accelerator equipped with a novel dual‐layer multi‐leaf collimator (MLC). The study explores the integration of Shuffle Attention (SA) mechanisms and deep imbalance regression techniques to enhance the precision and robustness of deep learning‐based PSQA predictions. It ensures relative prediction robustness in the extreme imbalance distribution of gamma passing rate (GPR) values. Methods Data from 214 FF‐IMRT treatment plans covering various treatment sites comprising 1394 beam orientations and corresponding Portal Dosimetry verification data were collected. Fluence maps calculated for each beam orientation served as inputs for the ResNet model. First, the SA module was introduced to improve the prediction accuracy of ResNet, resulting in the proposed Att‐ResNet model. Furthermore, to ensure prediction robustness in the GPR values with extreme imbalance distribution, we incorporated the Label Distribution Smoothing (LDS) technique, ultimately forming the ALDS‐ResNet method. Results ALDS‐ResNet exhibited smaller mean absolute error (MAE) values than ResNet across all gamma criteria (1%/1 mm: 2.035 vs. 1.824, 2%/2 mm: 1.416 vs. 1.178, 3%/3 mm: 0.951 vs. 0.787). ALDS‐ResNet also demonstrated lower MAE values than ResNet for complex but important plan samples (GPR < 85, 1%/1 mm: 10.163 vs. 4.985, 2%/2 mm: 7.443 vs. 3.272, 3%/3 mm: 5.031 vs. 2.940). Compared to ResNet, ALDS‐ResNet achieved higher Pearson correlation coefficient (CC) values at 2%/2 mm and 3%/3 mm gamma criteria, measuring 0.7864 and 0.7852, respectively. Conclusions The deep learning model based on ResNet shows promise for predicting GPR values in linacs with dual‐layer MLC. Integrating attention mechanisms with deep learning networks enhances the accuracy of PSQA predictions. The LDS technique is attributed to the substantial improvement in failed plan GPR prediction accuracy and robustness. Specifically, the deep learning model tailored for dual‐layer MLC linacs can be an auxiliary tool for physicists in identifying PSQA failure plans.

Purpose This study aims to investigate the influence of the magnetic field on treatment plan quality using typical phantom test cases, which encompass a circle target test case, AAPM TG119 test cases (prostate, head-and-neck, C-shape, multi-target test cases), and a lung test case. Materials and Methods For the typical phantom test cases, two plans were formulated. The first plan underwent optimization in the presence of a 1.5 Tesla magnetic field (1.5 T plan). The second plan was re-optimized without a magnetic field (0 T plan), utilizing the same optimization conditions as the first plan. The two plans were compared based on various parameters, including con­formity index (CI), homogeneity index (HI), fit index (FI) and dose coverage of the planning target volume (PTV), dose delivered to organs at risk (OARs) and normal tissue (NT), monitor unit (MU). A plan-quality metric (PQM) scoring procedure was employed. For the 1.5 T plans, dose verifications were performed using an MR-compatible ArcCHECK phantom. Results A smaller dose influence of the magnetic field was found for the circle target, prostate, head-and-neck, and C-shape test cases, compared with the multi-target and lung test cases. In the multi-target test case, the significant dose influence was on the inferior PTV, followed by the superior PTV. There was a relatively large dose influence on the PTV and OARs for lung test case. No statistically significant differences in PQM and MUs were observed. For the 1.5 T plans, gamma passing rates were all higher than 95% with criteria of 2 mm/3% and 2 mm/2%. Conclusion The presence of a 1.5 T magnetic field had a relatively large impact on dose parameters in the multi-target and lung test cases compared with other test cases. However, there were no significant influences on the plan-quality metric, MU and dose accuracy for all test cases.