O-ring linac systems improve radiotherapy efficiency but require rigorous pretreatment verification due to increased delivery uncertainty in IMRT/VMAT. Existing methods face limitations: measurement-based approaches incur setup errors, while calculation-based methods (e.g., Monte Carlo) need machine-specific validation per AAPM TG-219.
To commission the O-ring linac model in RadCalc Monte Carlo software and establish its clinical dosimetric accuracy.
The additional radiation to light field offset (ARLFO) parameter in RadCalc was adjusted from −0.08 cm (−0.8 mm) to +0.08 cm (+0.8 mm) (nine sets). TG-119 and clinical benchmark cases were used to design IMRT/VMAT plans. Verification plans were measured experimentally, and all plans were imported into RadCalc for secondary dose calculation. Triangular gamma analysis (3%/2 mm) compared Monte Carlo-simulated, measured, and TPS-calculated doses. The optimal ARLFO was determined by weighted averaging of gamma pass rates. The model was validated on 10 clinical cases per site (head, thorax, abdomen, pelvis).
Commissioning identified the optimal ARLFO parameter as −0.02 cm (−0.02 mm). Under the gamma analysis criterion of 3%/2 mm, the comparison between TPS-calculated doses and Monte Carlo-calculated doses for 10 clinical plans across four anatomical sites yielded: 98.9% ± 0.8% (head and neck), 98.5% ± 0.8% (thorax), 99.6% ± 0.3% (abdomen), and 99.1% ± 0.5% (pelvis). For verification plans, the gamma pass rates between Monte Carlo-calculated doses and measured doses were 97.5% ± 2.8% (head and neck), 95.6% ± 2.8% (thorax), 96.3% ± 2.9% (abdomen), and 99.4% ± 0.5% (pelvis), while comparisons of Monte Carlo-calculated doses versus TPS-calculated doses reached 99.5% ± 0.8% (head and neck), 99.2% ± 1.0% (thorax), 99.5% ± 0.9% (abdomen), and 99.9% ± 0.2% (pelvis), demonstrating consistent dosimetric accuracy of the optimized model across all clinical sites.
This study establishes a commissioning methodology to determine the optimal ARLFO value for RadCalc, enabling clinics to achieve reliable independent plan verification for O-ring linac.
This study aims to enhance the preoperative diagnosis of non-mass breast lesions (NMLs) by validating radiomics-based machine learning models and assessing their performance alone and in combination with clinical ultrasound features to distinguish benign from malignant lesions.
A total of 123 NMLs from 119 patients with confirmed pathology were analyzed. Patients were split into a training cohort (n = 98) and a validation cohort (n = 25). From each ultrasound image, 1558 radiomics features were extracted. After dimensionality reduction and feature selection, 10 key features were retained. Predictive models were developed using logistic regression (LR), linear regression, support vector machine (SVM), random forests, Extremely Randomized Trees (Extra Trees), and Light Gradient Boosting Machine (LightGBM). A clinical model was built using LR based on ultrasound findings such as calcification, high resistance index, and axillary lymph node enlargement. A combined model incorporated both radiomics and clinical features. Model performance was evaluated using receiver operating characteristic (ROC) curves and decision curve analysis (DCA).
The LightGBM model achieved the highest radiomics-only performance (AUC: 0.932 training; 0.867 validation). The clinical model achieved AUCs of 0.837 (training) and 0.790 (validation). The combined model outperformed both, with AUCs of 0.973 (training) and 0.933 (validation), and showed superior clinical benefit in DCA.
Combining radiomics with clinical ultrasound data significantly improves diagnostic accuracy for NMLs, supporting better differentiation between benign and malignant lesions and aiding clinical decision-making.
To report and investigate a case of significant dose attenuation observed during quality assurance (QA) of an Elekta Unity MR-Linac system, and to elucidate its root cause and solution.
During a period of abnormal environmental conditions (elevated temperature and humidity), a >20% dose output reduction was detected at a gantry angle of 180°. A comprehensive investigation was undertaken, including dose output measurements at multiple gantry angles, beam quality assessment, mechanical accuracy checks, EPID image analysis, and gamma pass rate analysis using ArcCheck-MR. The source of the attenuation was localized through a process of elimination.
The investigation revealed severe dose attenuation specifically at angles traversing beneath the treatment couch, with a gamma pass rate (3%/3 mm) dropping to 39.1% at 180°. All other system parameters, including MRI quality and mechanical alignments, were within tolerance. Visual inspection after disassembly confirmed the presence of a substantial pool of water beneath the table, estimated to be at least 3.5 cm deep, which acted as an unintended attenuator. The water was attributed to condensation and drainage issues exacerbated by the failed air-conditioning system. Mitigation strategies, including hardware sealing, enhanced drainage, revised maintenance protocols, and the addition of a 180° dose check to daily QA, successfully resolved the issue.
This case highlights a previously unreported and critical failure mode for MR-Linac systems. It underscores the profound impact of environmental control on machine performance and patient safety. Robust HVAC maintenance, proactive monitoring of sub-table areas, and the inclusion of oblique-angle dosimetry in QA routines are essential for all Elekta Unity and similar high-precision radiotherapy systems.
Koufigar S, Ford E, He Y, Olsen S, Fagerstrom JM. A case study on SSD to SAD linear acceleartor calibration transition. J Appl Clin Med Phys. 2025;26:e70298.
Description of error:
There is a spelling error in the article title. The word “accelerator” is incorrectly written as “acceleartor.” The correct title is “A case study on SSD to SAD linear accelerator calibration transition.” The online version of the article has been corrected accordingly.
We apologize for this error.
To determine the gamma criteria that maximize the sensitivity and specificity of the transit dosimetry software PerFRACTION (Sun Nuclear Corporation) under five possible failure modes in external beam radiotherapy.
We simulated five failure modes that potentially introduce large changes in the absorbed dose distribution: (1) Linac hardware incidents. In a VMAT head and neck treatment plan, erroneous plans were created, introducing known errors in the MLC aperture, the collimator angle, and the monitor units. (2) Breathing management protocol incidents. Based on a 4D-CT of a lung treatment, we created expiration and inspiration CT images. A treatment plan was created using each CT and recalculated in the alternate CT. (3) Patient identification incidents. The treatment plan of one breast patient was recalculated on another patient, and vice versa. (4) Selection of the planning CT incidents. A lung patient had two planning CTs with the presence/absence of lung fluid. A treatment plan was generated for each CT and recalculated for the other. (5) Bolus positioning incidents. An erroneous treatment plan for a breast plan was created without the bolus. For the five failure modes, the transit images were compared using seven gamma criteria, both global and local. Receiver-operating characteristic (ROC) curves were generated based on the change in the PTV mean dose: > 5%, > 10%, and > 20%. The area under the curve (AUC) and optimal passing rate were calculated.
The global gamma criterion γ(10%/1 mm) maximizes PerFRACTION sensitivity and specificity to detect failure modes that introduce a change in the PTV mean dose exceeding 10%. The standard global gamma criterion, as γ(3%/3 mm), maximizes PerFRACTION sensitivity and specificity to detect deviations in the PTV mean dose above 5%.
A 10%, 1 mm global gamma parameter value produces the needed sensitivity and specificity to identify erroneous deliveries at a level of dose differences of 10% or greater.
To develop a knowledge-based deep model for sCT generation from a single MR volume in LINAC-based frameless SRS, enabling the MR-only workflow without extra CT simulation.
A total of 139 patients were included in the study, with 120 used for training and 19 for testing. A Deep Residual U-Net(DRU) was developed to generate sCT from patient-specific high-resolution T1 + Contrast MR volume, complemented by a healthy brain CT volume from the Visible Human Project that provides CT-specific anatomical knowledge. To simulate treatment conditions, a template immobilization mask was deformed to align with the patient-specific sCT anatomy, thereby creating a full sCTF volume. Four metrics, including PSNR, SSIM, RMSE, and MAE were derived to evaluate Hounsfield units(HU) accuracy of sCT compared to the ground-truth CT without immobilization masks. Single isocenter multi-target SRS plans developed with volumetric modulated arc therapy (VMAT) technique were recalculated within sCTF volumes to produce simulated dose distributions, which were compared with clinical plan dose distributions using the mean dose difference in the planning target volume(PTV) and gamma index evaluation.
In the test set, the generated sCT (statistics are reported as mean ± standard deviation) achieved a PSNR(Peak Signal-to-Noise Ratio) of 75 ± 4 dB, Structural Similarity Index (SSIM) of 0.99 ± 0.01, root mean square error (RMSE) of 11.9 ± 5.8 HU, and mean average error (MAE) of 1.4 ± 0.8 HU for brain tissues. When comparing sCTF dose calculation results against the original plans, gamma index passing rates were 95.8 ± 4.2% for the entire volume and 84.4 ± 15.0% within PTVs, using 3%/1 mm/15% threshold criteria. The median/ interquartile range of PTV dose differences were –2.0% and 2.3%, with all discrepancies below –5.0%.
This study successfully demonstrated the generation and validation of sCT images from single-modality MRI using a knowledge-based deep model. The results confirm that single-modality MRI without simulation CT scan effectively supports frameless SRS planning and integrates seamlessly into current clinical workflows.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：