RESUMO
PURPOSE: Quality assurance for stereotactic body radiation treatment requires that isocentric verification be ensured during gantry rotation at various angles. This study examined statistical parameters on Winston-Lutz tests to distinguish the deviation of angles from isocenter during gantry rotation using machine learning. METHOD: The Varian TrueBeam linac was aligned with the marked lines on the Ruby phantom. Eight images were captured while the gantry was rotating at a 45° shift. The statistical features were derived from IsoCheck EPID software. The decision tree model was applied to these Winston-Lutz tests to cluster data into two groups: precise and error angles. RESULTS: At 90° and 270° angles, the gantry exhibits isocentric stability compared to other angles. In these angles, the most statistical features were inside the range. Most variations were observed at 0° and 180° angles. In most tests, the angles 45°, 135°, 225°, and 315° showed reasonable performance and with less variation. CONCLUSION: The comprehensive statistical analyses for gantry rotation of angles assists expert radiotherapists in determining the contribution of each feature that highly affects gantry movement at specific angles. Misalignment between radiation isocenter and imaging isocenter, tuning of the beam at each angle, or a slight change in the position of the Ruby phantom can further improve the inaccuracy that causes the most variations. Better precision can effectively increase patient safety and quality during cancer treatment.
RESUMO
This study aims to predict isocentric stability for stereotactic body radiation therapy (SBRT) treatments using machine learning (ML), covers the challenges of manual assessment and computational time for quality assurance (QA), and supports medical physicists to enhance accuracy. The isocentric parameters for collimator (C), gantry (G), and table (T) tests were conducted with the RUBY phantom during QA using TrueBeam linac for SBRT. This analysis combined statistical features from the IsoCheck EPID software. Five ML models, including logistic regression (LR), decision tree (DT), random forest (RF), naive Bayes (NB), and support vector machines (SVM), were used to predict the outcome of the QA procedure. 247 Winston-Lutz (WL) tests were collected from 2020 to 2022. In our study, both DT and RF achieved the highest score on test accuracy (Acc. test) ranging from 93.5% to 99.4%, and area under curve (AUC) values from 90 to 100% on three modes (C, G, and T). The precision, recall, and F1 scores indicate the DT model consistently outperforms other ML models in predicting isocenter stability deviation in QA. The QA assessment using ML models can assist error prediction early to avoid potential harm during SBRT and ensure safe and effective patient treatments.