Patient‐specific quality assurance using machine log files analysis for stereotactic body radiation therapy (SBRT)

Abstract An in‐house trajectory log analysis program (LOGQA) was developed to evaluate the delivery accuracy of volumetric‐modulated arc therapy (VMAT) for stereotactic body radiation therapy (SBRT). Methods have been established in LOGQA to provide analysis on dose indices, gantry angles, and multi‐leaf collimator (MLC) positions. Between March 2019 and May 2020, 120 VMAT SBRT plans of various treatment sites using flattening filter‐free (FFF) mode were evaluated using both LOGQA and phantom measurements. Gantry angles, dose indices, and MLC positions were extracted from log and compared with each plan. Integrated transient fluence map (ITFM) was reconstructed from log to examine the deviation of delivered fluence against the planned one. Average correlation coefficient of dose index versus gantry angle and ITFM for all patients were 1.0000, indicating that the delivered beam parameters were in good agreement with planned values. Maximum deviation of gantry angles and monitor units (MU) of all patients were less than 0.2 degree and 0.03 % respectively. Regarding MLC positions, maximum and root‐mean‐square (RMS) deviations from planned values were less than 0.6 mm and 0.3 mm respectively, indicating that MLC positions during delivery followed planned values in precise manner. Results of LOGQA were consistent with measurement, where all gamma‐index passing rates were larger than 95 %, with 2 %/2 mm criteria. Three types of intentional errors were introduced to patient plan for software validation. LOGQA was found to recognize the introduced errors of MLC positions, gantry angles, and dose indices with magnitudes of 1 mm, 1 degree, and 5 %, respectively, which were masked in phantom measurement. LOGQA was demonstrated to have the potential to reduce or even replace patient‐specific QA measurements for SBRT plan delivery provided that the frequency and amount of measurement‐based machine‐specific QA can be increased to ensure the log files record real values of machine parameters.


| INTRODUCTION
Volumetric-modulated arc therapy (VMAT) for stereotactic body radiation therapy (SBRT) can deliver high doses to target volumes while sparing proximal organs at risk (OARs) by generating a rapid fall-off of dose outside the target in a hypofractionated regimen. 1 The steep dose gradients are enabled by simultaneous modulation of multileaf collimator (MLC) positions, gantry rotation speeds, and dose rates during single or multiple gantry rotations around a patient. 2 Conformal dose distribution can be generated by beam modulation in treatment planning system. However, extensive modulations utilized in SBRT can lead to deviation of delivered dose distribution from the planned one. This can be due to delivery system uncertainties such as MLC leaf position errors and gantry rotational instability. 1 Therefore, pretreatment quality assurance (QA) of modulated arc therapy is necessary for patient safety.
According to American Association of Physicists in Medicine (AAPM) Task Group No. 218, measurement-based patient-specific QA methods are widely used and are the core element of most QA programs. 3 A treatment plan with MLC leaf sequence file is computed on a homogeneous phantom with dosimeters to calculate the dose in QA geometry. The physical phantom is then irradiated to measure the actual dose distribution. Measured dose distribution can then be compared with the calculated dose distribution. 4 Although measurement-based patient-specific QA is commonly used clinically, previous studies have demonstrated the insensitivity of this method to the discrepancy of beam parameters such as gantry angle errors and MLC positioning errors during data transfer or machine delivery. 5 Therefore, analysis of machine log files has been suggested by several studies as an alternative. [6][7][8] It can identify the problems that are undetectable with measurement-based approach such as MLC positioning error. This is important as Budgell, et al. have demonstrated that accurate delivery of dose for IMRT fields require better than 1 mm leaf positioning accuracy. [8][9][10][11] It has been shown by Agnew, et al. that occasionally log file analysis might not be able to detect errors in MLC positions due to the failure of t-nut or motor. However, it could still be a powerful tool to ensure data transfer accuracy associated with MLC performance, provided that more frequent measurement-based machine-specific QA has been established. 12 Several studies have been performed to evaluate the efficiency and effectiveness of machine log file analysis for volumetric-modulated arc therapy (VMAT). 13,14 Chan et al. conducted a preliminary study on quantifying the deviations of monitor unit (MU) and gantry angles during delivery against the plan through log analysis for three head and neck and three prostate VMAT cases. 14 Mcgarry et al. quantified the plan delivery accuracy in terms of log file-derived MLC positioning error with single center-specific plan and single standard plan for each participating center in VMAT audits. 13 However, no comprehensive study with a large number of clinical cases has been conducted to evaluate simultaneously the deviations of MU, gantry angles, and MLC positions against the planned values using machine log files. The accuracy of all these beam parameters is essential to guarantee mechanical stability and dosimetrical accuracy.
Also, extensive study has not been performed to investigate the use of machine log analysis for VMAT SBRT plans using flattening filter-free (FFF) mode for various treatment sites. Since SBRT is indicated for localized small tumor, minor deviation introduced in MLCs might affect the dose distribution to a more significant extent compared with other radiotherapy treatment techniques. 4,15 This study is therefore aimed to demonstrate the effectiveness of an in-house developed trajectory log analysis program (LOGQA) for evaluating the delivery accuracy of VMAT SBRT plans comprehensively with the introduction of intentional errors for software validation.

2.A | Treatment plans
In this study, 120 VMAT SBRT were planned and delivered using the

2.B | Pretreatment QA measurements
Each VMAT SBRT treatment plan was calculated for the Arc-CHECK TM phantom (Sun Nuclear Corporation, Melbourne, FL, USA) using the Eclipse TM treatment planning system. [17][18][19] ArcCHECK TM was then irradiated to measure the dose distribution of each plan.
At the same time, machine log files were generated automatically.
Global gamma-index analyses with gamma criteria of 2% dose difference and 2 mm distance-to-agreement (DTA) were performed using the SNC software (Sun Nuclear Corporation, Melbourne, FL, USA). 20 The software compared the measured dose distributions with the planned dose distributions generated from the Eclipse TM system using a dose calculation grid size of 0.25 cm. Absolute doses were used and the points with doses less than 10% of the maximum doses were excluded in the gamma-index analysis. Passing rate of gamma-index analysis with 95% or above was recognized as a pass in verification based on our local practice.

2.C | Machine log files analysis
where proportional of time being shielded ¼ Fluence at any pixel in zone3 ¼ fractional MU between2 CPs Five graphs (Fig. 2-6) were plotted by LOGQA to demonstrate the accuracy of delivery for each plan, with quantitative indicators to define whether the comparison was passed or not (Table 1). Correlation coefficient is calculated to evaluate if the delivery is in good agreement with the plan.   ii Gantry angle errors of 1 degree and 2 degree for all the control points.
iii Differential dose index errors of 5% and 10% for 60 control points.
Gamma criteria of 2% dose difference and 2 mm distance-toagreement (DTA) were used to compare the original plans and the log files of the plans with errors.

3.A | Evaluation of machine log files
Between March 2019 and May 2020, 120 VMAT SBRT treatment plans were evaluated using LOGQA and all the parameters being checked were within our tolerance listed in Table 1 Table 2.

3.B | Gamma-index analysis of pretreatment measurements
Absolute gamma-index analyses from ArcCheck TM measurements were performed for each VMAT SBRT plan. The passing rate of all plans was larger than 95% with 2% dose difference and 2 mm spatial acceptance criteria, which would be recognized as a pass in verification based on our local practice.

| DISCUSSION
Many potential errors can arise during treatment planning and delivery, such as inaccurate dose calculation and errors in plan transfer and delivery. 21 There is a growing trend of performing machine log analysis for patient-specific QA as it is more sensitive to identify mechanical errors of the order of 1 mm and 1 degree. 22,23 Also, machine log analysis can catch the discrepancies related to plan transfer and delivery problems. 24,25 In present study, 120 VMAT

| CONCLUSION
Machine log analysis provides crucial information on VMAT SBRT plan delivery. End-to-end plan transfer and beam parameters accuracy check using LOGQA provides a robust and reliable QA method to reduce patient-specific QA measurement, provided that a comprehensive measurement-based machine-specific QA program is in place. Good agreement of the trajectory plots and ITFMs between the delivery and the plan indicates precise modulation of dose rate, gantry rotational speed and MLC leaf speed, guaranteeing reliable mechanical stability and dosimetrical accuracy.

CONFLI CT OF INTEREST
No conflict of interest.