Independent validation of machine performance check for the Halcyon and TrueBeam linacs for daily quality assurance

Abstract Purpose To evaluate the ability of the machine performance check (MPC) on the Halcyon to detect errors, with comparison with the TrueBeam. Methods MPC is an automated set of quality assurance (QA) tests that use a phantom placed on the couch and the linac's imaging system(s) to verify the beam constancy and mechanical performance of the Halcyon and TrueBeam linacs. In order to evaluate the beam constancy tests, we inserted solid water slabs between the beam source and the megavoltage imager to simulate changes in beam output, flatness, and symmetry. The MPC results were compared with measurements, using two‐dimensional array under the same conditions. We then studied the accuracy of MPC geometric tests. The accuracies of the relative gantry offset and couch shift tests were evaluated by intentionally inserting phantom shifts, using a rotating or linear motion stage. The MLC offset and absolute gantry offset tests were assessed by miscalibrating these motions on a Halcyon linac. Results For the Halcyon system, the average difference in the measured beam output between the IC Profiler and MPC, after intentional changes, was 1.3 ± 0.5% (for changes ≤5%). For Halcyon, the MPC test failed (i.e., prevented treatment) when the beam symmetry change was over 1.9%. The accuracy of the MLC offset test was within 0.05 mm. The absolute gantry offset test was able to detect an offset as small as 0.02°. The accuracy of the absolute couch shift test was 0.03 mm. The accuracy of relative couch shift test of Halcyon was measured as 0.16 mm. Conclusion We intentionally inserted errors to evaluate the ability of the MPC to identify errors in dosimetric and geometric parameters. These results showed that the MPC is sufficiently accurate to be effectively used for daily QA of the Halcyon and TrueBeam treatment devices.


| INTRODUCTION
Daily quality assurance (QA) of medical linear accelerators (linacs) is standard practice in radiotherapy clinics. These QA tests include checks of x-ray output, energy, and beam profile constancy as well as image-guided radiation therapy (IGRT) functionality as recommended by the American Association of Physicist in Medicine (AAPM) Task Group (TG) 142, 1 MPPG5a, 2 and TG-179. 3 In general, several devices are used each morning to perform this QA. Beam measurements are usually done with a dedicated beam check device containing multiple diodes and/or ion chambers. IGRT imaging checks are generally done with an imaging phantom that checks the alignment between treatment and imaging geometries as well as the accuracy of IGRT-guided couch shifts. This process can be time-consuming, adding 15 min (or more) to morning QA, and can have a high level of user dependency, especially the imaging tests.
Varian Medical Systems (Palo Alto, CA, USA) recently released the Machine Performance Check (MPC) system, which is a fully integrated self-check tool for assessing the performance of the linac.
MPC uses a vendor-supplied phantom which is placed on the couch at the H2 position, using either a separate bracket or a bracket that is already attached to the phantom (depending on the system).
Images of the phantom are acquired with various combinations of collimator positions, gantry angles, and couch positions through an automated sequence. Beam performance checks include beam output constancy, beam uniformity, and beam center shift. Geometry checks include the radiation isocenter size and a measure of coincidence with MV and kV imaging isocenters, as well as checking collimator and gantry readout accuracy, MLC and jaw positioning, and couch positioning accuracy. The MPC tests are executed as a special mode on the linac. All motions (gantry, collimator, and couch) are automatic, as is the analysis of the images. The MPC results are then presented relative to a baseline for the beam performance checks, and compared to absolute specifications for the geometric tests.
Users are not able to modify the MPC results or thresholds but can set beam baselines. MPC is available for both the new Halcyon (Varian Medical Systems) and the existing TrueBeam platforms (version 2.0 and newer). The tests on these two treatment delivery systems are similar, although the vendor-set thresholds vary between the systems. Other differences include the following: • The Halycon has a "virtual isocenter" where the patient (phantom) is setup outside of the treatment bore, and then moved to the true isocenter by couch motion. This movement is checked by the MPC for the Halcyon.
• TrueBeam systems include kilovoltage imaging, which was not available on the preclinical Halcyon 1.0 unit tested in this work.
• The Halcyon does not allow couch rotations. • The MPC is part of a built in safety mechanism on the Halycon.
On these, unit beams cannot be run in clinical mode unless the MPC has been run and passed on to that calendar day.
• The TrueBeam uses the isocal phantom, also used for imaging geometry calibrations, with a separate couch mount while the Halycon uses a similar phantom that is permanently attached to a couch mount.

| METHODS
Here, we summarize each specific MPC test, and then describe how we intentionally inserted errors and evaluated the ability of the MPC to detect these errors. Specific details on the MPC system can be found in the publications by Clivio and Barnes and in the vendor documentation.
To assess the ability of the MPC to detect changes in machine output or beam profile, we simulated errors by fully or partially inserting solid water into the beam path to change the linac's apparent output and symmetry. The MPC results were compared to measurements made with a 2D ion-chamber array (IC profiler, Sun Nuclear, Melbourne FL). Other groups have previously benchmarked this device for a variety of beam measurement, including profile measurements, symmetry, and flatness measurements, and relative output measurements. 8  %Uniformity Change ¼ 100 Â maxðRatioðx; yÞÞ À minðRatioðx; yÞÞ ð Þ (1) Again this is calculated for the inner region of the image. It represents the worst-case. The threshold of uniformity change is AE2% for both TrueBeam and Halcyon.

2.A.2 | Error-detection tests
To evaluate the sensitivity of the MPC beam output and uniformity constancy measurements, we inserted solid water slabs between the beam source and the MV imager to introduce changes in beam output or flatness and symmetry of the beam (Fig. 1)

2.B.2 | Error-detection tests
To assess the accuracy of the MPC MLC positioning test for the  motor/linear drive system has a specification accuracy of 0.08 mm.

2.C.2 | Couch error detection tests
The accuracy was confirmed to be <0.5 mm visually over a 10 cm travel, using a ruler. In order to evaluate the impact of using the stage on the results, the 0 mm shift results were compared with the results when the MPC phantom was setup, using the usual approach (with the bracket at H2). For the TrueBeam, the vertical couch shifts were not examined because the clearance between the linac and the linear motor stand was insufficient when the motor stand was mounted in a vertical orientation.
The movement of the couch to the actual isocenter was evaluated in the same way, except that no images were taken with the phantom at the first position (i.e., at the virtual isocenter outside the bore), so the phantom was shifted prior to initiating the MPC sequence.

| RESULTS
The results of the error detection tests are summarized in Table 2.
To evaluate the reproducibility of the MPC tests, the process was repeated three times, and the range in the reported values calculated. The ranges were 0.44% and 0.15% for the reported changes in output and uniformity, respectively, 0.04°for gantry measurements, and 0.05 mm for couch shifts in any direction.

3.A | Beam output and uniformity
We compared the change in beam output reported by the MPC and

3.B | MLC position
The differences between the MLC offset measured by MPC and the intentionally miscalibrated values are listed in

3.C | Couch translation
In order to assess any impact of the translational stages on the experimental setup, MPC was performed with the phantom attached to the stage, but without any induced shifts. The results were within the ranges found when MPC was repeated multiple times (reported

3.D | Gantry rotation
The differences between the absolute gantry angle measured by the Halcyon MPC and the known miscalibrated values are also listed in

4.A | Beam output and uniformity
These results indicate that the MPC can detect changes in beam output and uniformity with sufficient accuracy/precision for daily QA.
One important weakness in our study is in the way in which we changed the beam output. We inserted solid water into the beam, which has the additional impact of hardening the radiation beamthus, the changes detected by the MPC test are likely and also partially due to the overresponse of the portal imager to the low energy component of the beam, resulting in larger differences than may be found without this effect. Barnes and Greer tested the ability of the MPC to accurately detect output by intentionally adjusting the linac output (True-Beam) and found that MPC output agreed with ion chamber to within 0.17%, 5 which is better agreement than found in our study. Clivio also found better agreement between MPC and ion chamber. 4 Our approach was to try to change the output (and uniformity, and other parameters) using external means. Thus, when both results are considered together (internal and internal adjustments), this adds confidence to our conclusion that the MPC can detect changes in beam output and uniformity. Furthermore, although we did not report on the day-to-day stability of these measurements, other authors 4,5 have reported that the MPC beam output measurements accurately tracked other independent measurements of beam output and flatness and symmetry over an extended period of time. Therefore, the MPC appears to provide an adequate measurement of the beam output and quality for daily QA purposes. Although the MPC may also be appropriate for monthly beam constancy checks, we do not believe that this has been proven with a sufficient degree of confidence, so monthly QA with independent equipment is still warranted.

4.B | MLC position
The accuracy of the MPC MLC position test was well within the 1-2 mm tolerance suggested by TG142 for daily checks of the collimator size indicator (All beam collimation on the Halcyon is performed, using the MLCs), indicating that the MPC is sufficient for daily QA.

4.C | Mechanical checks (Couch translation and gantry rotation)
The largest disagreement between the couch translation (or absolute position) and the intentionally inserted error was less than 0.5 mm, well within the 1 to 2 mm tolerances suggested in TG-179 for couch shift accuracy (monthly and daily tolerances, respectively). These results support the use of the MPC for checking these parameters.
Although we do not yet have any data on the expected failure modes of the gantry rotation (e.g., is it even possible for the system to be incorrect for a small arc?), given that the tolerance for gantry angle indicators suggested by the AAPM TG 142 is 1.0°, the results presented here indicate that (with the current vendor threshold), the use of the MPC is appropriate for testing gantry angle errors. These results, therefore, support the use of the MPC checks to replace some of the mechanical checks that are often performed on a daily basis. They may also replace some of the mechanical checks that are performed monthly.
There are some limitations of our study. We did not perform any tests of the coincidence of imaging and treatment isocenters. Our early version of the Halcyon did not include kilovoltage imaging, so this will have to be tested once released. Our version of the MLC was also a preclinical version (including the lower bank being slaved to the upper bank, as mentioned above). Our MLC tests were also limited to changing the calibration for a single MLC leaf, although this limitation is somewhat mitigated because earlier authors have investigated this issue (Barnes). Finally, we did not assess changes in the MPC over time. We do not expect this to be a concern for mechanical tests, and other authors have followed the MPC output results over relatively extended periods.
In summary, our tests have shown that the MPC can detect errors in the beam constancy and geometric parameters with an accuracy that is appropriate for use as daily QA. In many cases, the MPC also has the potential to replace monthly QA checks, but additional work (especially regarding constancy) is needed before a solid conclusion can be given regarding the use of MPC for monthly QA.

| CONCLUSION
We showed that the MPC of both Halcyon and TrueBeam linacs can detect errors in the beam constancy and geometric parameters with an accuracy that means the MPC is appropriate for use as daily QA.

ACKNOWLEDG MENTS
We thank the following for their support during this work: Pekka