Use of a plastic scintillator detector for patient‐specific quality assurance of VMAT SRS

Abstract Purpose To evaluate a scintillator detector for patient‐specific quality assurance of VMAT radiosurgery plans. Methods The detector was comprised of a 1 mm diameter, 1 mm high scintillator coupled to an acrylic optical fiber. Sixty VMAT SRS plans for treatment of single targets having sizes ranging from 3 mm to 30.2 mm equivalent diameter (median 16.3 mm) were selected. The plans were delivered to a 20 cm × 20 cm x 15 cm water equivalent plastic phantom having either the scintillator detector or radiochromic film at the center. Calibration films were obtained for each measurement session. The films were scanned and converted to dose using a 3‐channel technique. Results The mean difference between scintillator and film was ‒0.45% (95% confidence interval ‒0.1% to 0.8%). For target equivalent diameter smaller than the median, the mean difference was 1.1% (95% confidence interval 0.5% to 1.7%). For targets larger than the median, the mean difference was ‒0.2% (95% confidence interval ‒0.7% to 0.1%). Conclusions The scintillator detector response is independent of target size for targets as small as 3 mm and is well‐suited for patient‐specific quality assurance of VMAT SRS plans. Further work is needed to evaluate the accuracy for VMAT plans that treat multiple targets using a single isocenter.

response dependent on the field size. In the well-defined geometry used for output factor measurement, the field size dependence can be corrected using factors determined by Monte Carlo calculations. 7 However, for patient-specific quality assurance (QA), the geometry and field size are not sufficiently well-defined to apply a correction factor. Radiochromic film (RCF) can be used for patient-specific QA, but it requires careful calibration and is labor intensive.
Lack of water equivalence is the fundamental source of field size dependence. Miniature plastic scintillators have been developed that are nearly water equivalent. 8,9 A commercial version of the detector described by Beddar et al., having dimensions 1 mm diameter and 3 mm long, has been investigated for measurements of depth dose and profiles of small fields, 10 determination of field size correction factors for other detector types, 11 small-field dose measurements in heterogeneous media, 12 and for patient-specific QA of small-field SRS plans. 13 Recently, a second generation has been developed that is smaller, 1 mm instead of 3 mm length. The near water equivalence coupled with small dimensions make this detector a promising candidate for point dose measurements of SRS plans, particularly those for which the dose distribution has significant dose gradient over the distance of 3 mm.

| MATERIALS AND METHODS
The active scintillator volume of the prototype detector (model W2, Standard Imaging, Madison, WI) was a cylinder of 1 mm diameter and 1 mm height. The scintillator was bonded to a 1 mm diameter polymethyl methacrylate (PMMA) optical fiber that was approximately 1 m long. The light output of the fiber was split between two photodetectors. One detector had an optical filter designed to transmit wavelengths in the range of the scintillation spectrum. This is referred to as the blue channel. The second detector collected the signal from the longer wavelengths, and is referred to as the green channel. The spectra of the scintillator and Cerenkov radiation are shown schematically in Fig. 1. The Cerenkov radiation has a broad spectrum that overlaps the scintillation spectrum and so the signal from the blue channel (s Blue ) results from both scintillation and Cerenkov radiation, whereas Cerenkov radiation is the primary source of signal from the green channel (s Green ). The portion of s Blue due to Cerenkov radiation is proportional to s Green . The scintillator output is proportional to the dose deposited in the scintillator volume. The dose may therefore be determined from the signals by 14,15 Rearranging, The Cerenkov light radiation correction factor k CLR can be determined by irradiating the detector using two field geometries that result in the same dose at the detector position but irradiate different lengths of the fiber. The correction factor is then given by where max and min refer to the signals for maximum and minimum volume of fiber irradiated, respectively. Once k CLR is known, k gain can be determined by irradiating the detector to a known dose: Alternatively, the Cerenkov light radiation correction factor k CLR and signal to dose conversion factor k gain can be determined by irradiating the detector to a known constant dose for range of field sizes and fitting a line to Eq. (2). Substituting into Eq. (1), the dose is given by A Bland-Altman plot of the W2 and radiochromic film measurement-to-plan ratios for the 60 patient plans is shown in Fig. 4    Mean difference (%) F I G . 7. Mean difference between the W2 and radiochromic film as a function of fixed k CLR and k gain calculated using the sessionspecific reference field. The shaded region shows the standard deviation.
The repeated measurements are summarized in Table 1. It was noted that the dose distributions were not well centered at isocenter and that the measurement location could have a dose gradient. The magnitude of the dose gradient at the measurement location is also given in Table 1. The standard deviation of the repeated measurements given in Table 1 demonstrates that the uncertainty of the W2 < 0.5% in regions of low-dose gradient. However, in high-dose gradient locations, the uncertainty will be larger due to positioning uncertainty. In the work reported here, the detector was placed at isocenter, which was not always at the low gradient center of the dose distribution.

| DISCUSSION
Accuracy of the measurement would likely be improved if the measurement location was shifted to the position of the maximum dose.
It is important to note that at this location, a detector positioning error will always result in underreporting of the dose.

| CONCLUSION
The scintillator detector is well-suited for patient-specific quality assurance of VMAT SRS plans. The detector response is nearly independent of target size for targets as small as 3 mm. Because the detector is near water equivalent, a dummy detector having a highdensity fiducial at the location of the scintillator is necessary to position the detector using a kilo-voltage image guidance system. The signal due to Cerenkov radiation generated in the optical fiber is similar to that generated by uniform fields smaller than 7 cm × 7 cm.
Further work is needed to evaluate the accuracy of the scintillator detector for multiple target, single isocenter SRS.

CONFLI CT OF INTEREST
The authors declare no conflict of interest.
T A B L E 1 Standard deviation of 10 repeated measurements and dose gradient at the measurement location.
Target equivalent diameter (mm)