Characterization of a 0.8 mm³ Medscint plastic scintillator detector system for small field dosimetry

The HYPERSCINT RP-200 (HS RP-200) dosimetry system consists of three parts (i) PSD probe, (ii) acquisition system, and (iii) software. The probe investigated in this work is a single-point, polyvinyltoluene detector, which has a sensitive volume of 0.8 mm³ given by its diameter of 1 mm and length of 1 mm. This probe is housed in a black nylon envelope and is optically coupled to a 20 m long 1 mm diameter PMMA optical fiber core clad in a polyethylene jacket. The fiber optical cable is attached to the HS RP-200 acquisition system with a proprietary connector. The signal is read out using Medscint's proprietary software. For the remainder of this paper, this particular set up consisting of the above mentioned probe, acquisition system, and software will be referred to as HS RP-200–1 × 1.

First, using a Varian TrueBeam^TM STx linear accelerator (linac) (Varian Medical Systems, Palo Alto, CA), a four-component calibration procedure specified by Medscint, including scintillation, fluorescence, Cherenkov, optical fiber spectral attenuation, and system dose calibration components, was performed. The scintillation and fluorescence components were obtained by performing a high intensity kV irradiation, in fluouroscopy mode, of only the probe end, followed by irradiation of only the optical fiber, for 60 s and 120 s respectively. The Cherenkov calibration and optical fiber spectral attenuation consisted of four components, and included irradiating two sections of optical fiber laying perpendicularly to the MV energy beam at isocenter with gantry angles of 45° and 315°. Irradiating different sections of fiber allows the HS RP-200 system to evaluate and subtract the Cherenkov component from the signal of the detector. Finally, the dose calibration was performed by irradiating the probe to a known dose. A repetitive integration time of 1 s was used for all measurements as recommended in the user manual.

The dosimetric performance tests were carried out under setup conditions with lights and monitor displays turned off to prevent signal contamination from ambient light (although the PSD probe is light-tight). Majority of irradiations were carried out with a Varian TrueBeam^TM STx linac equipped with an HD 120^TM MLC featuring leaves measuring 2.5 mm in width. These irradiations were delivered with either a 6 MV-WFF beam with a dose rate of 600 monitor units min⁻¹ (MU min⁻¹) or 10 MV-FFF beam with a dose rate of 2400 MU min⁻¹. Detector specific output ratio measurements were also conducted with Varian TrueBeam^TM linac equipped with Millennium MLC with a leaf width of 5 mm. In dose rate dependence measurements, the dose rate was varied from 40 to 2400 MU min⁻¹. The PSD probe was positioned in a solid water phantom (Gammex-RMI Ltd, Nottingham, UK, model 457–320) perpendicularly to the beam, at a depth of 10 cm of solid water serving as backscatter, and source-to-surface distance of 100 cm. The phantom had L × W × H dimensions of 30 cm × 30 cm × 20 cm. The measurement setup was applied to the subsequent subsections except for the study of angular dependence, where a cylindrical water phantom was used. The probe was connected to the HS RP-200 signal acquisition system, which was placed outside the linac bunker. The measurements were repeated on several separate occasions spanning over eight weeks to account for any potential setup and linac output variations.

2.2.1. Pre/post irradiation leakage

2.2.2. Short term repeatability

2.2.3. Dose–response linearity

2.2.4. Dose rate dependence

2.2.5. Field size dependence (output factors)

2.2.5.1. Field size specification

Following the IAEA TRS-483 guidelines (IAEA 2017), field size was defined as the full width at half maximum (FWHM) of in-plane and cross-plane profiles of the field at the position of the detector. Then, equivalent square field size S_clin was calculated as follows

$$\begin{eqnarray*}&&{S}_{\mathrm{clin}}=\sqrt{a\,\ast \,b},\end{eqnarray*}$$

where a is the length and b is the width of a rectangular field defined by the FWHM in-plane and cross-plane dosimetric profiles. Lateral beam profiles were measured using RTQA2 Gafchromic film (Ashland Advanced Materials, Bridgewater, NJ). The field size was MLC defined with jaws set a few millimeters outside MLC dimensions, as specified in table 1, to reduce jaw aperture uncertainties (Cranmer-Sargison et al 2011). Jaw and MLC settings for field sizes 1 × 1 cm² and larger reproduced settings by Bassinet et al (2013). In that study Bassinet et al, using a Clinac 2100, measured field output factors for active and passive detectors including shielded and unshielded diodes, LiF microcubes, films, ion chambers and microDiamond.

Table 1. Field sizes defined via nominal dimensions as given by combination of MLC and jaw positions, and equivalent field size as defined by FWHM of the in-plane and cross-plane profiles obtained from Gafchromic film.

MLC size (cm2)	Jaw size (cm2)	Sclin TrueBeamTM STx (cm2)	Sclin TrueBeamTM (cm2)
0.25 × 0.25	1 × 1	0.28	—
0.5 × 0.25	1 × 1	0.33	—
0.5 × 0.5	1 × 1	0.53	0.46
0.75 × 0.75	1 × 1	0.74	—
1 × 1	1.2 × 1.2	0.96	1.00
1.25 × 1.25	1.4 × 1.4	1.16	—
1.5 × 1.5	1.7 × 1.7	1.44	1.49
1.75 × 1.75	2 × 2	1.67	—
2 × 2	4.4 × 4.4	1.98	2.03
2.5 × 2.5	4.4 × 4.4	2.45	2.42
3 × 3	4.4 × 4.4	2.99	3.05
5 × 5	5 × 5	4.99	4.97
10 × 10	10 × 10	9.99	9.87

2.2.5.2. Output factors

IAEA TRS-483 (IAEA 2017) adopts the formalism by Alfonso et al (2008) and defines field output factors as the absorbed dose for a small (clinical) field, ${f}_{\mathrm{clin}},$ relative to the absorbed dose in a machine specific reference field, ${f}_{\mathrm{msr}},$ as

$$\begin{eqnarray*}&&{{\rm{\Omega }}}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}=\frac{{M}_{{Q}_{\mathrm{clin}}}^{{f}_{\mathrm{clin}}}}{{M}_{{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{msr}}}}\,\ast \,{k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}},\end{eqnarray*}$$

where ${M}_{{Q}_{\mathrm{clin}}}^{{f}_{\mathrm{clin}}}$ is the detector reading in the small field, ${M}_{{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{msr}}}$ is the detector reading in the machine specific reference field, and ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}$ is the small field output correction factor. The detector specific output ratio is defined as $\frac{{M}_{{Q}_{\mathrm{clin}}}^{{f}_{\mathrm{clin}}}}{{M}_{{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{msr}}}}.$

In this work detector specific output ratios were measured with the HS RP-200–1 × 1 system and then ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}$ factors were derived from comparison of these results to output factors measured with two PTW TN60019 microDiamond detectors. A brief comparison to literature is included where field output factors were published for the same fields.

2.2.5.3. PSD positioning

The effective point of measurement, as indicated on the probe at 2.5 mm from the tip, was placed at the center of the smallest linac-defined field (0.25 × 0.25 cm²) by means of in-plane and cross-plane profile measurements. Using the scanning approach, the detector was placed at the position which gave the maximum reading by shifting detector in lateral and superior–inferior directions.

2.2.5.4. Measuring output factors

Under reference conditions, the PSD was irradiated using Varian TrueBeam^TM STx linear accelerator to 100 MU per field size for field sizes of 0.25 × 0.25 cm² to 10 × 10 cm² as outlined in table 1. Additionally, for field sizes of 10 × 10 cm² down to 0.5 × 0.5 cm², as outlined in table 1, PSD was irradiated with 100 MU per field size with Varian TrueBeam^TM linac. Detector specific output ratios for a specific field size were defined as the ratio of the reading at that field size to the reading at a field size of 10 × 10 cm² (machine specific reference field). These output ratios were then compared with measurements performed using a PTW TN60019 microDiamond (PTW, Freiburg, Germany) corrected with small field correction factor following IAEA TRS-483 under the same irradiation conditions. The measured values were compared to published data.

2.2.6. Angular dependence

2.2.7. Estimated uncertainties

An average of three 10 min measurements under no irradiation of the PSD yielded a reading of 0.91 ± 0.3 cGy. Post irradiation readings indicated minimal leakage. The 5 s reading immediately post 500 MU irradiation (time = 0 s post irradiation) was 0.003 ± 0.005 cGy and the 5 s reading one minute post irradiation (time = 60 s post irradiation), was 0.005 ± 0.005 cGy, both well below the background reading. Comparing results on the same time scale of 5 s, both post-irradiation readings are comparable to the background reading of 0.008 ± 0.002 cGy/5 s.

The average of 10 readings taken for 200 MUs delivered per measurement in a 5 × 5 cm² field size was 113.33 cGy with a standard deviation (SD) of 0.04 ± 0.01%, where uncertainty is given by standard error (SE) defined by $\mathrm{SE}=\tfrac{\mathrm{SD}}{\sqrt{2\,\ast \,N-2}},$ where N is the number of samples .

Dose–response of this dosimetry system showed a linear relationship with an R² value of ~1 and standard error of fit 0.26 cGy (figure 1). However, for the lowest delivered dose of 5 MU, the deviation of the measured from the expected linear dose response was 5.9%. Otherwise, differences fell below ±1% for doses higher than 25 MU. At the measurement point in this setup 1 MU = 0.574 cGy.

Zoom In Zoom Out Reset image size

The HS RP-200–1 × 1 showed minimal dose rate dependence between 200 and 2400 MU min⁻¹. The PSD measured constant dose with largest discrepancy of 0.48% at dose rate of 100 MU min⁻¹, otherwise discrepancy was within 0.10% (figure 2). For dose rates below 100 MU min⁻¹, the discrepancy increased up to 1.56% at the lowest dose rate measured (40 MU min⁻¹).

Zoom In Zoom Out Reset image size

3.4.1. 6 MV-WFF beams

Figure 3 displays detector specific output ratios measured with HS RP-200–1 × 1 for 6 MV-WFF energy beam delivered with Varian Truebeam^TM STx for field sizes of 0.25 × 0.25 cm² to 10 × 10 cm². This figure also includes corrected field output factors measured with two PTW TN60019 microDiamond detectors and data from Lechner et al (2022) for field sizes of 0.5 × 0.5 cm² to 10 × 10 cm². For field sizes of 1 × 1 cm² and larger, the agreement between HS RP-200–1 × 1 and PTW TN60019 microDiamond detectors with appropriate corrections was within measurement uncertainty. For field size of 0.5 × 0.5 cm², the detector specific output ratio measured with HS RP-200–1 × 1 was 4.3% lower than corrected field output factor measured with PTW TN60019 (#1) microDiamond detector and 3.7% lower than that measured with PTW TN60019 (#2). Field output factors for field sizes smaller than 0.5 × 0.5 cm² were not measured with microDiamond because published small field correction factors for these field sizes have large reported uncertainties (Alhakeem and Zavgorodni 2018).

Zoom In Zoom Out Reset image size

The detector specific output ratios measured with HS RP-200–1 × 1 and corrected field output factors measured with two PTW TN60019 microDiamond detectors for Varian TrueBeam^TM 6 MV-WFF energy beam are displayed in figure 4. Again, agreement was within measurement uncertainty for field sizes of 1 × 1 cm² and larger. Largest difference was seen at the smallest field size of 0.5 × 0.5 cm² of 2.9% and 4.9% between HS RP-200–1 × 1 and PTW TN60019 (#1) and (#2), respectively.

Zoom In Zoom Out Reset image size

Comparison of detector specific output ratios for two Varian TrueBeam^TM models showed little difference. Detector specific output ratios measured with HS RP-200–1 × 1 on Varian TrueBeam^TM and Varian TrueBeam^TM STx agreed to within 3% for field sizes down to 0.5 × 0.5 cm². For the field sizes down to 0.6 × 0.6 cm², comparison of these output ratios to the output factors reported by Lechner et al (2022) for Varian TrueBeam^TM STx showed agreement within 1%. However, comparison to Lechner data for Varian TrueBeam^TM showed a difference of about 45% for the smallest field size of 0.5 × 0.5 cm².

3.4.2. 10 MV-FFF beams

Figure 5 displays detector specific output ratios measured with HS RP-200–1 × 1 for 10 MV-FFF energy beam delivered with Varian Truebeam^TM STx for field sizes of 0.25 × 0.25 cm² to 10 × 10 cm². This figure also includes corrected field output factors measured with two PTW TN60019 microDiamond detectors and data from Lechner et al (2022) for field sizes of 0.5 × 0.5 cm² to 10 × 10 cm². For field sizes of 2 × 2 cm² and larger the agreement between HS RP-200–1 × 1 and PTW TN60019 microDiamond detectors with appropriate corrections was within measurement uncertainty. For field size of 1 × 1 cm² the detector specific output ratio measured with HS RP-200–1 × 1 was 3.2% lower than corrected field output factor measured with PTW TN60019 (#1) microDiamond detector and 3.6% lower than that measured with PTW TN60019 (#2). For field size of 0.5 × 0.5 cm² the detector specific output ratio measured with HS RP-200–1 × 1 was 4.8% lower than corrected field output factor measured with PTW TN60019 (#1) microDiamond detector and 5.9% lower than that measured with PTW TN60019 (#2).

Zoom In Zoom Out Reset image size

The detector specific output ratios measured with HS RP-200–1 × 1 and corrected field output factors measured with two PTW TN60019 microDiamond detectors for Varian TrueBeam^TM 10 MV-FFF energy beam are displayed in figure 6. Again, agreement was within measurement uncertainty for field sizes of 2 × 2 cm² and larger. Largest difference was seen at the smallest field size of 0.5 × 0.5 cm² of 6.2% and 6.4% between HS RP-200–1 × 1 and PTW TN60019 (#1) and (#2), respectively.

Zoom In Zoom Out Reset image size

Little difference was seen between detector specific output ratios measured on two Varian TrueBeam^TM models. Detector specific output ratios measured with HS RP-200–1 × 1 on Varian TrueBeam^TM and Varian TrueBeam^TM STx agreed to within 2% for field sizes down to 0.5 × 0.5 cm². For the field sizes down to 0.6 × 0.6 cm², comparison of these output ratios to the output factors reported by Lechner et al for Varian TrueBeam^TM STx showed agreement within 4%. However, comparison to Lechner data for Varian TrueBeam^TM showed a difference of about 52% for the smallest field size of 0.5 × 0.5 cm².

The angular dependence, relative to gantry at 0°, of the investigated PSD is shown in figure 7. The angular dependence was mostly below 0.1%, and the maximum angular dependence was 0.15% ± 0.01% at a gantry angle of 260°.

Zoom In Zoom Out Reset image size

A dose of 0.91 cGy was recorded over a pre-irradiation measurement spanning 10 min. This is quite interesting considering that part of the calibration procedure was to perform a background reading which is to be subtracted from future measurements. It is therefore expected that a background reading with a correct calibration should be zero. However, the average length of an SRS/SRT treatment at our institution is well below 10 min. Considering an average dose per fraction of 11.7 Gy (based on patients from January 2021—May 2021, inclusive), the 0.91 cGy dark current is less than 0.1% of the average dose delivered, and can be considered clinically insignificant. Readings taken immediately prior to and immediately after irradiation indicated no leakage as all values were comparable to expected background levels.

The short term repeatability of HS RP-200–1 × 1 that was quantified through the standard deviation of 10 readings was 0.04 ± 0.01%. This finding surpasses the short-term repeatability of Extradin W1 PSD, which Carrasco et al (2015) reported as 0.25 ± 0.05% and 0.10 ± 0.07% for respective 15 and 75 cGy irradiations. Also, the percent deviation of 0.04% herein was found to be comparable to the Debnath et al (2021) reported deviations of 0.02% and 0.07% for 100 cGy irradiations for an inorganic scintillating detector in 10 × 10 cm² and 0.5 × 0.5 cm² fields, respectively.

Irradiation of the PSD in the range of 25–2000 MU showed a linear relation between delivered MU and measured signal with less than 1% deviation. At a delivered irradiation of 5 MU, the measured signal deviated by 5.9% from the expected value defined by linear fit. Considering that 5.9% is only 0.16 cGy under these irradiation conditions, this difference is likely to be negligible in many clinical situations. However, during VMAT delivery, considerable fraction of the dose is delivered in small portions, therefore, the impact of non-linearity of this PSD on VMAT dosimetry could be worthy of further investigation.

These results are similar to findings by Schoepper et al (2022) for the HS RP-100 readout system coupled to a 1 mm × 3 mm PSD probe. The group reported a difference of 2.93% at 5 MU, and differences lower than 1% for 10 MU and higher. Jean et al (2021) demonstrated a similar trend with differences over 2% for 5 MU doses and lower, and differences less than 1% for 10 MU and higher. They used the HS RP-100 acquisition system coupled to three PSDs with dimensions of 1 mm × 7 mm, 1 mm × 6 mm and 1 mm × 3.5 mm. The slightly poorer result of HS RP-200–1 × 1 can be explained by its smaller active volume compared to the probes used by Schoepper et al and Jean et al. The HS RP-200–1 × 1 demostrated noticeably worse linearity at low doses when compared to Exradin W1 and W2, which showed deviations of measured from expected dose of only 0.5% (over 0.15–6 Gy) and 0.05% (over 1–1000 MU), respectively (Carrasco et al 2015, Galavis et al 2019).

For dose rates between 200 and 2400 MU min⁻¹, the response was found to be independent of dose rate, given a difference of less than 0.10%. At 100 MU min⁻¹ a difference of 0.48% was observed. For lower dose rates, this difference increased upwards 1.56% for 40 MU min⁻¹. These findings are consistent with Carrasco et al (2015) who found a dose rate independence for Exradin W1 to within 0.5% for dose rates of 200–600 MU min⁻¹ and within 1% for 100 MU min⁻¹. Additionally, comparing to the HS RP-100 platform with 1 mm × 3 mm PSD probe, investigated by Schoepper et al (2022), they found dose rate independence to be within 1.18% for dose rates 200–600 MU min⁻¹, and up to 3.6% difference for dose rates lower than 200 MU min⁻¹. The findings in this report indicate an improved dose rate independence of HS RP-200–1 × 1 compared to larger PSD and earlier version of HS-RP-100 signal acquisition system found in literature.

Given the good agreement of measured 6 MV-WFF energy beam detector specific output ratios for HS RP-200–1 × 1 to field output factors measured with PTW TN60019 microDiamond detectors for field sizes 1 × 1 cm² and larger, we can make a conclusion that for this energy and these field sizes HS RP-200–1 × 1 small field output correction factor, ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}},$ equates to a unity (${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}=1$) within uncertainty of 2.5%. This can be expected considering the similarity of sensitive volume properties between Exradin W1 PSD and Medscint PSD used in this study. The small field output correction factors reported in TRS-483 for Exradin W1 are 1.000 for field sizes down to 0.4 × 0.4 cm² IAEA (2017). However, as our measurements showed up to 5% disagreement of detector specific output ratios and field output factors at field sizes smaller than 1 × 1 cm², further investigation is required to determine the values of ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}$ for Medscint PSD at such fields. Monte Carlo simulations incorporating an accurate detector model based on manufacturer's specifications could be useful in resolving this discrepancy. In our study, two microDiamond detectors were utilized, advantageous in terms of cross-validation and measurement consistency. Nonetheless, we acknowledge that augmenting our study with various detector types or dosimetry techniques, such as film dosimetry, would have further enhanced the comprehensiveness of our findings.

Based on the agreement of measured 10 MV-FFF energy beam detector specific output ratios for HS RP-200–1 × 1 to field output factors measured with PTW TN60019 microDiamond detectors, we can make a conclusion that for 10 MV-FFF beam small field output correction factor, ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}},$ for HS RP-200–1 × 1 equates to a unity (${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}=1$) for field sizes of 2 × 2 cm² and larger. For smaller fields, down to 0.5 × 0.5 cm², ${k}_{{Q}_{\mathrm{clin}},{Q}_{\mathrm{msr}}}^{{f}_{\mathrm{clin}},{f}_{\mathrm{msr}}}$ is expected to be within 6% of unity. Further investigation is required to more accurately determine small field output correction factors for 10 MV-FFF energy at field sizes smaller than 2 × 2 cm².

Lechner et al (2022) reported small field output factors measured on several linac models at multiple institutions using several detectors. For field sizes larger than 1 × 1 cm², the output factors measured in this study agreed with the data reported by Lechner et al within 3%. However, for fields of 1 × 1 cm² and smaller, considerable disagreement was found as shown in figures 4 and 6. Also, results reported by Lechner et al indicate a difference of over 45% between output factors measured on Varian TrueBeam^TM STx (HD MLC) and Varian TrueBeam^TM (Millenium MLC) for field size of 0.5 × 0.5 cm². The detector specific output ratio measurements performed in this study did not show such a large difference between output factors for these models of Varian TrueBeam^TM linacs. Our measurements on these two machines for beam energies of 6 MV-WFF and 10 MV-FFF showed output factor differences of up to 2.7% and 1.6% respectively, where the largest difference was measured at the smallest measured fields of 0.5 × 0.5 cm². These differences are much smaller than those reported by Lechner et al, and the reason for the difference between our results and those by Lechner et al still needs to be clarified.

The angular dependence, relative to gantry at 0°, of HS RP-200–1 × 1 was below 0.15%. This is in agreement with Carrasco et al (2015), who measured a maximum angular dependence of 0.34% for the Exradin W1, and Schoepper et al (2022) who measured a maximum angular dependence of 1.18% for HS RP-100 with 1 mm × 3 mm PSD probe. This finding indicates an isotropic response around the axis of symmetry for HS RP-200–1 × 1.