3.1. Pulse Response Characteristics of Electron Beam
The ISD was positioned in the beam path at a source-to-surface distance of 100 cm, which represents a standard setup for oncology QA procedures. The ISD was then irradiated by the linac beam from directly above (0° Gantry Rotation) with the operating conditions of a 6 MeV electron beam and a 500 monitor units/min (MU/min) dose rate, for a duration of 24 s.
Figure 4a shows the optical intensity (photon counts) for consecutive macro-pulses of beam radiation, over a duration of 24 s. The excellent time resolution of the MPPC detector (0.1 ms gate time) has meant that the data captured using the device allows the individual electron-beam micro-pulses of the linac to be detected. By zooming in on 20–20.01 s of data, we can see that a single micro-pulse is made up of multiple data points in
Figure 4b. Each point is the number of fluorescent photons received by the MPPC in 0.1 ms. The individual pulses here do not fully represent the shape of electron pulses; the falling edge observed is actually determined by the decay time of fluorescence.
In order to illustrate the average effect of the pulses, the red trace in
Figure 4 corresponds to the 1000-point moving average imposed on the data (the blue trace in
Figure 4). By comparing the blue trace and the red trace in
Figure 4, it can be observed that the blue trace shows an increasing trend in the first three seconds, while the red trace remains relatively stable. This appears to present a contradiction. To explain this phenomenon, we extracted data from the intervals of 1–2 s and 20–21 s for further analysis, as shown in
Figure 5. The 1 s integration period was chosen to ensure that a sufficient number of micro-pulses would be captured to observe the pattern of the electron-beam pulses. It is evident that although the micro-pulse intensity in
Figure 5a is relatively low, the frequency of micro-pulses is significantly higher than that in
Figure 5b. Within a one-second timeframe, there are 63 micro-pulses in
Figure 5a, whereas only 45 micro-pulses are observed in
Figure 5b. This ensures that the red trace representing the average remains stable throughout the process, both during the initial three seconds of micro-pulse instability and the subsequent 20 s when the micro-pulse intensity is relatively constant. This phenomenon, referred to as DRS (dose rate servo), involves a type used in some Varian machines called the pulse drop servo [
12], which regulates the dose rate by dropping individual micro-pulses.
The data presented in
Figure 4a were utilized to calculate the accumulated dose as recorded by the sensor. This calculation was performed by integrating the averaged dose values on a point-by-point basis. The results of
Figure 6 show the linear increase in dose measured by the ISD during the application of the irradiation in macro-pulses (as shown in
Figure 4a) from the linac, with a rate of increase of 2.94 × 10
5 per second, and the final value of the accumulated dose reaches 6.73 × 10
6.
From
Figure 5, it can also be observed that the micro-pulses of the electron beam do not exhibit periodicity. This phenomenon is markedly different from the pattern observed when the linac emits X-rays, where the frequency of the pulses remains stable. We conducted measurements under the same experimental conditions except for changing the radiation to X-rays with a dose rate of 600 MU/min, and the results are shown in
Figure 7. This indicates that the repetition rate (frequency) of the X-ray micro-pulses is 360 Hz, which was reported by O’Keeffe (2016) [
13].
The pattern of micro-pulse delivery over a time period of 2 s was selected and recorded at a selected time of approximately 10 s, which ensures that the recording period fully coincides with the active beam delivery phase of the macro-pulse. This was repeated and recorded in the same manner as
Figure 5, but with the dose rate varied from 100 MU/min (as in
Figure 5) to 500 MU/min in steps of 100 MU/min. The data zoomed in to the 2 s window at the time of 10 s, corresponding to every case, are shown, respectively, in
Figure 8a–e. As can be seen from
Figure 8, the micro-pulses at any dose rate, over the full measurement range of 100 to 500 MU/min, do not exhibit periodicity. This phenomenon is also entirely different from the regularity of X-ray emission micro-pulses [
13].
If the number of complete micro-pulses captured in each frame from
Figure 8a–e is counted and the accumulated value within two seconds is calculated, the following table can be constructed (see
Table 1). Although the micro-pulses of the electron beam do not exhibit periodicity, the data contained in
Table 1 clearly demonstrate that there is a linear relationship between the number of complete micro-pulses captured in a 2 s window and the dose rate, as shown in
Figure 9a. As the dose rate increases, a noticeable increase in the number of individual micro-pulses can be observed within a 2 s time window. Similarly, there is a clear linear relationship between the accumulated intensity within two seconds and the dose rate, as shown in
Figure 9b. However, both the linearity of the pulse count and the linearity of the accumulated intensity are slightly lower (R
2 = 0.9941 and 0.9966). This is mainly due to the instability of the pulses, as shown in
Figure 4a. If we compare the total accumulated count within 24 s with the dose rate, we find that the linearity of the dose rate is very excellent, as shown in
Figure 10. A linear regression analysis shows that the R
2 value in this case is 0.9993. Also, the intercept is very close to the (0, 0) point, the y axis intercept being at (0, −43548). This corresponds to a value of 0.65% of the maximum value on the y axis.
The results in
Figure 10 unequivocally demonstrate that although the amplitude of individual micro-pulses of electron-beam irradiation delivered by the linear accelerator may fluctuate, if the amplitude data are averaged over a sufficiently long interval (in this case 2 s), the fluctuations become insignificant, and the average amplitude value can highly accurately reflect the dose rate. It is crucial to confirm the mode of radiation delivery in pulse form, as it is essential for ensuring the accuracy of dose measurement by the ISD, especially if the ISD is intended for use as a real-time dosimeter in clinical applications.
3.3. Depth-Dose Experiment
The ISD’s depth-dose measurement was performed with a dose rate of 400 MU/min at 6 MeV in depth from 0 cm to 5 cm in steps of 1 mm. The results from these measurements were compared with similar measurements for the PTW31010 0.125cc cylindrical Ionization Chamber (Physikalisch-Technische Werkstätten [PTW], Freiburg, Germany). This PTW 31010 miniature ionization chamber was calibrated at the National Institute of Metrology, China, in December 2023. It has been used for routine absolute dose measurements in G3 CyberKnife treatments and has demonstrated good repeatability.
Figure 13 shows the results of the depth-dose profile for the 6 MeV electron beam measured using ISD and IC. The Dmax for the 6 MeV beam measured using the IC occurred at a depth of 15 mm, while the ISD’s D
max was found to be at around 7 mm. When using the ISD to measure the point of maximum dose for an electron beam, a significant forward shift is observed compared to the point of maximum dose measured by the IC. This is contrary to the results obtained when measuring X-rays, as shown in
Figure 14.
The phenomenon of the maximum dose point shifting backward when ISD is used to measure X-ray has been reported and explained by Qin et al. (2019) [
14] and He et al. (2024) [
15]. This is because the absorption of radiation by water or human tissues is caused by the energy loss of secondary electrons, and there is only a small number of backscattered electrons in the shallow layer of water. But for ISD, it can directly absorb the scattered photons to emit fluorescence, resulting in a larger signal output. The difference in response mechanisms between water and inorganic scintillators ultimately leads to the phenomenon of the maximum dose point shifting backward under X-ray irradiation.
For the electron beam, according to the formula for collisional stopping power of electrons:
Ionization loss increases slowly with electron energy at high energies. Conversely, at low energies, ionization loss is inversely proportional to electron energy. When high-energy electrons propagate in water, their range is very short, and energy is rapidly lost, converting them into low-energy electrons. As ionization loss increases significantly, the response of IC quickly rises to reach the maximum dose point. With the depletion of electron energy, ionization loss decreases rapidly, causing the PDD curve to drop sharply and eventually approach zero.
For ISD, based on the formula for the radiative stopping power of electrons:
When electrons pass near atomic nuclei, they lose energy, which is emitted as bremsstrahlung radiation under the influence of the Coulomb field. This radiative loss is proportional to the electron energy and the material’s Z
2. At shallow water depths, the combination of high-energy electrons and the high atomic number (Z) of the scintillator leads to a rapid generation of substantial bremsstrahlung within the ISD, exceeding that produced in water (over 4–20 MeV energy range). Owing to the luminescence mechanism of Gd
2O
2S:Tb, these photons can be directly absorbed to generate fluorescence, resulting in a steeper and faster rise of the PDD curve for the electron beam compared to the actual curve measured by an ionization chamber [
14,
15]. As the electron energy decreases rapidly, the number of newly generated bremsstrahlung photons diminishes quickly, leading to a decline in the PDD curve and causing the maximum dose point to shift forward.
At water depths greater than 30 mm, the response intensity measured by the ISD remains higher than expected and does not decrease to zero. This is due to the fact that although the intensity of bremsstrahlung radiation is low, the propagation distance of bremsstrahlung photons is long, allowing them to penetrate depths much greater than 30 mm. For the ISD, due to the dependence of the photoelectric effect on the atomic number, low-energy bremsstrahlung photons can still interact with Gd2O2S:Tb via the photoelectric effect, producing a certain number of secondary electrons and resulting in some dose absorption. Therefore, there is always a certain level of intensity at depths below 30 mm. In addition, the generation of Cerenkov may also cause some response. High-energy electrons in water can produce Cerenkov radiation, which, although relatively weak in intensity, still contributes to a measurable level of intensity at water depths below 30 mm. The actual impact of Cerenkov radiation will be carefully evaluated in our future work using bare optical fibers and Monte Carlo simulations.
As the ratio between the radiation dose received by water (D
IC) and the dose absorbed by the scintillator (D
ISD) confirms, the depth-dose response measured by the ISD can be calibrated by multiplying the correction coefficient.
Figure 15 shows the curve of this ratio as a function of depth.
The energy of the electron beam was adjusted to 9 MeV, and the PDD experiment was repeated under identical experimental conditions, as shown in
Figure 16. It can be observed that as the electron energy increases, the range of the electrons becomes longer, resulting in a significant backward shift in the maximum dose point in the PDD curves measured by both ISD and IC.
3.4. Beam Profiling Experiment
The off-axis ratio curves of the ISD in the x-axis direction were experimentally tested under a 10 × 10 cm
2 radiation field with 6 MeV and 9 MeV electron beams. These measurement data were compared with those obtained using a 0.125cc cylindrical Ionization Chamber, as shown in
Figure 17.
From
Figure 17a, it can be observed that under the irradiation condition of 6 MeV electron beams, the in-of-field region measurement results of ISD are highly consistent with those of IC, and both dose responses tend to flatten. In the penumbra region, the curve measured by ISD is also consistent with that of IC. However, there is a certain difference in the out-of-field region, which is similar to the reason for the non-zero response intensity observed in PDD experiments at depths greater than 30 mm. This is mainly due to bremsstrahlung radiation generated by the interaction of high-energy electrons with water, which propagates into the out-of-field region and interacts with the scintillation material, thereby producing a certain intensity.
As shown in
Figure 17b, we can see that under the irradiation condition of 9 MeV electron beams, the test results of ISD in both the in-of-field and out-of-field regions are similar to those under the 6 MeV condition. However, there is a noticeable difference in the penumbra region, where the curve measured by ISD shows a slower rate of decline compared to the results obtained by IC. The reason for this difference is evident: as the electron energy increases, the probability of bremsstrahlung radiation production significantly increases, as shown in Equation (2). This additional X-ray contamination interacts with the scintillation material even in regions where electrons cannot easily reach, leading to an over-response phenomenon and increasing the width of the penumbra region.
From the OAR experiment of this electron beam, it can be seen that under the 6 MeV condition, ISD demonstrates a considerable level of accuracy in measuring radiation dose; however, under the 9 MeV electron-beam condition, a certain degree of calibration is required for ISD to function properly.