Development and evaluation of a monoxide-based flexible skin dosimeter for radiotherapy at photon energies

Radiation therapy uses high-energy radiation that can cause various side effects depending on the patient's exposure. In particular, the radiation exposure of the skin to reach the target volume produces side effects. Therefore, side effects as skin damage can be reduced during clinical trials using various skin dosimeters such as films and glass detectors to determine the skin dose exposure. However, accurately measuring the doses using these dosimeters is challenging due to human body curvature. In this study, a flexible skin dosimeter was developed using the photoconductor materials mercury oxide (HgO) and lead oxide (PbO). The performance of the proposed dosimeter was evaluated by measuring reproducibility, linearity, dose rate independence according to dose, and percent depth dose (PDD) at photon energy beam. The results showed that the flexible skin dosimeter using HgO material has high applicability as a skin dosimeter due to its stability compared to PbO. The results also provided useful insights for the radiation therapy field, particularly in areas where radiation measurement is difficult, depending on the human curvature. The proposed apparatus could serve in various radiation detection areas as a flexible, functional material.


A
: Radiation therapy uses high-energy radiation that can cause various side effects depending on the patient's exposure. In particular, the radiation exposure of the skin to reach the target volume produces side effects. Therefore, side effects as skin damage can be reduced during clinical trials using various skin dosimeters such as films and glass detectors to determine the skin dose exposure. However, accurately measuring the doses using these dosimeters is challenging due to human body curvature. In this study, a flexible skin dosimeter was developed using the photoconductor materials mercury oxide (HgO) and lead oxide (PbO). The performance of the proposed dosimeter was evaluated by measuring reproducibility, linearity, dose rate independence according to dose, and percent depth dose (PDD) at photon energy beam. The results showed that the flexible skin dosimeter using HgO material has high applicability as a skin dosimeter due to its stability compared to PbO. The results also provided useful insights for the radiation therapy field, particularly in areas where radiation measurement is difficult, depending on the human curvature. The proposed apparatus could serve in various radiation detection areas as a flexible, functional material.

K
: Radiotherapy concepts; Dosimetry concepts and apparatus; Radiation damage to detector materials (solid state)

Introduction
Radiation therapy uses high-energy radiation to treat cancer patients. However, high-energy radiation used in radiation therapy affects both tumor and normal tissues, causing side effects. Among other things, the skin is directly exposed to radiation because it must pass through the skin to reach the target volume during treatment. In addition, the side effects of hair loss, erythema, and blisters occur most frequently in radiation treatment because of their high sensitivity to radiation [1,2]. Therefore, skin side effects due to radiation therapy should be minimized by accurately measuring the radiation dose exposed to the skin. Current skin doses are calculated using a treatment planning system; however, the accuracy of dermal doses has an uncertainty of ±20% [3][4][5][6]. Therefore, in clinical patient treatment, dosimeters are used to verify the skin dose, such as film, glass dosimeter (GD), optically stimulated luminescent dosimeter (OSLD), and thermo luminescent dosimeter (TLD). Nevertheless, with these skin dosimeters, accurate dose measurements are challenging due to the curvature of the human body surface [7][8][9]. In addition, real-time measurement is not possible in EBT3 films, which is often used for skin dose measurement. Thus, clinical patient treatment requires a digital dosimeter capable of real-time measurements with flexible materials with no functional loss when bent along the human body curve. However, thus far, flexible dosimeter studies are insufficient, even though various materials have been studied in the field of radiation detectors. Among them, lead oxide (PbO) and mercury oxide (HgO) have high atomic numbers ( Hg : 80, Pb : 82, O : 8) and density (11.14 g/cm 3 , 9.53 g/cm 3 ). Therefore, they have been actively used in direct conversion studies [10][11][12][13]. These materials can be used to produce flexible materials through the particle in binder (PIB) method, which is conducted by mixing them with silicone binder in polycrystalline form. Because the PIB method is simple for large-scale manufacturing and simpler than the single crystal manufacturing method, this method can lower the manufacturing unit price [11][12][13][14][15]. This study evaluates the performance of unit cell dosimeters based on HgO and PbO to lay the foundations for developing flexible large-area dosimeters.
Therefore, a flexible skin dosimeter using the PIB method was developed in this study. Moreover, its applicability is demonstrated by evaluating reproducibility, linearity, dose rate independence, and percentage depth dose (PDD), according to the assessment items of the quality assurance for radiation treatment.

Experimental method
In this study, a flexible skin dosimeter was produced using the PIB method, which is simple to apply to photometric substances and silicone binders. The performance of the skin dosimeter was evaluated.

Fabrication of flexible skin dosimeter
For the lower electrode, a heat-resistant film was applied with indium tin oxide (ITO). The radiation absorption layer was manufactured by mixing 99.99% purity HgO, PbO (Kojundo Chemical Laboratory Inc., Japan) material, and silicone binder at a ratio of 4 : 1 and applying a screen printing technique to frames of 1 × 1 cm 2 . Megavoltage radiation dosimetry Gafchromic EBT3 (International Product, Wayne, U.S.A., NJ) for surface measurement having a thin thickness (thickness of 0.280 mm) was used. In this study, the dosimetry thickness was reduced to 0.150 mm to reduce the attenuation rate further [16][17][18]. Subsequently, the radiation absorption material was dried for 12 hours at a fixed temperature of 40 • C. The vector placement method was used to create an upper electrode on the top of the radiation absorption material. The upper electrode used gold (Sigma Aldrich Inc. U.S.A.) with 99.999% purity to collect charges. Moreover, the size of the upper electrode was 0.8 × 0.8 cm 2 .  Figure 6 shows the experimental set-up. The performance of the flexible skin dosimeter was evaluated considering the reproducibility, linearity, dose rate independence, and PDD of the HgO and PbO flexible skin dosimeter at 6 MV and 10 MV. A LINAC system (Infinity; Elekta AB, Stockholm, Sweden) was used for measurements. Considering the 6 MV and 10 MV max photon energy, the build-up materials were set at 1.5 cm and 2.1 cm, respectively. The build-up material used a Slab phantom of equivalent tissue thickness (PTW, RW3, Germany). The source-to-surface distance (SSD) was set to 100 cm. Waveforms were acquired by the oscilloscope (WaveSurfer 510, Teledyne LeCroy, U.S.A.) to collect signal values from radiation. When radiation was irradiated to the photoconductor, electrons were emitted. The emitted electrons flowed into the oscilloscope due to the applied voltage. The oscilloscope calculated the accumulated charge amount in the form of a graph and digital data. The charge was calculated from the collected waveforms using ACQ software (Biopac, AcqKnowledge 4.2, Canada). At this time, a drive voltage of 1 V/μm was applied to the circuit using electrometers (Keithley, 6517A, U.S.A.). Table 1 shows the radiation irradiation conditions considered in the experiment.

Evaluation
In this study, precision and accuracy were evaluated through reproducibility and linearity measurements. In addition, dose rate independence and PDD were evaluated to analyze the response characteristics. Flexible skin dosimeters were irradiated ten times repeatedly for reproducibility measurements. The signals obtained from the first beam were normalized to evaluate the response characteristics of repeated irradiations. Reproducibility assessment can be expressed in RSD based on the amount of the acquired signals. The RSD was calculated as follows: where and Ave are the measured signal value and mean average signal value, respectively. Moreover, is the number of measurements. The evaluation criteria were set within 1.5% of the RSD value corresponding to the 95% confidence level [19]- [22]. Linearity was evaluated through the coefficient of determination (R 2 ) of the linear regression that irradiated radiation by gradually increasing the dose in 3, 10, 50, 100, 200, 300, and 400 MU under 50 MU/min conditions. The evaluation criteria for R 2 were not less than 0.999. The stability of the signals and the possibility of developing a flexible skin dosimeter were evaluated through the reproducibility and linearity results.
The dose rate independence was evaluated by increasing the doses gradually by irradiating from 1 to 400 MU under 100, 300, and 400 MU/min. The measured signals were normalized to 200 MU, and the RSD ( = 3) for 100 MU measurements was calculated and compared to the diode. The dose rate dependence was evaluated in the same experiment as in [20].
PDD was obtained by increasing the Slab Phantom thickness from 0.1 cm to 25 cm. The results were normalized to calculate the percentage based on the max point and compared with those from a thimble chamber.

Reproducibility
Reproducibility was analyzed to evaluate the stability of the flexible skin detector signal made of a monoxide material and silicone binder. Figure 2 shows the reproducibility results from repeated irradiation. After analyzing the reproducibility by irradiating radiation ten times, the relative standard deviation (RSD) value of the HgO flexible skin dosimeter was 1.7% and 2.3% at 6 MV and 10 MV, respectively. The RSD value for the PbO flexible skin dosimeter was 1.4% and 1.4% for 6 MV and 10 MV, respectively.

Linearity
Linearity was analyzed to evaluate the accuracy of the output signals according to the irradiation dose. Figure 3

Dose rate independence
The dose rate independence was evaluated, analyzing a value related to the response characteristics by dose-rate.

Percent depth dose
PDD was obtained by measuring the dose at depth and increasing the slap phantom from 0.1 cm to 25 cm. Figure 6 shows the PDD measurement results of the flexible skin detector. The PDD result of the flexible skin dosimeter shows the ideal Dmax value at the Dmax point for 6 MV and 10 MV energy to represent the ideal PDD graph. In particular, for the PDD value at 10 cm 6 MV, HgO was 69.98%, and PbO was 71.15%. These values compared to the PDD the thimble chamber values of HgO and PbO differed by 2.4% and 3.75%, respectively. At 10 MV, HgO was observed at 72.24%, and PbO was 73.02%. When compared these values with the PDD of the thimble chamber, the difference was 0.14% and 0.92%, respectively.
The above results provide a foundation for developing a flexible large-area dosimeter capable of obtaining real-time 2D skin dose maps, focusing on evaluating the performance of unit cell dosimeters based on HgO and PbO.

Discussion
In this study, HgO and PbO were analyzed at 6 MV and 10 MV energy to evaluate their applicability as flexible skin detectors. The reproducibility measurement of HgO showed an RSD of 1.7% at 6 MV and 2.3% at 10 MV. The results for the HgO also showed linearity. In this study, the RSD was higher than the baseline of 1.5%, indicating that the signal is unstable.
In contrast, PbO materials were present at 1.4% at 6 MV and 1.4% at 10 MV. These results indicate that the signal was stable because it satisfied a 95% confidence interval with values below the 1.5% threshold [19].
In the linearity analysis, HgO showed very good linearity with a value of 0.9999 at 6 MV. However, at 10 MV, the result was 0.9986, which was lower than the benchmark of 0.9990. In contrast, PbO showed excellent linearity of 0.9994 at 6 MV and 0.9992 at 10 MV. Consequently, the graph was linear with respect to the dose.
Concerning the dose-rate independence evaluation, the RSD value for the HgO was 0.5% at 6 MV and 2.0% at 10 MV. Moreover, the RSD value for the PbO was 0.3% at 6 MV and 0.3% at 10 MV. In other studies, the standard deviation of the HgI 2 -, diamond-, and silicon diode-based dosimeter for 6 MV energy was 4.1%, 4.3%, and 0.3%, respectively. When the results are compared with those of the manufactured flexible skin dosimeter at 6 MV, the latter has a similar or better dos rate independence than other dosimeters [19].
In the studies regarding skin dose measurements, the result value at the surface differed from each detector due to the build-up of the component itself due to skin uncertainty. For instance, in [23], the dose of each surface dosimeter showed a difference of more than 27% for 6 MV. When comparing and analyzing the PDD value at 10 cm depth of the thimble chamber, the difference between HgO and PbO represented 2.4% and 3.7%, respectively, under 6 MV conditions. At 10 MV energy, the difference between HgO and PbO was 0.14%, and the difference between PbO was 0.92%. These results could be due to experimental variables in the experimental set-up and air gap generation using the Slab phantom. The drop in the results of HgO could be due to the charge of the flexible skin detector remaining in the electric field after the study.
The irregular particle size and silicone binder of the HgO material used in the manufacture of flexible skin dosimeters were not evenly mixed. Thus, a gap between the two materials was present, which is believed to cause unstable results, such as the above, because of the capturing effect of electric charges [20]. In the future, it is necessary to supplement the problem after a sufficient roll-mill process.