Dosimetric Measurements using Radioluminescence of Beryllium Oxide

Abstract— The application of ionizing radiation in cancer therapy requires dosimetric verification and quality assurance of the applied therapeutic dose. Conventionally, ionization chambers are used for this purpose. At Technische Universität Dresden an alternative approach to the ionization chamber is under investigation. The detector system is based on a beryllium oxide (BeO) probe coupled to an optical fiber. Radioluminescence in the material generates photons proportional to the dose absorbed in the probe. The photons are counted by very sensitive time resolving single photon counting heads. Connected to the cylindrical BeO probe with a diameter of about 1 mm is a long, thin and flexible fiber, which allows measurements in complex and narrow geometries. Because of the system properties, real-time dosimetry in fields of high dose rates and steep gradients with a high spatial resolution is possible. The first experiments with the detection system were performed in photon and electron fields. In photon fields, the problem of Cherenkov radiation arises, which has already been successfully eliminated by temporal discrimination of the detected events. The current research concentrates on the use in fields of heavy charged particles. In this application, saturation effects appear in the high linear energy transfer (LET) area, like in other luminophores. In consequence, a correction function is required. Such a correction function has been determined for proton fields as a function of the LET-connected dependencies in the spectral composition of the radioluminescence light and has been already successfully tested.

I. INTRODUCTION
The presented measuring system consists of a BeO probe connected to a long and thin fiber. In the BeO probe, photons are emitted when exposing the probe to ionizing radiation. This effect is called radioluminescence. The generated signal is proportional to the dose absorbed in the probe. This principle offers several advantages for clinical use, like a high spatial resolution as well as resistance against external influences, such as magnetic fields and temperature changes. Unfortunately, drawbacks appear which have to be addressed according to the incident radiation type.

II. MEASURING PRINCIPLES
A. Beryllium Oxide Probe and Light Guide
The cylindrical BeO probe has a diameter and a height of 1 mm each and is coupled to the light guiding fiber with transparent epoxy (figure 1). To reduce ionization in the fiber, its diameter is only 0.2 mm. The length of the fiber is 5 m to achieve flexibility in the measuring setup [1]. Furthermore, the probe and the light guide are encased by a black coating to shield the fiber from external light. When ionizing radiation hits the BeO probe, photons are emitted whose number is proportional to the dose. They are guided by the light guide to the detection unit of the system.

B. Integration of a Beam Splitter
In case of photon or electron radiation, the radioluminesence spectrum in BeO has one broad maximum [2].
In contrast, if protons are the incident particles, the spectrum has two maxima and a minimum between them. To separate these maxima, a beam splitter was integrated in the measuring setup whose edge lays in the minimum of the spectrum. The beam splitter divides the photons with wavelengths over 347 nm and the ones with wavelengths under 347 nm. This gives the possibility to differentiate the two groups of photons in the analysis. Therefore, the corresponding events are collected in two separated channels [3]. Both exits of the beam splitter are connected to an own single
photon counting head (SPCH) based on the μPMT technology. These SPCHs by Hamamatsu are very sensitive, so that every photon can be counted. The sensitive area of the SPCHs amounts only 3 mm2 to reduce direct ionization [1]. With this measurement setup the ratio of the number of photons with low wavelengths and the ones with high wavelengths can be determined. This ratio is called γ in the following.

III. MEASUREMENTS
A. Depth Dose Curves in a Water Phantom As shown in [2], in a photon field the depth dose curve measured in a water phantom follows the course of the reference data, when using the method of gated discrimination to eliminate the Cherenkov radiation. Gated discrimination uses the fact, that the decay time of the Cherenkov radiation is much smaller (~ps) than the luminescence lifetime of BeO, which amounts 27 μs. In consequence, the luminescence of BeO is only measured after the decay time of the Cherenkov radiation [2]. In the next step, the detector system was tested in a homogenous (5 x 5) cm2 proton field at the experimental room of University Proton Therapy Dresden (UPTD) using the pencil beam scanning (PBS) nozzle. To examine the dependence of the signal to the residual range of the incident particles, the BeO probe is moved linearly in the water
phantom (figure 2) to record a depth dose curve for a quasi-mono-energetic beam of 110.7 MeV. In proton fields, the measured curves are correlated to the characteristic proton depth dose curve, but with rising LET (deeper position) saturation effects appear (see figure 3). A possible solution for this problem is a correction function. This function can be derived from the ratio between the two detection channels, which is mentioned above as γ. It was observed that γ increases with the penetration depth and consequently increases with a rising LET.

IV. ANALYSIS IN PROTON BEAMS
The correction function to convert the measured light signal into a dose-to-water value (𝐷real) should have the form 𝐷real =𝑓(𝛾)∙𝑀BeO ,
where Dreal names the reference dose, measured with an Advanced Markus Chamber TM34045 (PTW, Freiburg, Germany), f(𝛾) describes the correction function and MBeO is used as symbol for the measuring effect of the BeO probe with wavelengths smaller than 347 nm. This channel was choosed, because the curve has a steeper gradient (see figure 3). Accordingly, the correction function can be determined by approximating the dependency of Dreal/MBeO on γ. Figure 3 shows the depth dose curve, measured with the Markus Chamber, the depth dose curves measured with the BeO probe and the corrected data measured with the BeO probe. Comparing the BeO data and the data of the Markus Chamber, the saturation effect in the BeO probe in high LET region is clearly recognizable.

V. DISCUSSION
The measured curve of the BeO probe follows the course of the characteristic depth dose curve of protons. For not underestimating the dose, the correction function is necessary. The corrected data fit pretty well with the data of the Markus Chamber, hence the position of the Bragg-Peak can be
determined very precisely. However, the maximum of the Bragg Peak was overestimated by the corrected data by 2.3%, because of fluctuations in the ratio Dreal/MBeO.

VI. CONCLUSION AND OUTLOOK
The data measured by the BeO probe were successfully corrected for photon irradiation by gated discrimination and for proton irradiation with a correction as a function of spectral information. In further experiments, the behavior of the BeO probe in fields of charged particles, which are heavier than protons, will be investigated. Furthermore, the small size of the BeO probe has the potential for spot-wise dosimetry or small-field-dosimetry.

ACKNOWLEDGMENT
The experimental part of the UPTD facility has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 730983 (INSPIRE).