Initial testing of a pixelated silicon detector prototype in proton therapy

Abstract As technology continues to develop, external beam radiation therapy is being employed, with increased conformity, to treat smaller targets. As this occurs, the dosimetry methods and tools employed to quantify these fields for treatment also have to evolve to provide increased spatial resolution. The team at the University of Wollongong has developed a pixelated silicon detector prototype known as the dose magnifying glass (DMG) for real‐time small‐field metrology. This device has been tested in photon fields and IMRT. The purpose of this work was to conduct the initial performance tests with proton radiation, using beam energies and modulations typically associated with proton radiosurgery. Depth dose and lateral beam profiles were measured and compared with those collected using a PTW parallel‐plate ionization chamber, a PTW proton‐specific dosimetry diode, EBT3 Gafchromic film, and Monte Carlo simulations. Measurements of the depth dose profile yielded good agreement when compared with Monte Carlo, diode and ionization chamber. Bragg peak location was measured accurately by the DMG by scanning along the depth dose profile, and the relative response of the DMG at the center of modulation was within 2.5% of that for the PTW dosimetry diode for all energy and modulation combinations tested. Real‐time beam profile measurements of a 5 mm 127 MeV proton beam also yielded FWHM and FW90 within ±1 channel (0.1 mm) of the Monte Carlo and EBT3 film data across all depths tested. The DMG tested here proved to be a useful device at measuring depth dose profiles in proton therapy with a stable response across the entire proton spread‐out Bragg peak. In addition, the linear array of small sensitive volumes allowed for accurate point and high spatial resolution one‐dimensional profile measurements of small radiation fields in real time to be completed with minimal impact from partial volume averaging.

been tested in photon fields and IMRT. The purpose of this work was to conduct the initial performance tests with proton radiation, using beam energies and modulations typically associated with proton radiosurgery. Depth dose and lateral beam profiles were measured and compared with those collected using a PTW parallelplate ionization chamber, a PTW proton-specific dosimetry diode, EBT3 Gafchromic film, and Monte Carlo simulations. Measurements of the depth dose profile yielded good agreement when compared with Monte Carlo, diode and ionization chamber.
Bragg peak location was measured accurately by the DMG by scanning along the depth dose profile, and the relative response of the DMG at the center of modulation was within 2.5% of that for the PTW dosimetry diode for all energy and modulation combinations tested. Real-time beam profile measurements of a 5 mm 127 MeV proton beam also yielded FWHM and FW90 within AE1 channel (0.1 mm) of the Monte Carlo and EBT3 film data across all depths tested. The DMG tested here proved to be a useful device at measuring depth dose profiles in proton therapy with a stable response across the entire proton spread-out Bragg peak. In addition, the linear array of small sensitive volumes allowed for accurate point and high spatial resolution one-dimensional profile measurements of small radiation fields in real time to be completed with minimal impact from partial volume averaging.

| INTRODUCTION
As imaging techniques continue to evolve, the targets being presented for treatment in our clinical practice are becoming smaller in size, oftentimes requiring beams of less than 1.0 cm in diameter for treatment. This technical challenge is compounded by expanding treatment sites, including functional radiosurgery, which demand that small beams be delivered with high precision to high doses. These small beam sizes require alternative dosimetry methods including diodes, 1,2 micro-ion chambers, 3,4 and film 5,6 for measuring beam output, as standard radiotherapy ion chamber devices exhibit partial volume averaging due to their relatively large sensitive volume (SV) size. Protons have the added complication that their Linear Energy Transfer or LET varies as a function of depth (or energy), which can significantly impact detector response. Proton beam scanning and intensity-modulated proton therapy (IMPT) is another area where accurate and efficient methods for real-time measurements with high spatial resolution are necessary. In the case of proton beam scanning, not only is an understanding of the machine output at a specific point essential to accurate dose delivery, but accurate beam profile information at various depths in water is also critical for accurate treatment planning and reproducible beam delivery.
For small-field and beam profile measurements, radiochromic film has often been seen as the standard metrology device, providing high spatial resolution for such applications. However, in the case of proton therapy, radiochromic film can exhibit a varying response to changing LET. 6 Additionally, radiochromic film also exhibits a number of technical and ease-of-use limitations that can limit its deployment in regular clinical QA programs. Chief of these is that postexposure processing limits radiochromic film's ability to provide real-time data. 7 In addition, artifacts can be introduced in dose measurements due to properties of the film itself (e.g., variations in active layer or substrate thickness, postexposure intensification), environmental and handling effects (e.g., temperature, light sensitivity), and scanner response (e.g., lateral position artifact, 8 film orientation, 9 dust, fingerprints, film curl, and local-, inter-and intrascanner factors). 7 Finally, inter-and intrafilm lot and scanner variation can result in inconsistent dose mapping between film lots, scanner models, and environmental processing conditions. 10 Recent improvements in technology and techniques, such as multiple-channel dosimetry, 11 simplified calibration and intralot recalibration, 10 calibration-less relative dosimetry, 12 and film that is less sensitive to light have significantly mitigated these concerns.
They have not, however, eliminated them entirely.
In an effort to identify more efficient and real-time methods of small-field dosimetry, the team at the Center for Medical Radiation Physics (CMRP) at the University of Wollongong have continued solid-state development of small pixelated arrays of monolithic silicon diode detectors. The prototype device tested in this work is referred to as the dose magnifying glass (DMG). It is a pixelated silicon detector that has the potential to provide not only point dose measurements with a high degree of spatial resolution, but also beam profile measurements in real time. However, while this device has been tested in photon therapy, IMRT, and stereotactic radiotherapy, [13][14][15] it had yet to be tested in proton therapy. Accordingly, the uniformity of response across the proton spread-out Bragg peak (SOBP) was unclear.
The goal of this work was to evaluate a DMG prototype in proton fields typically associated with radiosurgical applications at Loma Linda University Medical Center (LLUMC). The response of the device was compared to a commercially available PTW proton diode, a PTW plane-parallel ionization chamber, and Gafchromic EBT3 film.
Additionally, an in-house developed and validated Geant4-based Monte Carlo application was used for comparison. Both depth dose and lateral profiles were evaluated. It was hypothesized that these tests would not only evaluate the DMG against established forms of metrology, but also identify future directions for development.

| METHODS
The DMG is an array of 128 n+-strips of 2 mm length and 20 lm width on a p-type silicon substrate 380 lm thick. The pitch (or SV center-to-center separation) is 100 lm with a p-stop implantation between the microstrips (80 lm) for compensation of the accumulation layer generated by irradiation of the thick silicon dioxide. [13][14][15] A schematic of the DMG SV array is displayed in Fig. 1. The assembly was pre-irradiated to a dose of 4 MRad with Co-60 before deployment for testing.
The DMG is glued and wire bonded onto a 400 lm thick kapton carrier to provide the proper fan out of the signal for connecting the readout electronics and minimize the impact of these connections on the proton scatter conditions. The detector is used in passive mode (no bias applied at the contacts) and readout is carried out with the detector configured as a planar detector with a common electrode p+ from the same side as the n+ strips. For this work, the DMG was mounted to a Lucite probe holder ( Fig. 2) for rigidity and to facilitate mounting in a water tank; however, the compact nature of the F I G . 1. Schematic of two n+ SV elements in the DMG detector. Note the entire DMG is comprised of 128 SV elements. device allows for a wide range of mounting options, including within specialized phantoms and probes. The two TERA06 chips read out the 128 detector channels simultaneously. 16 Each channel of the TERA06 is a charge-to-frequency converter equipped with a 16-bit counter that records the number of times the input charge (accumulated into a capacitor during the integration time) exceeds the quantum charge, settable by an analog potentiometer. A digital reset is used for zeroing the counter values immediately after each frame is acquired. A graphical user interface (GUI) has been custom designed at the CMRP and provides the operator with all the controls to acquire the data from the detector in real time and also perform preprocessing of the raw binary data.  23 The treatment nozzle configuration and proton energies were chosen to mimic the experimental conditions described above (Fig. 5). In particular, simulated 127 or 157 MeV protons were delivered through a single-stage scattering system to

| RESULTS
The DMG can be operated as a simple dosimeter, with charge measured over either a single SV or multiple SV's across a given linear displacement. The single SV option allows for point measurements of a high spatial resolution (2 mm long and 20 lm wide) to be made, which is of significant benefit in very small or high-gradient radiation fields. The disadvantage of this mode of operation is that the small SV requires longer acquisition times to accrue sufficient statistics. It is envisaged that the DMG would be operated more typically as a tested. Additionally, the DMG data do not show a decrease in response over the SOBP region and the shape of the SOBP compares well with the other detector modalities. Both of these features indicate an LET independence of this device for clinical proton therapy energies, albeit additional measurements may be necessary with discrete proton energies to validate this.
In routine QA applications, it is expected that the DMG will be operated using all SV elements to gain both point dose measurement data as well as beam profile data. Data were collected for 5 mm diameter, 127 and 157 MeV proton beams at various depths along the depth dose profile (Figs. 8 and 9). Excellent agreement in profile shape was observed for all measurement locations when compared with Monte Carlo and EBT3 film data. Note that the edge of the detector array is at a displacement of 1.

| DISCUSSION
The beam profiles measured with the DMG showed excellent agreement with Monte Carlo and Gafchromic film data. The high spatial resolution of the device is comparable to that of film and allows for measurement of the smallest fields currently used in clinical proton therapy (e.g., 5 mm diameter). Unlike film, that requires postprocessing and special handling, the DMG provides readout that is refreshed in real time via a graphical user interface. The real-time nature of the data acquisition is very useful, especially for proton pencil beam scanning applications, where efficient real-time feedback to accelerator staff is important during the acceptance testing and commissioning process. When comparing with other data sets or if further analysis is required, the data can be output for graphing and analysis using any of the common computer-based tools for this purpose.
The DMG tested in this project is a prototype instrument and certain refinements are necessary prior to clinical deployment in proton therapy. Firstly, several channels were removed from analysis because they either were unresponsive or produced excessively high amounts of electronic noise that could not be corrected for. Such channels may be caused by issues in the manufacturing process and may be exacerbated by sensitive cabling and connections. The loss of data can impact measured results and lead to the generation of an incomplete data set, however, in this study the suppression of these channels in our prototype did not significantly impact the beam profile agreement with either Monte Carlo or EBT3 film. In addition, for production DMG detectors, scanning laser systems such as those used in high energy physics 24 could be used to evaluate the DMG systems prior to deployment allowing for selection of units that have all of their SV elements operating within acceptable limits.
Beam profile measurements for 157 MeV protons exhibited increased noise and statistical uncertainty, which can be traced to the cabling system in this existing prototype. The aluminized ribbon cables between the DMG probe and the DAQ system proved to be very delicate and underwent degradation as the experiments progressed. The ruggedness of this component will have to be addressed as we move toward a clinical system.
When applying a detector system to scanned proton therapy, the detector's ability to handle high instantaneous dose rates needs to be considered. In this study, the DMG was tested in a pulsed pro-  60 Gy/min. While these dose rates are rarely seen in passively delivered clinical proton therapy applications, they can be encountered in proton beam scanning applications. Additionally, as the trend of radiation treatment has been toward higher dose rates and faster treatments, especially when applied to SRS and SBRT treatments, it can be foreseen that a uniform response at higher dose rates would further widen the clinical application of the DMG. An improvement in this regard is already under investigation, as we believe that uniform response with dose rates exceeding 12 Gy/min (average) may be achieved by reducing the integration time of the system. By doing so, it is hypothesized that counts will not be missed when the collected charge is too high (i.e., in a high dose-rate environment).
The SVs and a pitch of 2 mm) in dose mapping on a medical linac. 26,27 The DMG tested here has demonstrated that it can be used to provide real-time measurements of both point dose and beam profiles of pencil proton beams with up to 0.02 mm resolution. Such resolution is key in obtaining a precise picture of the sharp dose gradients when the detector is properly orientated with the radiation field. This information, while very useful for collimated beams used in proton radiosurgery and functional radiosurgery, is also especially useful in pencil beam scanning applications for both commissioning and routine QA. During the commissioning process, the collection of accurate beam profile data as a function of depth in water for the complete clinical range of proton energies is essential for evaluation of system performance/stability and also for input to the treatment planning system (or validation of a Monte Carlo system which generates data for the treatment planning system). 28 Such measurements are typically completed using radiochromic film and/or scanned ion chamber detectors; 28,29 however, a pixelated silicon detector such as the DMG may provide a means for measuring these profiles in real time with greater efficiency as profile data can be measured at each depth without the need for lateral detector scanning. Additionally, for daily QA of pencil beam proton systems, many centers use ion chamber arrays such as the I'mRT MatriXX (IBA Dosimetry, Schwarzenbruck, Germany) which is comprised of a 2D array of 32 9 32 ion chambers with a center-to-center spacing of 7.6 mm and a SV size of 4 mm diameter and 5 mm height. The DMG may help augment these systems, especially if it can be developed as a nozzle-mounted 2D pixelated transmission sensor system. 30 The improved spatial resolution of the DMG would allow for more accurate daily analysis of the proton beam size and shape for evaluation of accelerator/beam transport performance.
F I G . 9. Measured and simulated beam profiles at varying depths in water for a 157 MeV proton beam. Geant4 simulation data are also provided for comparison. A simulated central beam axis depth dose profile utilizing 0.25 9 0.25 9 0.25 mm 3 voxels with DMG measurement locations marked is provided for reference (top).

| CONCLUSION
The DMG prototype tested here proved to be a useful device at measuring depth dose profiles in proton therapy, with negligible variation in response across the clinical proton SOBP. In addition, the small SV allowed for accurate point measurements of small radiation fields to be completed without the partial volume averaging exhibited by larger ion chamber detectors. The device's high spatial resolution and linear SV arrangement also allowed for beam profiles to be mea-

CONF LICT OF I NTEREST
The authors declare no conflict of interest.