Development and Characterization of Low Cost Nanoscintillator-Based Radiation Detection Systems Using 3D Printing Technology

Abstract

The fields of medical health physics, imaging, and radiotherapy have pushed the development and implementation of numerous radiation monitoring systems. Furthermore, detection and measurement of ionizing radiation is essential for many industrial processes. Various detection systems including ion chambers, thermoluminescent detectors, electronic portal imaging devices, semiconductor detectors, and scintillation-based systems have been developed to suit this need. Diagnostic imaging systems most often make use of large arrays of inorganic scintillation crystals. These crystals must be grown using specialized equipment in a laboratory environment. Furthermore, the crystal geometry is limited to relatively small volumes, and production time is on the order of months. Plastic scintillation materials have also been extensively studied for dosimetry applications. These detectors offer high sensitivity with lower production cost and a production timeline on the order of days. Plastic scintillators are most often created by extrusion, casting, and injection molding. These techniques allow for larger volume detectors, but their geometry is still limited in most cases to regular geometric shapes. In recent years, advancements in 3D printing technology have been proposed as alternative manufacturing methods for radiation detectors. These techniques offer the ability for rapid prototyping and allow for at-will creation of complex detector geometries that would otherwise be prohibitively time consuming and expensive using current scintillator manufacturing methods. Furthermore, the wide availability of affordable off-the-shelf consumer 3D printers allows detector manufacturing outside of laboratory environments. The primary focus of this dissertation is the development and characterization of 3D printed radiation detectors using [Y1.903; Eu0.1, Li0.16] scintillating nanoparticles suspended in a printable glycol-modified polyethylene terephthalate (PETG) filament. We assess this technology for use in two applications: (1) as a real-time x-ray imaging screen, and (2) as an inorganic scintillation detector element in a fiber-optic probe dosimeter. (1) The imaging screen was characterized by investigating the accuracy of the scintillation image vs incident exposure patterns, the radiation stability of the detectors, and their ability to differentiate tissue thickness and material density in biologically relevant samples. Scintillation images were captured using a smartphone camera situated outside of the primary x-ray field. A housing apparatus was designed to hold the detector plane perpendicular to the field, and above an optical grade mirror angled 45° relative to the camera. Accuracy of the scintillation image was investigated using cutout-patterned lead masks to attenuate portions of the incident x-ray field. Localization of photons generated in the detector volume was quantified for 5 printed samples using local contrast between adjacent areas of the scintillation image corresponding to shielded and unshielded regions of the detector surface. We calculated the difference in scintillation intensity between these regions of the scintillation image were 7.97 ± 5.4% times higher than measured for the baseline shielded areas. Radiation damage effects on scintillation light output due to prolonged exposures was assessed using 6 detector samples. One detector was used as a control group, while the remaining 5 accumulated absorbed dose using a Cs-137 Irradiator to provide lifetime doses ranging from 1.3 – 14 kGy. The average surface scintillation intensity for each detector was measured relative to the control detector prior to and post irradiation. Relative scintillation intensity showed no discernable change due to the lifetime accumulated dose values investigated. Performance of the detector screen imaging biologically relevant samples was assessed in three stages. Firstly, the ability of the scintillation image to show increased attenuation due to material thickness was demonstrated by imaging a mouse femur. The image showed clear signal difference in thicker portions of the bone, allowing for a pseudo-topological reconstruction of the femur based on pixel gray values in the smartphone camera image. The second stage was demonstrating signal differentiation from attenuation differences due to material density in the range of biological tissues. Tissue-equivalent phantoms representative of lung, breast, soft tissue, brain, 1 year-old bone, and adult bone were used in this study. The phantoms were imaged in groups at various x-ray fields of tube potential from 40-120 kVp. Minimal differences in tissue differentiation were seen across this energy range. Our results suggest the material density threshold for differentiation lies between 0.08 and 0.15 g/cm3. The third stage of the printed detector screen assessment focused on imaging anatomical features of a complete biological sample using a plasticized mouse. Scintillation images were captured corresponding to 120 kVp x-ray projections of 4 regions of the mouse. Specifically, regions centered on the head, neck, torso, and hindquarters were imaged. Radiochromic film was placed on top of the detector plane to provide a comparison x-ray projection image. These scintillation images demonstrated the presence of prominent skeletal structures, and the torso image showed clearly defined lung volumes, a region of increased attenuation representative of the mouse liver, and hints of a gradient of attenuation for overlapping organs of the digestive track. These investigations provide proof-of-principle for the use of 3D printed real-time imaging screens. (2) A fiber-optic probe detector was developed using an aluminum brace to couple 3D printed detector chips to an acrylic light guide in order to funnel scintillation photons into the terminal end of a 0.6 mm diameter optical fiber. The probe detector was fitted with 1 mm thick and 2mm thick detector chips printed at maximum scintillation nanomaterial concentration. Fluorescence spectrometer measurements of these two configurations showed comparable scintillation intensity under 130 kVp x-ray excitation, suggesting that the observed scintillation photons are primarily generated on the surface of the printed detector chip. The probe detector light output was then measured with 1 mm thick chips printed at scintillator loading concentrations of 1, 5, 10, 25, and 35% by weight. Fluorescence spectrometer measurements showed monotonic increase in scintillation intensity vs detector chip concentration. Dose response curves for probe detector fitted with 35% printed chips under 80, 160, and 240 kVp excitation were plotted using a NIST-traceable ion chamber as a gold standard for dose measurement. The detector signal was shown to have a strongly linear relationship to incident dose rate for all three energy x-ray fields. The lower detection limit for 80, 160, and 240 kVp exposures was calculated to be 3.55 ± 0.16 cGy/min, 4.09 ± 0.18 cGy/min, and 4.93 ± 0.22 cGy/min respectively. We conclude that 3D printed scintillation detectors are viable for use in optical fiber dosimetry systems, In addition to investigations into 3D printed radiation detectors, this dissertation also serves to extend the applications and physical characterization of the novel Nano-FOD detection system. This detector makes use of inorganic scintillating nanomaterials coupled with an optical fiber and photodiode to provide real-time dose rate measurements. This work builds on previous characterization studies by implementing a methodology for determining lower detection limits using signal vs dose rate calibration curves. Lower detection limits for 5 Nano-FOD detectors were calculated for 60, 80, 100, 120, 150, 200, and 250 kVp x-ray fields. We observed roughly 30% standard deviation in detection limits among the five sampled Nano-FODs at each energy level measured. In addition to this measurement, we quantified sensitivity variations using dose rate calibration factors for all fibers at each energy level. We also explored the capacity of the Nano-FOD system for in vivo measurement of I-131 in small animal applications. This proof-of-concept study focused on in vitro measurement of 103 mCi of I-131 mixed with 2ml of stabilizing solution inside of a lead shielded glass vial. Two Nano-FOD detectors were used in the investigation, one of which was shielded from β particles via an acrylic sheath. Measurements for each detector were taken over a period of 20 days in order to observe the decay behavior of the Nano-FOD signals. The signal of the shielded fiber was subtracted from the unshielded fiber signal after accounting for differences in diode sensitivity, detector sensitivity, and γ attenuation due to the acrylic sheath. The first two of these correction factors were calculated using data from lower detection limit investigations. The difference in incident γ dose rate on the two detectors due to attenuation was derived computationally using the FLUKA Monte Carlo simulation package to model our experimental geometry. Nano-FOD signal from β- emissions was isolated using this two-fiber subtraction method and shown to decay with a half-life of 7.73 ± 0.31 days. These results demonstrate the viability of the two-fiber subtraction method for I-131 β- dose measurement using the Nano-FOD system.