인체 내 양성자 빔 모니터링을 위한 감마 버텍스 영상 장치 개발

Abstract

In proton therapy, highly conformal dose can be delivered to the target volume while minimizing the dose to adjacent normal tissues and critical organs as compared with the conventional radiotherapy using x-ray or electron beam. However, the proton dose distribution, especially the beam range, might deviate from the planned one due to dose calculation errors, organ motions during the treatment, anatomical changes, or patient setup errors. The uncertainty of the proton dose distribution necessitates an additional safety margin around the target volume, thus it restricts the degree of freedom of proton treatment. In order to fully utilize advantages of proton therapy and guarantee the patient safety, a means of monitoring in vivo proton dose in real-time is required. For this purpose, a prompt gamma imaging, which measures prompt gammas which have a close correlation with the proton beam range, was suggested, and the possibility for proton beam monitoring was experimentally demonstrated. For the clinical application, however, an optimized detection system for high energy gamma-ray imaging should be developed. The present dissertation suggests a new emission imaging method for high energy gamma-rays, gamma electron vertex imaging (GEVI), and experimentally demonstrates the feasibility for proton beam monitoring. To determine the emission position, in GEVI method, an incident gamma-ray is converted to an electron by Compton scattering, and then the trajectory and energy of the converted electron are measured by two hodoscopes and a calorimeter, respectively. To configure the imaging system, dedicated signal processing systems for component detectors were developed. To reduce data acquisition channels for a double-sided silicon strip detector (DSSD, hodoscope), a multiplexing based low-noise signal processing system was designed and constructed. For a plastic scintillation detector (calorimeter), an in-house amplifier module was used. Using these developed systems, characteristics of component detectors were estimated in terms of an energy resolution, a timing resolution, and a noise level. To experimentally demonstrate the proposed imaging principle, a proof-of-principle imaging system was constructed. From imaging experiments with gamma-ray sources and prompt gammas generated by proton nuclear interactions, it was confirmed that the GEVI method has a great potential for high energy gamma-ray imaging. Based on experimental results, a prototype GEVI imaging system was developed and tested for therapeutic proton beams to see the feasibility for proton beam monitoring. For the quantitative analysis, a 50% distal falloff position of the prompt gamma distribution was determined by using a methodology based on a 3rd-order polynomial fitting. For 80, 120, and 150 MeV proton beams, imaging sensitivities (= effective events per each primary proton) were 6.91 × 10^-7, 9.04 × 10^-7, and 1.15 × 10^-6, and the falloff position having a linear relation with the beam range was determined with the uncertainty of 0.91, 0.86 and 0.72 mm for 6.24 × 10^9 protons. Next, proton beams of 126 126, 132, 138, 144, 150, 156, 162, and 168 MeV were delivered to the solid water phantom, and prompt gammas were imaged by the imaging system which was fixed at 150 mm depth of the phantom. With experiment results, it was confirmed that the proton beam range can be verified with the distal falloff position of the measured distribution. In addition, prompt gamma distributions were imaged by changing the incident beam position, and we confirmed that the developed prototype system can monitor the proton beam with lateral (= depth) movement within 10 cm and vertical movement within ± 3 cm. In the present study, the GEVI imaging system was successfully developed, and the imaging performance was experimentally estimated. The developed imaging system can monitor the beam delivery position in the phantom and the beam range by using the correlation with the distal falloff position. We expects that in vivo proton beam monitoring can be performed using the developed imaging system. In addition, developed key techniques of a low noise signal processing, a multi-channel data acquisition, and an image reconstruction can be also utilized to develop various radiation imaging systems.