Search

Brain positron emission tomography (PET) plays an important role in the study of human brain function for biomedical. In order to significantly improve the quality of brain PET imaging, increase detection efficiency, and display images in real time, a specialized large-area panel detection system for brain PET imaging is designed, and preliminary experimental tests are carried out for this design. The PET detection system consists of two large-area probes, each containing a 3×3 array of position-sensitive photomultiplier tubes and multiple tightly arranged array-type Cerium and Yttrium Doped Lutetium Oxyorthosilicate (Ce:LYSO) scintillation crystals. A 22Na point source with 1 μCi was used to test the crystal properties of LYSO crystal arrays with pixels sized from 0.7 mm×0.7 mm to 2.0 mm×2.0 mm. And two LYSO crystal arrays with pixels of 1.5 mm×1.5 mm were selected for preliminary tomography of one-to-one coincidence detection. The experimental results show that the average FWHMs of position discriminating profiles of LYSO crystal arrays with various pixels are between 0.38 mm and 0.49 mm, and the average values of energy resolution (FWHM) are between 20% and 30%. In the initial image tomography test of 22Na point source, the resolution of reconstructed image is obtained as 1.5 mm by filtering back projection algorithm, which verifies the feasibility of large-area panel detection system.

To study the PET imaging performance of cross-strip cadmium zinc telluride detector with a size of 16.4 mm×16.4 mm×5 mm, which can determine the 3-D position information of 511 keV gamma photons. The 32 electrodes of the detector are all read out using charge sensitive preamplifiers and filter amplification modules, and then input high-speed analog-to-digital converters to sample the data, and obtain the full pulse shape of the signal in the detector, and use the ADC built-in FPGA to perform the online signal processing, digital processing such as peak finding and filtering. The energy resolution of the CZT detector for 511 keV gamma photons is 6.35%. According to the electrode structure of the CZT detector, 3-D position coordinate calculation formulas for the interaction of gamma photons can be obtained. The CZT detector realizes a 3-D imaging function with a sub-millimeter resolution.

In-beam PET imaging provides a noninvasive method to monitor the dose distribution in real-time for particle therapy such as carbon ion therapy and proton therapy. This work designed in-beam PET prototypes based on free PET including cylindrical, open-PET, C-shaped, dual-head plate and rectangular PET scanners and simulated in-beam PET imaging based on these scanners in GATE, a Monte Carlo simulation platform. The scanners were composed of 32 detector modules through free combination and had similar field of view, which could be used for dose monitoring in carbon therapy for the different part of human body and small animal. Each detector module was composed of LYSO array coupled with SiPM. The crystal array contained 20×20 crystals measuring 1.5 mm × 1.5 mm × 10 mm with 1.6 mm pitch. A PMMA phantom was irradiated by the periodic carbon ion pencil beam with an energy of 230 MeV/u and was monitored by the in-beam PET imaging using GATE macros. PET images of the generated positrons were reconstructed by using a MLEM algorithm. Results showed that the shift of image’s peak recreated the variation of positron activity, making it feasible for range verification in the carbon ion therapy by use of free PET scanners. This work compared the reconstructed images between the experimental measurement and simulation based on the dual-head plate scanner containing 8 detector modules. Both peak positions of experimental and simulated images were at 27 mm depth, which validated the feasibility of free PET prototypes for range verification and dose monitoring in carbon ion therapy.

To obtain accurate radiotherapy for carbon therapy, it is critical to accurately monitor the three-dimensional (3D) dose distribution of patients. This paper designs a double-layer Compton imaging system and a simple back-projection algorithm. Using the Geant4 toolkit, the structure of the Compton imaging system is optimized, the detection efficiency and imaging performance of the Compton imaging system are evaluated, and the 3D dose distribution of 200 MeV/μ carbon ion beam bombarding PMMA target is analyzed. In addition, four pairs of LYSO crystals with 1.5 mm×1.5 mm×10.0 mm and 12×12 pixels were used to perform Compton imaging experiments on a 22Na point source with a diameter of about 3 mm. Due to the deviation of the reconstructed dose distribution, factors affecting the precision of 3D dose monitoring are also analysed. For the 0.847 MeV γ point source, the reconstructed image has an FWHM increased to 2.38 mm on the cross-section and 7.02 mm on the coronal plane. The compton imaging system detects the 4.439 MeV prompt γ-rays generated by carbon ion beam bombardment of PMMA target, and the deviation of the reconstructed 3D dose distribution from the real 3D dose distribution is 9.3%. Using the 22Na point source imaging experiment of the LYSO Compton imaging system prototype, a reconstructed image with a FWHM of 4.05 mm was obtained, which verified the practicability of the compton imaging method.