Feasibility Study on Hybrid Medical Imaging Device Based on Compton Imaging and Magnetic Resonance Imaging

Abstract

Objective(s) There are many attempts to integrate between positron emission tomography (PET) and magnetic resonance imaging (MRI). Even though MRI provides very high-resolution anatomical images, generally it does not provide functional or metabolic information in the human body. Combining anatomical information with functional/metabolic information can offer better clinical care to the patient. There is, however, a challenge in the development of PET/MRI; that is, MR-compatible PET detectors should be first developed against the high magnetic field environment, which is not easy at the current state of technologies. Compton imaging could address the problem by using semiconductor detectors. Furthermore, Compton imaging has 3-dimensional imaging capability at a fixed position; hence, it does not need a ring-type structure, which makes it relatively easy to fit into an existing MRI system. The multi-tracing capability of Compton imaging also allows us to use several radiopharmaceuticals at the same time. In this study, we proposed a hybrid medical imaging device that combines Compton and MR imaging modalities. Material & Method(s) The proposed system consists of four Compton camera modules inserted into a cylindrical magnet. Each module consists of four double-sided silicon strip detectors (DSSDs) as the scatterers and four pixellated CdZnTe (CZT) detectors as the absorbers. The DSSDs have a dimension of 5 cm × 5 cm × 0.15 cm with 16 orthogonal strips on each side. The CZT detectors, each of which is composed of 256 pixels, have a dimension of 2.5 cm × 2.5 cm × 0.5 cm. The Compton camera modules were placed at 30-degree intervals in order to take a form of curvature. The imaging performance of the proposed system was determined by Geant4 Monte Carlo simulations. For this simulation, this study used a tomographic head phantom, which was taken from a Korean voxel model (HDRK-Man). Result(s) The simulated Compton images were then reconstructed by using a list-mode expectation maximization (LM-EM) algorithm and merged on the MR images. Conclusion(s) In this presentation, we report the performance of the proposed hybrid imaging system and discuss the feasibility of obtaining the anatomical images and the functional images simultaneously.