研究目的
To develop and prototype a high-sensitivity, high-resolution, and low-cost dedicated brain PET system using a novel hemispherical detector arrangement with add-on detectors to improve sensitivity and uniformity for early diagnosis of dementia and neuroscience studies.
研究成果
The developed helmet-chin and helmet-neck PET prototypes demonstrated high sensitivity (exceeding 10% at the top brain region), uniform spatial resolution of 3-4 mm, and the ability to resolve small structures (e.g., 2.2 mm rods). The neck position for add-on detectors provided higher sensitivity and wider FOV compared to the chin position. The systems offer a compact, low-cost alternative to whole-body PET scanners with potential for improved brain imaging in dementia diagnosis and neuroscience. Future work should focus on motion suppression, automated registration, and further clinical validation.
研究不足
The prototype system is a CT-less design, requiring separate acquisition of attenuation maps, which may introduce registration errors. The study used a limited number of detectors (54), which is fewer than commercial systems, potentially affecting overall performance. Head motion during clinical tests caused blurring in images, and no motion compensation methods were implemented. The system's compact size and chair-style setup may not be suitable for all patients or clinical environments. Further optimization is needed for automated registration and motion correction.
1：Experimental Design and Method Selection:
The study involved designing and building a prototype PET system with a hemispherical detector arrangement using 4-layer depth-of-interaction (DOI) detectors. The system included add-on detectors at the chin or neck positions to enhance sensitivity. Image reconstruction methods such as filtered back-projection (FBP) and iterative list-mode OP-OSEM algorithms were employed, incorporating detector response function modeling, normalization, attenuation correction, scatter correction, and random correction.
2：Sample Selection and Data Sources:
Phantoms (e.g., double cylinder pool phantoms, small hot-rod phantom, cylindrical pool phantom, brain phantom) and a healthy human volunteer were used for performance evaluation and clinical testing. Radioactive tracers like 18F-FDG were injected for imaging.
3：List of Experimental Equipment and Materials:
Equipment included 4-layer DOI detectors with Zr-doped GSO crystals (2.8×2.8×7.5 mm3), high-sensitivity 64-channel flat-panel photomultiplier tubes (R10551-00-64, Hamamatsu Photonics), tungsten end shields, data acquisition system, and various phantoms made of materials like PMMA. The prototype system had a gantry adjustable for sitting patient style.
4：8×8×5 mm3), high-sensitivity 64-channel flat-panel photomultiplier tubes (R10551-00-64, Hamamatsu Photonics), tungsten end shields, data acquisition system, and various phantoms made of materials like PMMA. The prototype system had a gantry adjustable for sitting patient style.
Experimental Procedures and Operational Workflow:
4. Experimental Procedures and Operational Workflow: The prototype was assembled with detectors arranged in a hemisphere shape. Measurements involved placing point sources, phantoms, and human subjects in the system, acquiring data with specified energy and time windows, and reconstructing images using FBP and iterative methods. Clinical tests were conducted with low-dose 18F-FDG injections and separate CT scans for attenuation correction.
5：Data Analysis Methods:
Data were analyzed for spatial resolution (FWHM), sensitivity, count rate performance, and image quality using statistical methods and software tools like in-house registration software and CUDA for GPU-accelerated computations.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
