| 胡欣宇,宋颖,谭展,等.单光子发射计算机断层成像技术的演变与发展现状[J].中华放射医学与防护杂志,2026,46(3):322-328.Hu Xinyu,Song Ying,Tan Zhan,et al.Evolution and current development of single-photon emission computed tomography technology[J].Chin J Radiol Med Prot,2026,46(3):322-328 |
| 单光子发射计算机断层成像技术的演变与发展现状 |
| Evolution and current development of single-photon emission computed tomography technology |
| 投稿时间：2025-03-17 |
| DOI：10.3760/cma.j.cn112271-20250317-00090 |
| 中文关键词: 单光子发射计算机断层成像 准直器 半导体探测器 图像重建 |
| 英文关键词:Single-photon emission computed tomography Collimator Semiconductor detector Image reconstruction |
| 基金项目: |
|
| 摘要点击次数: 3303 |
| 全文下载次数: 60 |
| 中文摘要: |
| 单光子发射计算机断层成像(SPECT)作为一种重要的核医学分子影像技术,近年来在系统结构、探测器性能、图像重建算法及多模态集成等方面取得了显著进展。在硬件方面,准直器结构不断优化,从传统平行孔准直器发展至针孔型、会聚型、心脏专用准直器及新兴自准直技术,有效提升了成像灵敏度与空间分辨率;同时,探测器材料也实现了从NaI(Tl)闪烁体向CZT半导体的跨越,提高了能量分辨率和计数效率。在软件层面,图像重建算法经历了从计算效率较高但图像质量受限的滤波反投影(FBP)方法,逐步发展为基于期望最大化原理的有序子集期望最大化(OSEM)迭代算法,显著提升了图像质量。尽管OSEM相较传统迭代算法在重建效率上已有改善,但由于其仍需多次迭代计算,计算负担较大,重建时间相对较长。近年来,随着深度学习方法的引入,进一步提升了图像质量,并在某些重建框架中实现了对计算效率的优化,推动了图像重建性能的整体发展。此外,SPECT图像定量化技术日趋成熟,依托于衰减校正、散射校正及准直器响应建模,可实现体素级放射性活度的准确计算,广泛应用于心肌灌注显像、肿瘤代谢分析等领域。随着SPECT/CT、SPECT/MRI等多模态融合系统的兴起,SPECT正逐步向高分辨率、定量化和个体化精准诊疗方向发展,并展现出广阔的临床应用前景和研究价值。 |
| 英文摘要: |
| Single-photon emission computed tomography (SPECT) is a crucial nuclear medicine molecular imaging technique that has achieved significant progress in recent years in system architecture, detector performance, image reconstruction algorithms, and multimodal integration. In terms of hardware, collimator structures have been continuously optimized. The evolution from traditional parallel-hole collimators to pinhole, converging, cardiac-specific, and emerging self-collimation technologies has effectively enhanced imaging sensitivity and spatial resolution. Concurrently, detector materials have transitioned from NaI(Tl) scintillators to cadmium zinc telluride semiconductors, improving energy resolution and count efficiency. In terms of software, image reconstruction algorithms have progressed from the computationally efficient but image-quality-limited filtered back projection method to iterative algorithms based on the expectation maximization principle, notably the ordered subset expectation maximization algorithm that has significantly improved image quality. Although the ordered subset expectation maximization algorithm offers enhanced reconstruction efficiency compared to traditional iterative algorithms, it still requires multiple iterative computations, resulting in substantial computational burden and relatively long reconstruction times. In recent years, the introduction of deep learning method has further improved image quality and, in certain reconstruction frameworks, optimized computational efficiency, thereby driving the overall advancement of image reconstruction performance. Furthermore, SPECT image quantification techniques have matured. Relying on attenuation correction, scatter correction, and collimator response modeling, these techniques enable accurate voxel-level quantification of radioactivity and are widely applied in fields such as myocardial perfusion imaging and tumor metabolic analysis. With the emergence of multimodal fusion systems such as SPECT/CT and SPECT/magnetic resonance imaging, SPECT is progressively advancing towards high-resolution, quantitative, and personalized precision diagnosis and treatment, demonstrating broad clinical application prospects and substantial research value. |
| HTML 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|