高分辨率小动物Micro Insert PET系统的模拟研究

在西门子Micro PET小动物成像系统中接入高分辨率的Insert探测器环,组成了Micro Insert PET系统,可以进一步提高小动物PET系统的图像分辨率。本文采用蒙特卡罗模拟软件GATE定义Focus-220小动物PET和Micro Insert两种系统,在同步仿真运算中,给出了三种类型的符合数据,即PET Scanner和PET Scanner之间的符合、PET Scanner和Insert之间的符合、Insert和Insert之间的符合。用滤波反投影算法重建了Na-22单点源模型、多点源模型、LZU模型的PET图像。其中,多点源模型的径向图像分辨率的仿真结果与美国华盛顿大学的Micro Insert实验数据,在忽略正电子射程、非线性效应项以及噪声等的贡献下,两者符合得很好。研究表明,使用Micro Insert系统能够将小动物PET系统的图像分辨率提高到亚毫米水平。     

多通道平行束准直器灵敏度补偿性能

点源扩展函数PSF作为SPECT成像的一个重要参考,反映出相应准直器灵敏度和空间分辨率等的特性。本文在多通道平行束准直器MPB的几何框架下进行物理模型构建,解释了由探测器晶体厚度引起的PSF变化以及基于该变化进行的公式修正,指出了不利的因素,其结果与蒙特卡罗模拟基本吻合。在PSF基础上,结合蒙特卡洛模拟,指出了在准直器的拉伸过程中,斜平行束部分OPB的灵敏度变化特性和对准直器的补偿作用;通过两点源逐渐接近的过程,并与高灵敏度平行束准直器HSPB对比,指出MPB准直器在保持高灵敏度下,还能有高空间分辨率的原理,提出了一种设计高灵敏度和高空间分辨率准直器的方法。     

基于蒙特卡罗模拟3-DPET图像重建的研究

应用蒙特卡罗模拟软件GATE对GeDiscoveryLSPET/CT扫描系统进行模拟,获得三维投影数据,分别采用三维有序子集最大期望值法（3-DOSEM）与三维滤波反投影法（3-DFBP）进行图像重建,从空间分辨率、成像的对比度以及信噪比等方面对3-DOSEM法与3-DFBP法的重建结果进行比较。结果显示,对蒙特卡罗方法模拟获得的三维投影数据,采用3-DOSEM法重建比3-DFBP法重建得到的图像质量好。     

高分辨率小动物SPECT成像系统的设计与性能评估

为实现低成本、高分辨率的小动物单光子发射断层成像(SPECT),设计了一种基于临床探测器和针孔准直器的小动物SPECT成像系统。综合考虑视野尺寸、空间分辨率和绝对探测效率等指标,分别给出了单针孔和七针孔准直器的两种最优化设计方案,并通过对比两种设计方案在全视野范围内的平均空间分辨率和探测效率,对成像性能进行了评估。结果表明,与单孔准直器相比采用七针孔准直器设计可同时获得更高的空间分辨率和探测效率。经蒙特卡罗模拟成像实验验证,该七针孔准直器在临床探测器上可达0.5mm级的超高空间分辨率,且成本远低于采用专用探测器的现有小动物SPECT系统。     

凸度测量过程中点扩展函数的蒙特卡罗模拟与解析表达研究

散射校正是高精度凸度测量中的技术难点。本文采用点扩展函数(PSF)理想模型估计的方法,建立了点扩展函数的蒙特卡罗(MC)计算模型,分析了PSF随射线能量和钢板厚度的变化关系,并提出用多个单能的线性组合来代替连续能谱计算的方法,取得了较好的效果。采用双高斯函数模型,结合蒙特卡罗模拟的计算结果,对凸度测量系统的单能点扩展函数进行了解析求解,并给出0.18MeV射线下,凸度测量系统PSF的解析公式,为凸度测量系统的散射校正奠定了基础。     

HPGe探测器点源效率与其空间位置的函数关系

本文采用蒙特卡罗(MCNP)模拟与实验相结合的方法,计算了HPGe探测器(美国Ortec公司产品,相对效率40%,对1332keV的分辨率为1.8keV)对空间不同位置、不同能量点源的探测效率,对Masayasu等[1]提出的点源径向表达式作了修正,改进了与实验值(或MC模拟值)的一致性;给出了HPGe探测器效率随点源位置变化的解析表达式;证明了应用MCNP模拟方法确定探测器对于点源的探测效率函数的可靠性和实用性,也有助于对体源效率的测量和计算。     

多平行束准直器模拟研究

在平行束准直器中引用多通道技术设计了三通道的多平行束（Multi-parallel Beam,MPB）准直器。计算机模拟表明,MPB准直器在保持较高的空间分辨率下,有效地改善了系统的灵敏度。放射药物匀均分布的圆柱数字模具和Jaszczak数字模具（phantom）模拟表明,随重叠比例的增加,未出现重建环状伪影;虽然靠近准直器表面空间分辨率有所下降,但在FOV（Field of View）中心区域空间分辨率近似等同于平行束准直器,重建图像的匀均度有了改善。投影数据的重叠表明,MPB的投影有放大作用。该研究为高成像质量的临床SPECT成像提供一种新的准直器设计方法。     

基于SART算法的CL硬化伪影校正方法研究

X射线薄板层析成像(CL)系统是一种针对板状构件的新型断层成像系统,该系统中存在射束硬化现象。本工作研究了一种基于SART算法的硬化伪影校正算法（SART-BHC算法）,通过对该算法的研究及对CL系统矩阵的计算,使该方法能应用于CL中。通过模拟与实验相结合的方式,利用蒙特卡罗方法获得CL投影的模拟数据,在实际CL中获得实验数据,然后利用SART-BHC与FPB算法对两种数据分别重建。结果表明,SART-BHC算法可应用于CL中,不需要任何先验知识,能很好地校正硬化伪影,并对层间混叠有抑制作用。     

高分辨micro-SPECT成像系统针孔准直器的优化设计

为了最大限度地减少γ射线在针孔准直器上的透射和散射效应,采用蒙特卡罗方法对单光子发射计算机断层(SPECT)成像系统使用的针孔准直器进行优化.模拟计算结果与实验测试得到的系统灵敏度和空间分辨率曲线相符,达到了高分辨小动物成像的要求.     

CT代数重建算法的快速实现

代数重建算法比解析重建算法在重建图像质量上有很大优势,但代数重建算法计算量较大,因而使其在实际应用中受到许多限制.针对代数重建算法中系数矩阵计算占绝大部分重建时间的情况,提出了一种将系数矩阵预先计算,并将系数矩阵计算结果存储于硬盘的方法.该方法大大提高了代数重建算法速度,并能够使用计算复杂的精确投影系数矩阵,从而可提高...     

A high-resolution small animal SPECT system developed at Tsinghua

An SPECT system dedicated to small animal imaging shall be of a millimeter spatial resolution or even better.This study was aimed at achieving 0.5-mm spatial resolution for a small animal SPECT system at low cost.It was developed from a single-head clinical SPECT scanner,with a seven-pinhole collimator and a four-degree-offreedom motion control stage.Several key techniques were developed,including high-resolution image reconstruction algorithm,high accuracy geometrical calibration method,and optimized system matrix derivation scheme.The system matrix was derived from Monte-Carlo simulation and de-noised by fitting each point spread function to a two dimensional Gaussian function.Experiments of point source and ultra micro hot rod phantom were conducted.With a spatial resolution of 0.5-0.6 mm,this system provides a practical way for low-cost high-resolution animal imaging on a clinic SPECT system.     

高分辨针孔SPECT三维图像重建的实现

采用一种新颖的查找表法来实施有序子集期望值最大化（OSEM）的针孔SPECT图像重建算法,有效地将针孔响应修正包含在查找表中,进而实现高分辨针孔SPECT的三维图像重建。标准Jaszczak模具的成像实验结果表明：该方法不但能显著减少图像重建过程的计算时间和内存占用,且能显著改善重建图像的分辨率。因此,查找表法是能够快速实现针孔SPECTOSEM图像重建算法,并能将针孔响应修正包含在查找表中以获取高分辨断层图像的有效方法。     

Phantom experiments on a PSAPD-based compact gamma camera with submillimeter spatial resolution for small animal SPECT

We demonstrate a position sensitive avalanche photodiode (PSAPD) based compact gamma camera for the application of small animal single photon emission computed tomography (SPECT). The silicon PSAPD with a two-dimensional resistive layer and four readout channels is implemented as a gamma ray detector to record the energy and position of radiation events from a radionuclide source. A 2 mm thick monolithic CsI:Tl scintillator is optically coupled to a PSAPD with a 8mm×8mm active area, providing submillimeter intrinsic spatial resolution, high energy resolution (16% full-width half maximum at 140 keV) and high gain. A mouse heart phantom filled with an aqueous solution of 370 MBq (99m)Tc-pertechnetate (140 keV) was imaged using the PSAPD detector module and a tungsten knife-edge pinhole collimator with a 0.5 mm diameter aperture. The PSAPD detector module was cooled with cold nitrogen gas to suppress dark current shot noise. For each projection image of the mouse heart phantom, a rotated diagonal readout algorithm was used to calculate the position of radiation events and correct for pincushion distortion. The reconstructed image of the mouse heart phantom demonstrated reproducible image quality with submillimeter spatial resolution (0.7 mm), showing the feasibility of using the compact PSAPD-based gamma camera for a small animal SPECT system.     

CT图像重建方法在半影成像中的应用

CT图像重建方法和半影成像是两种传统的辐射图像诊断方式。将CT图像重建方法应用到半影图像数据处理中,探索出一种新的半影图像处理方法。在新的半影编码孔图像数据处理中,首次应用了与图像诊断系统点扩散函数无关的编码孔图像处理方法,解决了传统编码孔图像处理中由于图像诊断系统点扩散函数不确定带来的技术难题。在理论研究的基础上进行了半影模拟成像及图像重建,获得了理想的研究结果。     

Characterizing the MTF in 3D for a Quantized SPECT Camera Having Arbitrary Trajectories

The emergence of application-specific 3D tomographic small animal and dedicated breast imaging systems has stimulated the development of simple methods to quantify the spatial resolution or Modulation Transfer Function (MTF) of the system in three dimensions. Locally determined MTFs, obtained from line source measurements at specific locations, can characterize spatial variations in the system resolution and can help correct for such variations. In this study, a method is described to measure the MTF in 3D for a compact SPECT system that uses a 16 × 20 cm(2) CZT-based compact gamma camera and 3D positioning gantry capable of moving in different trajectories. Image data are acquired for a novel phantom consisting of three radioactivity-filled capillary tubes, positioned nearly orthogonally to each other. These images provide simultaneous measurements of the local MTF along three dimensions of the reconstructed imaged volume. The usefulness of this approach is shown by characterizing the MTF at different locations in the reconstructed imaged 3D volume using various (1) energy windows; (2) iterative reconstruction parameters including number of iterations, voxel size, and number of projection views; (3) simple and complex 3D orbital trajectories including simple vertical axis of rotation, simple tilt, complex circle-plus-arc, and complex sinusoids projected onto a hemisphere; and (4) object shapes in the camera's field of view. Results indicate that the method using the novel phantom can provide information on spatial resolution effects caused by system design, sampling, energy windows, reconstruction parameters, novel 3D orbital trajectories, and object shapes. Based on these measurements that are useful for dedicated tomographic breast imaging, it was shown that there were small variations in the MTF in 3D for various energy windows and reconstruction parameters. However, complex trajectories that uniformly sample the breast volume of interest were quantitatively shown to have slightly better spatial resolution performance than more simple orbits.     

全覆盖型动物PET系统探测器灵敏度的研究

为了提高动物PET系统探测器的灵敏度,增加探测器几何立体角的覆盖比例。在原有的Mi-cro PET系统环形探测器两端,对称添加两块中间有孔的平板探测器构成了紧密型动物PET系统。研究紧密型动物PET系统灵敏度的方法为:基于Gate程序建立系统模型,在点源、面源和体源三种类型的放射源条件下进行Monte Carlo模拟。模拟结果表明:物理结构上的改进有效地提高了探测器灵敏度,使符合事件数为传统的Micro PET的4.48倍;同时点源和面源的模拟结果说明,紧密型探测器结构使得整个探测区域空间均匀性变好,有利于数据采集和图像重建;而平板探测器中间的孔削弱了对探测器灵敏度的改善,尤其是对于放置在探测区域边缘的放射源。应用该全覆盖型结构有望进一步提高PET系统的探测灵敏度并有助于提高图像质量。     

A complete implementation of list-mode reconstruction for PET

Tomography reconstruction algorithm is one of the key components of positron emission tomography (PET) scanners,most PET scanners use statistical iterative reconstruction algorithms from data in sinograms currently. However tomography reconstruction using list-mode data has many unique advantages,in recent years great attention has been paid to it,being in the process of rapid development and improvement.In this paper,using experimental data of small animal PET scanner Eplus-166,exploiting ordinary subsetized list-mode EM(S-LMEM) algorithm and orthogonal distance-based ray-tracer(OD-RT),we eventually achieve list-mode tomography reconstruction.System response matrix(SRM),which establishes mapping relationship between the image and the projection space,is one key problem in iterative reconstruction algorithm.OD-RT is based on an optimization Siddon’s algorithm to calculate the SRM,generating line-of-response(LOR) which is approximately Gaussian-shaped,achieving better modeling of detector response function(DRF).The results demonstrate that image resolution recovery achieves the inherent properties of the scanner and that on-the-fly ray-tracer for real-time calculation of system response matrix is feasible for dynamic reconstruction.Meanwhile,the optimal parameters for calculating SRM are found by experiments.     

Image reconstruction from Compton scattering data

A new image reconstruction method was developed for a Compton camera. A simulation to determine a γ-ray source position was performed by using the simulation tool, GEANT4. An image reconstruction was made in two steps. First, a three dimensional image was constructed and projected in one selected plane, then the points from each ellipse was picked up by taking the peak points of a density distribution of crossing points between the ellipse and the first step image. The second step procedure improved the accuracy and the spatial resolution of a position determination significantly, comparing with the image obtained by only the first step. The accuracy and the resolution for a point source were obtained to be about 0.02 mm and (1.35 0.15) mm, respectively. The same procedure was applied to an imaging of the distributed y-ray source.     

How crystal configuration affects the position detection accuracy in pixelated molecular SPECT imaging systems?

It is well known that inter-crystal scattering and penetration(ICS-P) are major spatial resolution limiting parameters in dedicated SPECT scanners with pixelated crystal.In this study,the effect of ICS-P on crystal identification in different crystal configurations was evaluated using GATE Monte Carlo simulation.A ~(99m)Tc pencil-beam toward central crystal element was utilized.Beam incident angle was assumed to vary from 0° to 45° in 5° steps.The effects of various crystal configurations such as pixel-size,pixel-gap,and crystal material were studied.The influence of photon energy on the crystal identification(CI) was also investigated.Position detection accuracy(PDA) was defined as a factor indicating performance of the crystal.Furthermore,a set of ~(99m)Tc point-source simulations was performed in order to calculate peak-to-valley(PVR) ratio for each configuration.The results show that the CsI(Na)manifests higher PDA than NaI(TI) and YAP(Ce).In addition,as the incident angle increases,the crystal becomes less accurate in positioning of the events.Beyond a crystal-dependent critical angle,the PDA monotonically reduces.The PDA reaches 0.44 for the CsI(Na) at 45° beam angle.The PDAs obtained by the point-source evaluation also behave the same as for the pencil-beam irradiations.In addition,the PVRs derived from flood images linearly correlate their corresponding PDAs.In conclusion,quantitative assessment of ICS-P is mandatory for scanner design and modeling the system matrix during iterative reconstruction algorithms for the purpose of resolution modeling in ultra-high-resolution SPECT.