4D micro-CT for cardiac and perfusion applications with view under sampling.
4D micro-CT for cardiac and perfusion applications with view under sampling.
复制标题
DOI：
10.1088/0031-9155/56/11/011
复制
发表时间：
2011-06-07
影响因子：
3.5
通讯作者：
Johnson GA
中科院分区：
文献类型：
--
作者：
Badea CT;Johnston SM;Qi Y;Johnson GA
微型CT目前用于临床前研究中提供解剖学信息。然而，人们对在相同系统上获得功能测量非常感兴趣。我们在此报告4DMicro-CT的新策略，着重于两个应用：i)基于回溯性门控的心脏成像和ii)使用多次对比剂注射/旋转范例的肺灌注。双源微型CT系统用于脉冲X射线曝光(80kVp，70 mA，10ms)和20投影/秒的采样率下的图像采集。心脏Micro-CT方案包括使用脂质体血池造影剂，该造影剂含有123 mg I/ml，通过尾静脉导管以0.01ml/g体重的剂量输送。使用回溯门控实现了对自由呼吸小鼠的快速扫描。心电和呼吸信号被用来将投影分类为10个心脏时相。肺灌注方案使用常规造影剂(IsoVue 370)。造影剂的总体积被限制在0.4毫升，并由微型注射器分4次注射，间隔2分钟，以允许清除。每一次注射都与动物的旋转同步，4次旋转中的每一次都以与前一次旋转的起始角22.5°的角度偏移开始。当使用传统的滤波反投影算法时，心脏和灌注方案都会导致投影角的不规则分布，从而在重建中造成明显的条纹伪影。为了减少这些伪影，我们构造了我们的Micro-CT系统的点扩展函数(PSF)，并分析了重建数据在傅立叶域中的分布。这使我们能够通过反卷积来校正不规则的角度不一致，并识别数据缺失的区域。这些区域用来自用所有可用投影重建的高质量但时间平均的图像的数据填充。我们在图形处理器(GPU)上实现了该算法的部分功能，以减少运行时间。利用数值模拟和微型CT实验数据对该方法进行了验证。仿真结果表明，反卷积方法在保留时间信息的同时，成功地去除了条纹伪影。在各向同性体素大小为88微米、时间分辨率为10ms的条件下，进行了小鼠心脏微CT扫描，图像质量良好。以176微米和687毫秒的时间分辨率获得了小鼠的4D肺血流灌注图像。与FBP重建相比，去卷积图像的条纹降低率为70%，对比噪声比提高了2.5倍。与所提出的方法相关的辐射剂量在典型的微型CT剂量范围内(心脏检查0.17Gy.灌注研究0.21Gy.本文介绍的低剂量4DMicro-CT成像方案可应用于高通量纵向研究，应用范围广泛，包括药物安全性和心肺表型。
Micro-CT is currently used to provide anatomical information in preclinical studies. There is however significant interest in obtaining functional measurements on the same systems. We report here novel strategies for 4D Micro-CT with a focus on two applications: i) cardiac imaging based on retrospective gating and ii) pulmonary perfusion using multiple contrast injections/rotations paradigm. A dual source micro-CT system is used for image acquisition with pulsed x-rays exposures (80 kVp, 70mA, 10 ms) and a sampling rate of 20 projections/sec. The cardiac micro-CT protocol involves the use of a liposomal blood pool contrast agent containing 123 mg I/ml delivered via a tail vein catheter in a dose of 0.01 ml/g body weight. Fast scanning of free breathing mice is achieved using retrospective gating. The ECG and respiratory signals are used to sort projections into 10 cardiac phases. The pulmonary perfusion protocol uses a conventional contrast agent (Isovue 370). The total volume of contrast agent is limited to 0.4 ml and is delivered by a micro-injector in 4 injections separated by 2 minute intervals to allow for clearance. Each injection is synchronized with the rotation of the animal, and each of the 4 rotations is started with an angular offset of 22.5° from the starting angle of the previous rotation. Both cardiac and perfusion protocols result in an irregular angular distribution of projections that causes significant streaking artifacts in reconstructions when using traditional filtered back projection algorithms. To alleviate these artifacts, we construct the point spread function (PSF) of our micro-CT system and analyze the distribution of the reconstructed data in the Fourier domain. This enables us to correct for irregular angular inconsistencies via deconvolution and also identify regions where data is missing. These regions are filled in with data from a high quality but temporally averaged image reconstructed with all available projections. We implemented parts of this algorithm on a graphics processing unit (GPU) to reduce run time. The proposed method was validated with numerical simulations and experimental micro-CT data. The simulations indicate that deconvolution successfully removes the streaking artifacts while preserving temporal information. 4D cardiac micro-CT in a mouse was performed with adequate image quality at isotropic voxel size of 88 microns and 10 ms temporal resolution. 4D pulmonary perfusion images were obtained in a mouse at 176 microns and 687 msec temporal resolution. Compared with FBP reconstruction, the streak reduction ratio is 70% and the contrast to noise ratio is 2.5 times greater in the deconvolved images. The radiation dose associated with the proposed methods is in the range of a typical micro-CT dose (0.17 Gy for the cardiac study and 0.21 Gy for the perfusion study). The low dose 4D micro-CT imaging protocols presented here can be applied in high throughput longitudinal studies in a wide range of applications, including drug safety and cardiopulmonary phenotyping.
登录
查看更多内容
影响因子：
3.8
作者：
Johnston, Samuel M.;Johnson, G. Allan;Badeaa, Cristian T.
通讯作者：
Badeaa, Cristian T.
影响因子：
6.7
作者：
Drangova, Maria;Ford, Nancy L.;Holdsworth, David W.
通讯作者：
Holdsworth, David W.
影响因子：
3.1
作者：
Carlson, Stephanie K.;Classic, Kelly L.;Russell, Stephen J.
通讯作者：
Russell, Stephen J.
影响因子：
3.8
作者：
Badea, Cristian T.;Hedlund, Laurence W.;Johnson, G. Allan
通讯作者：
Johnson, G. Allan
影响因子：
1.9
作者：
DHAWAN, AP;RANGAYYAN, RM;GORDON, R
通讯作者：
GORDON, R