Three-dimensional registration of myocardial perfusion SPECT and CT coronary angiography
Three-dimensional registration of myocardial perfusion SPECT and CT coronary angiography
复制标题
DOI：
10.1007/bf02984607
复制
发表时间：
2005-05
影响因子：
2.6
通讯作者：
H. Nakajo;S. Kumita;Keiichi Cho;T. Kumazaki
中科院分区：
文献类型：
--
作者：
H. Nakajo;S. Kumita;Keiichi Cho;T. Kumazaki
在这项研究中，我们描述了一种新的三维配准技术的CT冠状动脉造影（CTCA）与门控心肌灌注SPECT. Methods 12例已知或怀疑CAD谁接受CTCA与门控SPECT回顾性研究。使用CTCA数据手动追踪冠状动脉及其分支，并进行三维重建。门控SPECT数据注册和映射到一个左心室的二进制模型提取的CTCA数据使用手动，刚性和非刚性registration methods.ResultsThree-dimensional重建和体积可视化的两种模式，成功地实现了所有患者。所有3种配准方法的视觉质量均较好，非刚性配准的结果显著优于其他配准方法（p < 0.05）。使用非刚性配准的三维配准的成本函数（235.3 ± 13.9）显著优于手动和刚性配准的成本函数（分别为218.5 ± 15.3和223.7 ± 17.0）。观察员间的再现性误差是在可接受的范围内的所有方法，并有没有显着差异的methods.ConclusionThis技术的图像配准可以帮助整合的信息从门控SPECT和CTCA，并可能有临床应用缺血性心脏病的诊断。
ObjectiveIn this study, we describe a new technique for three-dimensional registration of CT coronary angiography (CTCA) and gated myocardial perfusion SPECT.MethodsTwelve patients with known or suspected CAD who underwent CTCA and gated SPECT were enrolled retrospectively. Coronary arteries and their branches were traced using CTCA data manually and reconstructed in three-dimensions. Gated SPECT data were registered and mapped to a left ventricle binary model extracted from CTCA data using manual, rigid and nonrigid registration methods.ResultsThree-dimensional reconstruction and volume visualization of both modalities were successfully achieved for all patients. All 3 registration methods gave better quality based on visual inspection, and nonrigid registration gave significantly better results than the other registration methods (p < 0.05). The cost function for three-dimensional registration using nonrigid registration (235.3 ± 13.9) was significantly better than those of manual and rigid registration (218.5 ± 15.3 and 223.7 ± 17.0, respectively). Inter-observer reproducibility error was within acceptable limits for all methods, and there were no significant difference among the methods.ConclusionThis technique of image registration may assist the integration of information from gated SPECT and CTCA, and may have clinical application for the diagnosis of ischemic heart disease.