Veress Alexander I, Weiss Jeffrey A, Huesman Ronald H, Reutter Bryan W, Taylor Scott E, Sitek Arek, Feng Bing, Yang Yongfeng, Gullberg Grant T
Department of Bioengineering, The Scientific Computing and Imaging Institute, University of Utah, Salt Lake City, UT, USA.
Ann Biomed Eng. 2008 Jul;36(7):1104-17. doi: 10.1007/s10439-008-9497-9. Epub 2008 Apr 24.
The objective of this research was to assess applicability of a technique known as hyperelastic warping for the measurement of local strains in the left ventricle (LV) directly from microPET image data sets. The technique uses differences in image intensities between template (reference) and target (loaded) image data sets to generate a body force that deforms a finite element (FE) representation of the template so that it registers with the target images. For validation, the template image was defined as the end-systolic microPET image data set from a Wistar Kyoto (WKY) rat. The target image was created by mapping the template image using the deformation results obtained from a FE model of diastolic filling. Regression analysis revealed highly significant correlations between the simulated forward FE solution and image derived warping predictions for fiber stretch (R (2) = 0.96), circumferential strain (R (2) = 0.96), radial strain (R (2) = 0.93), and longitudinal strain (R (2) = 0.76) (p < 0.001 for all cases). The technology was applied to microPET image data of two spontaneously hypertensive rats (SHR) and a WKY control. Regional analysis revealed that, the lateral freewall in the SHR subjects showed the greatest deformation compared with the other wall segments. This work indicates that warping can accurately predict the strain distributions during diastole from the analysis of microPET data sets.
本研究的目的是评估一种称为超弹性变形的技术直接从微型正电子发射断层扫描（microPET）图像数据集测量左心室（LV）局部应变的适用性。该技术利用模板（参考）图像数据集和目标（加载）图像数据集之间的图像强度差异来生成体力，使模板的有限元（FE）表示发生变形，从而与目标图像配准。为了进行验证，模板图像定义为来自Wistar Kyoto（WKY）大鼠的收缩末期microPET图像数据集。通过使用舒张期充盈有限元模型获得的变形结果对模板图像进行映射来创建目标图像。回归分析显示，模拟的正向有限元解与图像衍生的变形预测在纤维拉伸（R（2）=0.96）、圆周应变（R（2）=0.96）、径向应变（R（2）=0.93）和纵向应变（R（2）=0.76）方面具有高度显著的相关性（所有情况p<0.001）。该技术应用于两只自发性高血压大鼠（SHR）和一只WKY对照的microPET图像数据。区域分析显示，与其他壁段相比，SHR受试者的外侧游离壁变形最大。这项工作表明，通过对microPET数据集的分析，变形可以准确预测舒张期的应变分布。
