Improved Multimodality Data Fusion of Late Gadolinium Enhancement MRI to Left Ventricular Voltage Maps in Ventricular Tachycardia Ablation
Improved Multimodality Data Fusion of Late Gadolinium Enhancement MRI to Left Ventricular Voltage Maps in Ventricular Tachycardia Ablation
复制标题
DOI：
10.1109/tbme.2012.2233738
复制
发表时间：
2013-05-01
影响因子：
4.6
通讯作者：
Nezafat, Reza
中科院分区：
文献类型：
--
作者：
Roujol, Sebastien;Basha, Tamer A.;Nezafat, Reza
电解剖电压标测（EAVM）通常在瘢痕相关室性心动过速（VT）导管消融前进行，以定位心肌基质并指导消融手术。EAVM用于定位消融导管的位置，并提供左心室解剖结构和瘢痕的三维重建。然而，EAVM测量仅代表内膜瘢痕，没有透壁或心外膜信息。此外，EAVM是一个耗时的过程，具有高度的操作员依赖性，并且具有低采样密度，即，空间分辨率晚期钆增强（LGE）磁共振成像（MRI）允许无创评估疤痕形态，可以描绘三维疤痕结构。尽管LGE可作为VT消融的路线图用于识别致瘤基质，但其效用非常有限。为了识别VT基质，需要将EAVM识别的基质作为金标准与LGE-MRI瘢痕特征之间的相关性。要做到这一点，必须开发一个系统来融合这些模态的数据集。在这项研究中，提出了一种融合LGE-MRI和EAVM数据的配准管道。提出了一种新的曲面配准算法，将全局疤痕区域的匹配作为配准过程中的一个附加约束。一个预备的标志注册最初进行加速算法的收敛。进行了数值模拟，以评价在识别EAVM或LGE-MRI数据集中的标志时存在错误以及呼吸或心脏运动引起的其他错误的情况下配准的准确性。随后，在接受室性心动过速消融术的10例患者队列中评价了申报融合系统的准确性，其中EAVM和LGE-MRI数据均可用。与界标配准和曲面配准相比，该方法在配准误差上有明显改善。所提出的数据融合系统允许在室性心动过速消融中融合EAVM和LGE-MRI数据，配准误差小于3.5 mm。
Electroanatomical voltage mapping (EAVM) is commonly performed prior to catheter ablation of scar-related ventricular tachycardia (VT) to locate the arrhythmic substrate and to guide the ablation procedure. EAVM is used to locate the position of the ablation catheter and to provide a 3-D reconstruction of left-ventricular anatomy and scar. However, EAVM measurements only represent the endocardial scar with no transmural or epicardial information. Furthermore, EAVM is a time-consuming procedure, with a high operator dependence and has low sampling density, i.e., spatial resolution. Late gadolinium enhancement (LGE) magnetic resonance imaging (MRI) allows noninvasive assessment of scar morphology that can depict 3-D scar architecture. Despite the potential use of LGE as a roadmap for VT ablation for identification of arrhythmogenic substrate, its utility has been very limited. To allow for identification of VT substrate, a correlation is needed between the substrates identified by EAVM as the gold standard and LGE-MRI scar characteristics. To do so, a system must be developed to fuse the datasets from these modalities. In this study, a registration pipeline for the fusion of LGE-MRI and EAVM data is presented. A novel surface registration algorithm is proposed, integrating the matching of global scar areas as an additional constraint in the registration process. A preparatory landmark registration is initially performed to expedite the convergence of the algorithm. Numerical simulations were performed to evaluate the accuracy of the registration in the presence of errors in identifying landmarks in EAVM or LGE-MRI datasets as well as additional errors due to respiratory or cardiac motion. Subsequently, the accuracy of the proposed fusion system was evaluated in a cohort of ten patients undergoing VT ablation where both EAVM and LGE-MRI data were available. Compared to landmark registration and surface registration, the presented method achieved significant improvement in registration error. The proposed data fusion system allows the fusion of EAVM and LGE-MRI data in VT ablation with registration errors less than 3.5 mm.