An off-line gating method for suppressing motion artifacts in ICUSsequence
An off-line gating method for suppressing motion artifacts in ICUSsequence
复制标题
DOI：
10.1016/j.compbiomed.2010.09.006
复制
发表时间：
2010-11
影响因子：
7.7
通讯作者：
Sun Zheng;Yan Qi
中科院分区：
文献类型：
--
作者：
Sun Zheng;Yan Qi
冠状动脉内超声（ICUS）成像是一种基于导管的有创成像方式，已广泛应用于冠状动脉疾病的临床诊断。由心脏周期性运动和血液脉动引起的运动伪影是阻碍ICUS纵向切割可视化和从ICUS图像序列进行血管三维重建的主要因素。本工作的目的是提出一种基于ICUS图像特性的离线门控方法，以抑制在体内采集的连续拉回ICUS图像序列的运动伪影，该图像序列覆盖多个心动周期，而无需ECG信号的帮助。在分析每帧ICUS信号强度特征的基础上，首先构造差分矩阵。然后，从矩阵中提取潜在的心动周期信息，通过在矩阵中寻找一条累积差值最小的全局最优路径，得到各ICUS帧在连续心动周期中的对应关系。最后，选择每个在每个心动周期的结束时间采集的帧的子序列以获得门控序列。该方法的主要优点是，它适用于在图像采集期间受试者心率可能不规则的那些ICUS序列。此外，不需要对ICUS图像执行任何高级图像分析。结果给出了患者的临床图像数据，以验证所提出的方法。
Intracoronary ultrasound (ICUS) imaging, an invasive catheter-based imaging modality, has been widely used in clinical diagnosis of coronary artery diseases. Motion artifacts due to cyclic cardiac motion and pulsatile blood are the main factors that hinder the visualization of ICUS longitudinal cut and 3D vessel reconstruction from ICUS image sequence. The aim of this work is to propose an off-line gating method based on ICUS image properties to suppress motion artifacts for continuous pullback ICUS image sequences acquired in vivo covering several cardiac cycles without the help of ECG signals. Based on analysis of intensity features of each ICUS frame, a difference matrix is firstly constructed. Then, potential information about cardiac cycle is extracted from the matrix and the correspondence of each ICUS frame in successive cardiac cycles is obtained through finding a globally optimal path with the lowest cumulative difference value in the matrix. Finally, a subsequence of frames each acquired at the end-diastole of each cardiac cycle is selected to obtain the gated sequence. The main advantage of this method is that it is applicable to those ICUS sequences where the subject’s heart rate may be irregular during image acquisition. Also, any high-level image analysis performed to ICUS images is not needed. Results are given for clinical image data of patients in order to validate the proposed method.