基于散斑跟踪技术的超声心动图心动周期估计  

Cardiac cycle estimation of echocardiography with speckle tracking

作　　者：金培培[1] 孙丰荣[1] 刘芳蕾[1] 姚桂华[2] 

机构地区：[1]山东大学信息科学与工程学院,山东济南250100 [2]山东大学齐鲁医院心内科,山东济南250012

出　　处：《山东大学学报（工学版）》2017年第2期94-99,共6页Journal of Shandong University（Engineering Science）

基　　金：山东省自然科学基金资助项目(ZR2014FM006);国家自然科学基金资助项目(81671703);山东省重点研发计划资助项目(2015GS118026)

摘　　要：基于块匹配散斑跟踪技术,提出估计超声心动图心动周期的方法。选择超声心动图心肌壁中感兴趣区域(region of interest,ROI),跟踪和记录每帧超声图像上ROI区域内心肌组织上的散斑或散斑模式的位置和运动。通过计算ROI区域内散斑的位移,用最小绝对误差和(sum of absolute differences,SAD)算法估计心动周期的长度和起点,并用仿真超声序列图像对方法进行验证。试验结果表明超声心动图心动周期的理论值与实际值有较高的一致性。该方法能够有效、可靠地估计心动周期长度和起点。A cardiac cycle estimation method of echocardiography was proposed based on block matching speckle track- ing. The position and movement of speckles or speckle pattern within the area of myocardial tissue on echocardiographic images was tracked and recorded, by selecting the regions of interest （ROI） inside the myocardial wall. By calculating the displacement of speckles in the ROI, the length and the start of cardiac cycle was estimated with the sum of absolute differences （SAD） algorithm, and the method was validated by simulated echocardiographic sequences. The experiment results showed a high consistency between theoretical values and actual ones of the cardiac cycle. This method could es- timate the length and the start of cardiac cycle effectively and reliably.

关 键 词：超声心动图 散斑跟踪 块匹配 菱形搜索 心动周期 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]