Graduate School of Frontier Sciences, The University of Tokyo, 3F, 178-4-4, Wakashiba, Kashiwa City, 277-0871, Chiba, Japan.
Medical Systems Engineering Division 2, Hitachi Aloka Medical, Ltd., 3-1-1, Higashikoigakubo, Kokubunji, Tokyo, 185-0014, Japan.
Med Image Anal. 2017 Jan;35:554-569. doi: 10.1016/j.media.2016.08.002. Epub 2016 Sep 15.
By tracking echocardiography images more accurately and stably, we can better assess myocardial functions. In this paper, we propose a new tracking method with deformable Regions of Interest (ROIs) aiming at rational pattern matching. For this purpose we defined multiple tracking points for an ROI and regarded these points as nodes in the Meshfree Method to interpolate displacement fields. To avoid unreasonable distortion of the ROI caused by noise and perturbation in echo images, we introduced a stabilization technique based on a nonlinear strain energy function. Examples showed that the combination of our new tracking method and stabilization technique provides competitive and stable tracking.
通过更准确和稳定地追踪超声心动图图像,我们可以更好地评估心肌功能。在本文中,我们提出了一种新的跟踪方法,该方法使用可变形的感兴趣区域 (ROI) 进行合理的模式匹配。为此,我们为 ROI 定义了多个跟踪点,并将这些点视为无网格方法中的节点,以插值位移场。为了避免由于回波图像中的噪声和干扰而导致 ROI 的不合理变形,我们引入了一种基于非线性应变能函数的稳定技术。实例表明,我们的新跟踪方法和稳定技术的结合提供了具有竞争力和稳定的跟踪。