| Joint segmentation and motion estimation of left ventricle with an anisotropic biomechanical model |
| |
| 作者姓名： | ZHUANG Ling LIU Huafeng BAO Hujun |
| |
| 作者单位： | 1. State Key Laboratory of CAD & CG, Zhejiang University, Hangzhou 310027, China; 2. State Key Laboratory of Modern Optical Instrumentation, Zhejiang University, Hangzhou 310058, China |
| |
| 基金项目： | Supported by the National Program on Key Basic Research Project (2003CB716104)，National Natural Science Foundation of China (Grant No 60403040 and 60021201) |
| |
| 摘 要： | Accurate estimation of shape and motion variability of the heart has important clinical implications for heart disease diagnosis. In this paper, we describe a framework which performs simultaneous segmentation and motion analysis of the left ventricle from 3D cardiac images. The left ventricle is formulated as an anisotropic linear elastic body represented by a set of tetrahedrons constructed from sampling nodes. The elastic body is evolved by both internal force and external force, and the external force imposed on each node is an integration of edginess measure, shape coherence measure, prior distributions of the myocardial tissues and temporal driven measure. The displacement field and the segmentation result are simultaneously solved at the equilibrium state between the internal energy and the external energy. The experiment with 3D canine MR images illustrated its abilities and benefits.
|
| 关 键 词： | left ventricle motion simultaneous estimation deformable model image segmentation. |
Joint segmentation and motion estimation of left ventricle with an anisotropic biomechanical model |
| |
| ZHUANG Ling,LIU Huafeng,BAO Hujun.Joint segmentation and motion estimation of left ventricle with an anisotropic biomechanical model[J].Progress in Natural Science,2007,17(2):226-229. |
| |
| Authors: | ZHUANG Ling LIU Huafeng BAO Hujun |
| |
| Institution: | 1. State Key Laboratory of CAD & CG, Zhejiang University, Hangzhou 310027, China; 2. State Key Laboratory of Modern Optical Instrumentation, Zhejiang University, Hangzhou 310058, China |
| |
| Abstract: | Accurate estimation of shape and motion variability of the heart has important clinical implications for heart disease diagnosis. In this paper, we describe a framework which performs simultaneous segmentation and motion analysis of the left ventricle from 3D cardiac images. The left ventricle is formulated as an anisotropic linear elastic body represented by a set of tetrahedrons constructed from sampling nodes. The elastic body is evolved by both internal force and external force, and the external force imposed on each node is an integration of edginess measure, shape coherence measure, prior distributions of the myocardial tissues and temporal driven measure. The displacement field and the segmentation result are simultaneously solved at the equilibrium state between the internal energy and the external energy. The experiment with 3D canine MR images illustrated its abilities and benefits. |
| |
| Keywords: | left ventricle motion simultaneous estimation deformable model image segmentation |
|
| 点击此处可从《自然科学进展(英文版)》浏览原始摘要信息 |
| 点击此处可从《自然科学进展(英文版)》下载免费的PDF全文 |
