计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (22): 244-248.

• 工程与应用 • 上一篇    

基于可形变模型的左心室形状恢复与运动跟踪

高秀梅1,金 忠2,王元全3   

  1. 1.淮阴师范学院 计算机科学系,江苏 淮阴 223001
    2.南京理工大学 计算机科学技术学院,南京 210094
    3.北京理工大学 计算机科学技术学院,北京 100081
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-08-01 发布日期:2007-08-01
  • 通讯作者: 高秀梅

Deformable model based shape recovery and motion tracking of left ventricle

GAO Xiu-mei1,JIN Zhong2,WANG Yuan-quan3   

  1. 1.Department of Computer Science,Huaiyin Teachers College,Huaiyin,Jiangsu 223001,China
    2.School of Computer Science and Technology,Nanjing University of Science and Technology,Nanjing 210094,China
    3.School of Computer Science and Technology,Beijing Institute of Technology,Beijing 100081,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-08-01 Published:2007-08-01
  • Contact: GAO Xiu-mei

摘要: 从心脏核磁共振图像恢复左心室的3D形状并跟踪其运动,是当前医学图像分析领域的重要研究课题之一。提出了一种基于可形变模型的左心室3D形状恢复与运动跟踪方法,该模型充分考虑了左心室的实际几何形状,将左心室描述为一个一端封闭、非轴对称、轴线弯曲的广义圆柱壳;模型的参数为函数,用全局参数即可捕捉左心室的局部形变;模型的求解纳入基于物理学的可形变模型框架之下,这样,左心室的形状恢复与运动跟踪变成一个动态的数据拟合过程。根据跟踪的结果可以观察左心室的容积、质量等在一个心动周期中的变化情况,计算每博容积、射血分数等心功能指标,为心脏疾病诊断提供直观的临床参考。

关键词: 核磁共振图像, 形状恢复, 运动跟踪, 可形变模型, 左心室

Abstract: Recovering the shape and motion of the left ventricle from cardiac Magnetic Resonance Images(MRI) is a primary
topic in the community of medical image analysis.A new deformable model,which makes use of the priori knowledge of the left ventricle’s geometry sufficiently,for analyzing the shape and motion of the left ventricle,is presented.This new model is a generalized cylinder which is one-end closed,not axial symmetry and axis-curved.Parameters of the proposed model are functions which can capture the global and local deformation merely from global parameters.Using a physically-based approach,geometric models are converted into dynamic models,which deform under the forces computed from the data points and conform to the given dataset.From the tracked volumetric shape,one can monitor the evolution of the volume and mass of the left ventricle during one heart cycle and some clinically established indexes such as stroke volume and ejection fraction can be calculated; these measures are important for clinical use.

Key words: magnetic resonance images, shape recovery, motion tracking, deformable model, left ventricle