扩散张量图像的插值方法综述  

A Review on Interpolation Methods for Diffusion Tensor Images

作　　者：蒋帆 王远军[1] JIANG Fan;WANG Yuan-jun(Institute of Medical Imaging Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学医学影像工程研究所

出　　处：《波谱学杂志》2019年第3期392-407,共16页Chinese Journal of Magnetic Resonance

基　　金：国家自然科学基金资助项目(61201067);上海市自然科学基金资助项目(18ZR1426900)

摘　　要：扩散张量成像能够通过获得组织内水分子的三维位移分布来研究脑部结构,是近年来医学成像技术的研究热点.在采集、转换以及处理张量的进程中,研究者需要进行插值处理来提高图像分辨率或改善可视化效果.如在人脑模板构建、脑白质纤维追踪以及配准等应用中,张量插值是一个具有重要作用的处理步骤.本文对现有的张量插值方法进行综述,首先介绍了插值方法的理论内容,阐明各张量插值方法中所解决的技术问题及存在的局限性,然后介绍了常用的插值方法的评估指标,再利用现有的典型插值方法分别对仿真数据和真实数据进行插值实验和结果分析,最后对张量插值方法的未来发展趋势提出建议.Diffusion tensor imaging measures three-dimensional distribution of water molecule displacement in tissues,and has been widely used to study the structure of the brain.The technique has attracted wide research interests in recent years.Interpolation is often needed in the acquisition and reconstruction of diffusion tensor images to improve spatial resolution and/or facilitate visualization,especially when the images are used for white matter fiber tracking and registration in human brain.In this paper,the existing tensor interpolation methods are reviewed.The theory underlying the interpolation methods is described first.The technical challenges and limitations of the currently-used methods are then reviewed,followed by the introduction to the evaluation indices of the interpolation methods.The performance of typical interpolation methods is then compared with simulated data and real experimental data.The future direction of the development of tensor interpolation method is suggested.

关 键 词：扩散张量成像 张量插值 欧拉角 四元数 部分各向异性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] O482.53[自动化与计算机技术—计算机科学与技术]