医学图像配准综述

Review of Image Registration Methods for Medical Images

医学领域,医生需要比较不同的解剖影像信息和功能影像信息。将对应于同一物体的不同影像的像素点对应起来就是医学图像配准的主要任务。鉴于需要配准的图像的采集的方法和时间等方面不同,配准所采用的方法也不同。本文介绍了医学图像配准的新近方法。首先从图像配准的任务出发,介绍了配准过程的主要涉及的方面和配准流程。然后从配准时变换的参数角度考虑,将图像的配准分成两大类考虑:参数配准和非参数配准。每一类中给出基本的配准方法,并讨论其用法和特点。对参数配准,主要讨论它们各自不同的方法和适合处理的图像对;对非参数配准,侧重于考虑控制方程的由来和整体外力的计算。

Purpose： Review different methods in medical image registration.Method： Parametric and non-parametric registration methods. Result： Various registration methods and applications are presented. Conclusion： Different registration methods fit to different kinds of medical images and result differen applications in medical field. In medical field, researchers need to combine and compare the information given by functional and anatomical images. For the reason, image registration which can map points from one image to homologous points on an object h the second image is one of the fundamental tasks in image processing. Different numerical methods are introduced into registration due to the different images taken from different perspectives, times and devices. In this paper, we present an overview of recent methods for image registration in medical field. We start by the goal of image registration and introduce the framework of a registration process. The registration methods under review are classified into two categories： parametric image registration and non-parametric image registration. Typical methods in each category are presented. In each case, we provide the readers with general background information and properties or explanation of how these methods work.