摘要： 基于互信息的配准方法具有自动化程度高、配准精度高等优点，近来已成为医学图像处理领域的热点。基于互信息的配准方法实质上是一种对灰度进行统计和计算的方法，同一图像采用不同的窗宽窗位表示必然会影响配准结果。该文在分析窗宽窗位对图像质量的影响和基于互信息配准方法的影响的基础上，进行一系列的医学图像配准试验。在分析配准结果的基础上，给出基于互信息的配准方法所采用的合理窗宽窗位的建议。

关键词: 医学图像配准, 互信息, 窗宽窗位

Abstract: Registration methods based on mutual information are popular in medical image processing fields, which are accepted as the most accurate and automated methods. Essentially, registration methods based on mutual information are based on doing statistics of gray values of images. Therefore, if different representations of window widths and window levels of one image are used, registration accuracy will be affected. Based on analysis of window widths and window levels for image quality, and based on the affects of mutual information registration method, series of medical images registration tests are carried out. According to these results and analysis, the recommended window widths and window levels to be used in mutual information registration are given.

Key words: medical image registration, mutual information, window width and window level

中图分类号: 

• TN911.73