| 武杰,聂生东,汪红志,张学龙,周滟,许建荣.基于期望值最大化方法的磁共振图像人脑组织分割[J].中国医学影像技术,2007,23(10):1558~1561 |
| 基于期望值最大化方法的磁共振图像人脑组织分割 |
| Brain tissue segmentation of MRI based on expectation maximization method |
| 投稿时间：2007-02-18 修订日期：2007-08-20 |
| DOI： |
| 中文关键词: 期望值最大化 磁共振成像 组织分割 |
| 英文关键词:Expectation maximization Magnetic resonance imaging Tissue segmentation |
| 基金项目:上海市教委重点项目资助(06ZZ33);上海市教委项目资助(06EZ026)。 |
|
| 摘要点击次数: 3843 |
| 全文下载次数: 2298 |
| 中文摘要: |
| 目的 利用期望值最大化方法进行磁共振图像的人脑组织分割。方法 在分析当前常用的医学图像分割方法的基础上,提出一种基于统计理论的期望值最大化分割方法,并给出了相应的理论算法模型和实现步骤,最后用Visual C++ 6.0编程,并对磁共振大脑图像进行实验,并与应用SPM软件对同一幅图像的分割结果进行分析比较。结果 本文分割方法与SPM软件的分割结果非常接近,大脑灰质、白质、脑脊液等组织之间边界清晰,总体不确定性较小。结论 本文分割方法切实可行,分割效果较好,为进一步的磁共振图像分析和疾病研究提供了一种有效工具。 |
| 英文摘要: |
| Objective Using expectation maximization method for brain tissue segmentation of magnetic resonance imaging. Methods After analyzing the medical image segmentation Methods that are widely used now, expectation maximization segmentation method based on statistical theory was presented, and the relevant theoretic arithmetic model and realization protocol were summed up. Using the experiment data for MR brain image, the imaging segment code was carried out with Visual C++ 6.0. Then the segmentation Results were compared between expectation maximization method and SPM software. Results The result of this method is very close to the result of SPM, the boundary of gray matter, white matter and CSF in the brain is distinct, and whole uncertainty is relative low. Conclusion The segmentation technology mentioned in this paper is feasible, and the effect of segmentation is preferable, so it provides an efficient tool to analyze MR images and investigate disease farther. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
