| 基于K-Means及K-NN的磁共振颅脑图像分割初探 |
| |
| 引用本文： | 王臣,叶春涛. 基于K-Means及K-NN的磁共振颅脑图像分割初探[J]. 上海医学影像, 2011, 20(1): 9-11,14 |
| |
| 作者姓名： | 王臣 叶春涛 |
| |
| 作者单位： | 复旦大学附属华东医院放射科MRI室,200040 |
| |
| 基金项目： | 复旦大学附属华东医院资助(项目编号H-459) |
| |
| 摘 要： | 目的对K-均值聚类(K-Means)和K-最近邻规则(K-NN)方法在MR颅脑图像分割中的应用进行研究,分析二者优缺点并提出改进。方法利用K-Means算法和K-NN算法对脑组织进行分割。为了减少人脑的劳动使算法自动进行,提出使用K-Means方法提取K-NN方法的训练样本。结果 K-Means与K-NN及改良后的K-NN算法(KMN)能很好地从大脑结构中分割白质(white matter,WM)、灰质(grey matter,GM)和脑脊液(cerebrospinal fluid,CSF)。结论实验表明K-Means与K-NN能快速有效地分割脑组织,改进后的K-NN方法减少了人工参与,并能获得较好的分割效果。
|
| 关 键 词： | 磁共振脑图像 医学图像分割 K-均值聚类 K-最近邻规则 |
The head MRI segmentation based on K-Means and K-Nearest Neighbor rule |
| |
| WANG Chen,YE Chun-tao. The head MRI segmentation based on K-Means and K-Nearest Neighbor rule[J]. Shanghai Medical Imaging, 2011, 20(1): 9-11,14 |
| |
| Authors: | WANG Chen YE Chun-tao |
| |
| Affiliation: | WANG Chen,YE Chun-tao(Department of Radiology,Huadong Hospital Affiliated to Fudan University,Shanghai 200040,China) |
| |
| Abstract: | Objective To study the application of K-Means and K-NN method in the MRI brain image segmentation.To analysis and improve them.Methods This method had three main steps.Firstly,segment the MRI image using K-Means.Secondly,segment the MRI image using K-NN.Finally,proposed the improved K-NN method that using K-Means.Results Using our methods to segment 10 group MRI head clinical data,we got good results that gray matter,white matter and cerebrospinal fluid.Conclusion The experimental results shows that these m... |
| |
| Keywords: | MRI brain image Medical image segmentation K-Means K-NN |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
