| 孔维丹,朱朝喆,邹启红,左西年,严超赣,臧玉峰.采用分水岭分割方法获取功能磁共振激活图感兴趣区[J].中国医学影像技术,2008,24(9):1476~1479 |
| 采用分水岭分割方法获取功能磁共振激活图感兴趣区 |
| Acquiration of regions of interest (ROIs) with watershed algorithm on fMRI activation map |
| 投稿时间：2008-03-14 修订日期：2008-06-02 |
| DOI： |
| 中文关键词: 磁共振成像 感兴趣区 分水岭算法 功能分割 |
| 英文关键词:Magnetic resonance imaging Region of interest Watershed algorithm Functional segmentation |
| 基金项目: |
|
| 摘要点击次数: 3226 |
| 全文下载次数: 787 |
| 中文摘要: |
| 目的 为功能磁共振中的功能连接研究提供一种更为合适的感兴趣区(ROI)获取方法。方法 20名正常被试者完成视觉任务扫描,对视觉任务组水平激活图利用分水岭算法进行功能分割。结果 分水岭方法可得到单峰的、空间上互相独立的激活区。结论 分水岭方法可以将较大的、有多个激活峰值的功能团块分割为空间上相互独立的单峰脑区,克服了单纯采用激活程度阈值来产生ROI的方法之缺点,为下一步的功能连接分析提供了一种选择ROI的方法。 |
| 英文摘要: |
| Objective Watershed segmentation algorithm was applied to fMRI activation map to help selection of regions of interest (ROIs) for functional connectivity (FC) study. Methods Twenty normal subjects underwent two fMRI sessions: resting-state and visual response task. Watershed algorithm was used on the group activation map to generate single-peak functional regions. Five ROIs were selected, and FC analysis was performed between every two regions. Paired t-test was performed to investigate the FC differences between resting and task conditions. Results Single-peak and spatially isolated activated areas were obtained by watershed segmentation. Significant FC differences were found between resting and task conditions. Conclusion Watershed approach can be used in functional image segmentation. Large and connected areas with several activated peaks can be isolated into single-peak regions. It may help selection of ROIs for FC analysis. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
