• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>6th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2003) Pt.II; Nov 15-18, 2003; Montreal, Canada >Extraction and Application of Expert Priors to Combine Multiple Segmentations of Human Brain Tissue
【24h】

Extraction and Application of Expert Priors to Combine Multiple Segmentations of Human Brain Tissue

机译：结合人类脑组织多种分割的专家先验的提取和应用

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This paper evaluates strategies to combine multiple segmentations of the same image, generated for example by different segmentation methods or by different human experts. Three methods are compared, each estimating and using a different level of prior knowledge about the segmenters. These three methods are: simple label averaging (no priors), a binary expectation maximization (EM) method with independent per-label priors [Warfield et al, MICCAI 2002], and a simultaneous multi-label EM method with across-label priors [Rohlfing et al, IP MI 2003], The EM methods estimate the accuracies of the individual segmentations with respect to the (unknown) ground truth. These estimates, analogous to expert performance parameters, are then applied as weights in the actual combination step. In the case of the multi-label EM method, typical misclassiflcation behavior, caused for example by neighborhood relationships of different tissues, is also modeled. A validation study using the MNI Brain Web phantom shows that decision fusion based on the two EM methods consistently outperforms label averaging. Of the two EM methods, the multi-label technique produced more accurate combined segmentations than the binary method. We conclude that the EM methods are useful to produce more accurate segmentations from several different segmentations of the same image.
机译：本文评估了将同一图像的多个分割相结合的策略，例如通过不同的分割方法或由不同的人类专家生成的分割。比较了三种方法，每种方法都估计并使用了有关分段器的不同级别的先验知识。这三种方法分别是：简单标签平均（无先验），具有独立每个标签先验的二值期望最大化（EM）方法[Warfield等，MICCAI 2002]，以及具有跨标签先验的同时多标签EM方法[ [Rohlfing等人，IP MI 2003]，EM方法估计了与（未知）地面真实性有关的各个细分的准确性。这些估算值类似于专家性能参数，然后在实际组合步骤中用作权重。在多标签EM方法的情况下，还可以对典型的错误分类行为（例如由不同组织的邻域关系引起）进行建模。使用MNI Brain Web幻象进行的验证研究表明，基于两种EM方法的决策融合始终优于标签平均。在两种EM方法中，与二元方法相比，多标签技术可产生更准确的组合分割。我们得出结论，EM方法可用于从同一图像的多个不同分割中产生更准确的分割。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号