The Effects of Computational Method, Data Modeling, and TR on Effective Connectivity Results.
The Effects of Computational Method, Data Modeling, and TR on Effective Connectivity Results.
复制标题
DOI：
10.1007/s11682-009-9064-5
复制
发表时间：
2009-06
影响因子：
3.2
通讯作者：
Meyerand, M. Elizabeth
中科院分区：
文献类型：
--
作者：
Witt, Suzanne T.;Meyerand, M. Elizabeth
关键词：
随着有效连通性的使用变得越来越普遍，重要的是了解不同分析的结果如何相互比较，因为采用不同方法确定连通性的研究结果可能不会得出相同的结论。模拟的fMRI时间序列数据被用来比较结构方程模型、自回归分析、格兰杰因果关系和动态因果模型这四种更常用的计算方法的结果，以确定哪种方法可能更适合这项任务。结果表明，这三种方法都能检测到系统动力学的变化。结构方程模型似乎对tr或方差源的变化最不敏感，而格兰杰因果关系最敏感。结果还表明，改进数据分析的报告是必要的，使用效果统计来描述结果可能会消除使用不同方法确定连接性的跨研究结果比较的一些模糊性。
As the use of effective connectivity as become more popular, it is important to understand how the results from different analyses compare with each other, as the results from studies employing differing methods for determining connectivity may not reach the same conclusion. Simulated fMRI time series data were used to compare the results from four of the more commonly used computational methods, structural equation modeling, autoregressive analysis, Granger causality, and dynamic causal modeling to determine which may be better suited to the task. The results show that all three methods are able to detect changes in system dynamics. Structural equation modeling appeared to be the least sensitive to changes in TR or source of variance, and Granger causality the most sensitive. The results also suggest that improved reporting on data analyses is necessary, and employing an effect statistic to depict results may remove some of the ambiguity in comparing results across studies using differing methods to determine connectivity.
登录
查看更多内容
影响因子：
5.7
作者：
Bullmore, ET;Horwitz, B;Sharma, T
通讯作者：
Sharma, T
影响因子：
3.7
作者：
GEWEKE, JF
通讯作者：
GEWEKE, JF
影响因子：
4.8
作者：
Hampson, M;Peterson, BS;Gore, JC
通讯作者：
Gore, JC
影响因子：
16.2
作者：
Bokde, ALW;Tagamets, MA;Horwitz, B
通讯作者：
Horwitz, B
影响因子：
5.7
作者：
Horwitz, B
通讯作者：
Horwitz, B