请教:关于头动的组间差异分析

      关于头动参数的组间差异分析, 想请教严老师,之前投出去的一篇文章评审专家要求对头动参数进行组间差异分析并作为协变量进行分析,严老师,您能大概说一下怎样进行这个组间差异分析吗?谢谢!

 在严超赣老师的第三部分最新的视频里面讲到一些评估头动参数的方法,用DPARSFA勾选上选项之后就会自动生成相应的参数,进而做统计分析。如果简单一些的话,这篇文献 

Abnormal Intrinsic Brain Activity Patterns in Patients with Subcortical Ischemic Vascular Dementia

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0087880里面有一些描述,可以参考。另外严超赣老师这两篇文献Standardizing the intrinsic brain: towards robust measurement of inter-individual variation in 1000 functional connectomes. 及A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics.有详细论述。

      谢谢老师。还要请教老师一个问题,评审专家提的意见中有一个地方不明白,就是“ the root mean square of their motion parameters”这个是什么意思,为什么还有平方差?下面是评审专家的意见。
     “ Given the importance of differences in movement parameters which can confound resting state BOLD studies even after covarying movement parameters (work by the Power group), the authors should report the root mean square of their motion parameters and compare this between groups as well as run this as a covariate at the second level as a between-subjects variable.”

Power et al.的文章我没有仔细读过,但我记得那篇文章采用了frame wise density,什么计算总的头动情况,在Zang et al., 2007 Brain Dev有详细的公式,即每个时间点的头动幅度减去前一个时间点的头动幅度,取绝对值,如果有200个时间点,就会得到199个绝对值,将这些绝对值相加,6个头动的时间序列,得到6个总的头动的幅度。

root mean square (RMS) 与standard devination (SD) 相似。但我不确定Power的文章中是否采用。它反映的头动的fluctuation情况。frame wise density也反映了头动的fluctuation,但反映的是高频fluctuation;而RMS或者SD反映不出是高频还是低频的fluctuation;最近有人计算头动时间序列的ALFF,用它反映头动时间序列的低频fluctuation。