I always use ALL the voxels in the ROI. You don't want to compare a
subset of voxels in one subject to a subset in another subjects, then
you might be getting apples and oranges. This is of course a bigger
issue if you were to use a larger ROI.
The issue of adjustment: Depends not on the voxels, but on what you
told it to adjust for when you did the VOI extraction.
Best Regards, Donald McLaren
=================
D.G. McLaren
University of Wisconsin - Madison
Neuroscience Training Program
Office: (608) 520-0586
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain
PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED
and which is intended only for the use of the individual or entity
named above. If the reader of the e-mail is not the intended recipient
or the employee or agent responsible for delivering it to the intended
recipient, you are hereby notified that you are in possession of
confidential and privileged information. Any unauthorized use,
disclosure, copying or the taking of any action in reliance on the
contents of this information is strictly prohibited and may be
unlawful. If you have received this e-mail unintentionally, please
immediately notify the sender via telephone at (608) 520-0586 or
email.
On Mon, Aug 16, 2010 at 3:29 PM, Benjamin Mullin <[log in to unmask]> wrote:
> Dear Darren and others,
>
> I have a question related to the one listed below. I'm trying to ensure that all of my participants have the exact same seed region for my PPI analysis. I'm intending to use a 4mm radius sphere region that was derived from a separate GLM group analysis (which suggested that one group had increased activity while the other had decreased activity within this region).
>
> When I get to the step of extracting the time series from my seed region, I understand that it doesn't matter which contrast I use (it is only a matter of getting the voxels captured somehow). Is this also true for the statistical threshold used? I had planned on running the GLM as an ROI (using the regional mask I created) with a p threshold of .999. If I understand correctly, when I then extract the time series for that VOI, it will include all of the voxels in the mask. When I've run tests of this, it does result in each participant having the same number of voxels within their seed region. However, I want to make sure that the underlying time series is not being adjusted in some way.
>
> Thanks in advance for any insights. Ben
>
>
>
> Dear Fred
>
> On Tue, Aug 3, 2010 at 9:08 AM, Fred Sanders <[log in to unmask]> wrote:
>
> Dear SPM Experts,
>
> I am trying to run a PPI analysis and want to make sure I am interpreting it the correct way. Using the GLM, I compared two groups (Group A and Group B) using a simple task. I found that with the contrast Interest - Control Condition, Group A had greater activation in my ROI than Group B. I then wanted to run a PPI based on this ROI. However, Group B did not demonstrate any significant activation in the ROI in the contrast Interest - Control Condition. So I ran a contrast with Interest - Fixation. With this contrast, both groups showed activation in my ROI. Here are my questions:
>
> Question 1) To me, it does not back sense to run the PPI with the Interest - Control Condition b/c with this contrast in my ROI Group B does not demonstrate significant activation - is that correct??
>
>
> Correct in the sense that if a group does not show any activation in a region you cannot extract that VOI. As you suggest you can display activations using a different contrast, so that the VOI can be extracted. Note that you should think of the contrast you use to display the activations as simply a window to which voxels will be extracted. SPM goes to the xyz locations of the active voxels in your VOI, gets the actual data from your processed images and then adjusts the data for any effects of interest you specify. The contrast you use to display activations merely identifies the voxels to extract data from it does not affect the data that are extracted.
>
>
>
> Question 2) If Group A has greater activation than Group B in my ROI during the Interest - Fixation condition, how does this affect the PPI when I compare across the groups? My understanding is that the ROI activation serves as a regressor in the PPI. So even if one group has greater activation in the ROI, this should not matter. In other words, a region with a stronger interaction in Group A vs Group B still means the that there is stronger connectivity between the ROI and the region identified in the PPI in Group A. Is that correct?
>
>
> Two groups do not have to show equal activity in the source region in order for the PPI comparison to be valid, just like two groups would not need to have equivalent main effects in order for an interaction to be tested (indeed it is the main effects that are suspect when there is an interaction and not vice versa). The first eigenvariate from the ROI activation is included as part of the PPI design matrix, in order to discount the main effect of the ROI activity.
>
> Regards,
> Darren
>
>
>
> Thanks!!!
> Fred
>
>
>
> --
> Darren Gitelman, MD
> 710 N. Lake Shore Dr., 1122
> Chicago, IL 60611
> Ph: (312) 908-8614
> Fax: (312) 908-5073
>
|