Hello,
In my research, we had multiple subject completed 1 task, each subject 2 runs. And now I want to calculated the grand average source across all subjects.
I've already calculated the average source for each run and each subject. And then I average the source of 2 runs within each subjects, so I got 1 average source for 1 subject. After that, I also did Z score normalization for these averaged sources. Next, I projected this normalized average source for each subject onto the template (just as the tutorial Group analysis: Subject coregistration), and got the template sources for all subjects in "Group_Analysis" file.
Then I dragged all the template sources into the "Process"-run process- average-average file- everything. But the result shows nothing and the scale is invalid.
What was wrong with my step? How can I achieve my goal? Is there anyone who can help me?
Hi @xiongruochu,
Rather than the final average, the issue may have its origin in the previous steps. Most likely in the normalization step. See this post:
I recommend you to verify your data after each step of the process to understand what is going on:
- Subject averages
- z-score normalized Subject averages
- Projection on the Default Anatomy of the z-score normalized Subject averages
Hi, Cassani @Raymundo.Cassani
I've checked these steps, it seems ok. I did these steps as follows:
- Subject averages:
1)average all the trials for each run;
2)calculate noise covariance, head model, source(I used minimum 3)norm imaging, sLORETA, constrained)for each run;
average source of 2 runs for each subject (I dragged 2 source file into the Process, average by trial group (subject average))
last I got 1 averaged source file for each subject;
- z-score normalized Subject averages:
1)I dragged the that averaged source file into the Process, run--standardize--baseline normalization--set baseline and select z-score transformation
last I got 1 averaged normalized source file for each subject
- Projection on the Default Anatomy of the z-score normalized Subject averages
1)right click that averaged normalized source file-project sources--default anatomy--cortex_15002V
Then I checked the last projected source file, the scale is normal, and the activated areas were also presented.(figure below)
But after I calculated the projected source file for all subjects and averaged these projected source files, the result is just as what I mentioned above.
I guess something detail went wrong. But I cannot figure it out. I'm always following the tutorials.
If you plot each of the projected (on Default anatomy) z-score per subject. Do they all contain valid z-score values (i.e. the colormap does not show "Invalid scale")?
Yes, the projected sources have valid scale. like this:
Also, I checked the tutorial Workflow once again, I noticed that in the "Constrained cortical sources-- Average: Group analysis" part, it mentions "Rectify the cortical maps (process: Pre-process > Absolute value)", before the projection on a template. Is it because that I haven't rectified my source files?
