Error calculating the statistics for coherence maps

Hi,

I am working on EEG based resting state data for patients and controls and was analyzing the functional connectivity between specific scouts using coherence as a measure (NxN). (Source localization: sLORETA, headmodel: BEM(volume))

However, when I am trying to run a permutation test (independent) for statistical analysis between controls (File A) and patients (FileB) for the coherence maps, it is showing me an error stating:

  • Cannot select rows for connectivity (NxN) results.
  • No data read from Files A.

Could you please let me know how I can resolve this issue?

Thank you so much!

Best,
Ami

Hi Ami,

The issue relies because one or more Scouts were indicated in the Signal names or indices field of the Permutation test (independent). This is not a valid option as the coherence [NxN] file which was computed for N scouts, has one row and one column associated with the given label.

To have a coherence [1xN] file, 1 Scout vs N Scouts, you need to use Process 2 tab and place the same files in the A and B side, then run the process Coherence AxB [2021] and select the 1 Scout for A, and the N Scouts for B (scouts should be from the same Atlas). Then you can run the permutation test.

Best,
Raymundo

Hi Raymundo,

Thank you so much for your help.

Best,
Ami

Hi Raymundo,

As you had mentioned I calculated the coherence using the AxB process and then ran the permutation test. But I noticed that if I run the test for the subject averaged coherence file, then the resulting tmap file shows me an invalid scale. But if I do not average the coherence files per subject, and directly run the permutation test for all controls vs. patients, then the results show a valid tmap and p-values. So I wanted to confirm if this is correct or not?
Also, is there a way I could get the matlab script for the permutation test (unequal)?

Thank you so much.

Best,
Ami

Hi Raymundo,

Is there any feedback regarding this discussion?
Thank you so much for your help.

Best,
Ami

The “Invalid scale” label is shown because it is not possible to find a scale for the data. This is likely caused by no p values under the p-threshold, thus there is not data to show, so no scale is possible to find.

Check these posts

Thank you so much for your help.

Best,
Ami