Criterion for sign change in relative plots


I have noticed that Brainstorm output for a scout as a relative plot was different from the output I obtain with Matlab by simply averaging the vertices. I have understood that this inconsistency was linked to the invasion of the sign of one vertex (1 out of 18). However, when this vertex was part of a smaller scout (just two vertices), the sign did not change. Hence, I think that the sign changed "relative" to the other vertices, but what is the criterion? how are the other vertices taken into account?

Thanks a lot for any answer or comment.

Hi Christian,

Yes, the sign flip depends on the other sources in the cluster:

This is described in:

"We start by finding the dominant orientation of the scout, then flip the sign of the values that are not in the same direction (scalar product of the orientations < 0)."

The dominant orientation of the scout is obtained with SVD:


Thanks for the answer, that's clear! If possible, I would add this aspect in the description of the relative plot. Best, Christian