Quantification of Patterns on Different y-scales?

[cmwill69]

As shown in the attached figure 1, there are some very apparent recurrent patterns from 600ms to 1000ms, but on different scales, so from what I understand, they are not captured in the cross-recurrence plot as a result. Would you know of any methods that would allow me to capture the pattern?
I have realized that I could simply take the derivative to capture the mentioned pattern since it has almost the same trajectory throughout, but it seems like it makes the time series almost too chaotic and somewhat destroys the leader-follower dynamics shown in the original plot, which are the point of interest for CRQA in my paper. There is still an apparent leader throughout the trial, but it would be nice if there was a cleaner method of doing this. The derivative's CRP is attached below as figure 2. I am currently limiting the leader-follower analysis to 250ms around the point in time.