Interpreting amplitude envelope correlation (AEC) values

[maykooti1]

Hello,
I am using amplitude envelope correlation (AEC) to estimate MEG connectivity. I need some help understanding the output of AEC, which is resulting in a value between -1 and 1. How should I interpret the negative values? Is it similar to the Pearson correlation? in the Pearson correlation, the closer the value is to 1 or –1, the stronger the linear correlation and a value of 0 denotes no linear correlation

Thanks,
Mayadeh

[Francois]

This process "Amplitude Envelope Correlation" has been deprecated in June.
Please update Brainstorm and use the process "Envelope correlation" instead.
You won't have negative values with this one.
https://neuroimage.usc.edu/brainstorm/Tutorials/Connectivity#Envelope_correlation

[VictorRodriguez]

Hi @Francois,

I have a question, just out of curiosity. Why did you remove negative values? It is just from a "biological" point of view where negative coupligs do not make sense?

Thank you!

[Francois]

This is discussed also in these two topics:

• How is the orthogonalization symmetric in the "bst_henv.m" function? - #6 by Raymundo.Cassani
  => This topic concludes that there be can negative values, while in fact, the output values are positive only.
• Hilbert envelope correlation about assuming measures positive in bst_henv.m
  => This one never received any answer

@hossein27en @Sylvain @leahy @Raymundo.Cassani
Some text to explain the positive-only values should probably be added to the documentation for this method: https://neuroimage.usc.edu/brainstorm/Tutorials/Connectivity#Envelope_correlation

[Francois]

@hossein27en @Sylvain @leahy @Raymundo.Cassani @Marc.Lalancette Reminder

[Sylvain]

This implementation is aligned with those of MNE-Python's and Fieldtrip's. We will be looking into alternatives as soon as we have addressed other priorities in our development plan.