Positive values in ERD analysis


I conducted ERD analysis using Brainstorm. Most data I got were negative, however a few data were positive (the ERD was less than 20%). ERD should be negative. I have no idea why I got the positive values (too much noise?) Could you please tell me how to deal with the positive values? Thank you!


Does it mean that you have computed time-frequency maps and normalized them with the process "Standardize > Baseline normalization" with the option "Event-related perturbation (ERS/ERD)" ?

The measure is easy to understand: (values - mean_over_baseline) / mean_over_baseline * 100.
If you have positive values, it means you have higher values than in average in your baseline.

Make sure you only use this normalization method on values that are strictly positives (like power or magnitude of time-frequency maps).

Thank you, Francois!

  1. Yes, I used the option.
  2. The data I processed were from patients, so the data were very noisy. Could you please tell me how to deal with the values that are not strictly positive? and is it correct to obtain the positive values in ERD analysis?
    Much appreciated!

Positive values = values that are higher than the average over the baseline.
Negative values = values that are lower than the average over the baseline.
The interpretation depends on your hypotheses and your understanding of the experiment.