Need Help Interpreting PSD Output Files, Please

Discussions
1

Hi,

I just pre-processed two 10 minute-long segments of raw EEG data, then ran a Power Spectral Density analysis (Welch's). I would like some help unpacking the results, please:
I've attached the two .CSV output files below. In these output files, I'd like clarification as to what these values represent. Are they the Power (i.e. microvolts squared / Hz), the Logarithm of the Power or some other value altogether? Also, they are very low values in the negative range...is this to be expected or is it abnormal? In other words, do these values resemble the expected power values for these frequency bands in say a baseline, resting state?

Thank you for your help!
Chanda
2-BL_try02_clean data_psd_180526_1143.csv (273 Bytes)
2-MU1_try02_clean data_psd_180526_1237.csv (273 Bytes)

2

Answer from Sylvain by email:

HI Mona Lisa:

I am not sure how you obtained the .csv files: I suppose this is after exporting with Brainstorm. Are you using the grouping by frequency bands for computing the PSD ? I guess yes, the displays only show a few data points. That’s fine too.

Your best best would be to read the Brainstorm Matlab files, but if you don’t have Matlab, I understand the .csv option is best.

Now, the values you are looking at are power estimates from the PSD. They are small because they scale to about the square of the typical electrode voltage, which should be in the realm of microvolts. They are actually positive as 1.583022267E-12 means 1.58 x 10^-12.

Hope this clarifies,

Sylvain.