Why is there a difference between pwelch PSD & power of frequency bands?

Hi!

I have calculated the PSD of my EEG data using (I have 55 channels):

Fs = EEG.srate;
nfft = 2^nextpow2(699880);
 
for i = 1:55
    [pxx(i,:), freqs] = pwelch(EEG.data(i,:), hanning(699880), [], nfft, Fs);
end
 
mean_pxx = mean(pxx,1);

And I have gotten the attatched screenshot. I then wanted to calculate the power of each frequency band using the following code which yieleded the second attatched screenshot.

deltaIdx = find(freqs>1 & freqs<4);
thetaIdx = find(freqs>4 & freqs<8);
alphaIdx = find(freqs>8 & freqs<12);
betaIdx = find(freqs>13 & freqs<30);
gammaIdx = find(freqs>30 & freqs<70);

deltaPower = mean(mean_pxx(deltaIdx));
thetaPower = mean(mean_pxx(thetaIdx));
alphaPower = mean(mean_pxx(alphaIdx));
betaPower = mean(mean_pxx(betaIdx));
gammaPower = mean(mean_pxx(gammaIdx));

powers = [deltaPower thetaPower alphaPower betaPower gammaPower];
f = [1 2 3 4 5];
plot(f,powers, '-o', 'linewidth', 2)

Why does the PSD of all the data show little to no gamma band power but the decomposition of the bands show a lot of gamma band power? Thank you!

Answer 1

[Hamza Imtiaz]

Because the range of frequencies in the gamma band is 40 Hz (30-70) , whereas delta theta and alpha are all 4 Hz