Question about the harmonics in the EEG signal?

O

Odessamarin2012-10-22 17:02:32

Mathematics

Odessamarin, 2012-10-22 17:02:32

Hello, tell a dumb one ...
If after Fourier transform, a complex signal (EEG) in the spectrum there are peaks of 70, 140, 280 Hz ... can we assume that the fundamental frequency is 70, and the subsequent ones are its harmonics?
The question is, these hormones, what is it? It can be said that there is a frequency generator for both 140 and 280, but they are enhanced by the fact that they are superimposed on a more powerful 2x carrier 70Hz? Therefore, these peaks are on the spectrogram? Or is it precisely because of this multiplicity that they exist, and there is no generator of 140 and 280? Do they just stand out from the rest of the spectrum?
Thank you if someone can clarify in a nutshell)

Reply

Answer the question

In order to leave comments, you need to log in

5 answer(s)

O

Odessamarin, 2012-10-22
@Odessamarin

in the fact that these are pickups, and in the fact that a person has nothing to generate a signal of such a frequency ... neurons generate. and 70 is not limited there.

K

kbtsiberkin, 2012-10-22
@kbtsiberkin

Are the mentioned peaks very narrow? If so, then 140 may well be a consequence of 70. Is the suggestive harmonic at 210 Hz not observed? A series of 70-140-210-280 and further may well be realized with only one source at the lowest frequency. But on the condition that the generator of these same 70 Hz has a good quality factor and creates non-sinusoidal pulses that are very narrow in time, and not a continuous signal close to a sinusoid.
Such pulses have a rather slowly decaying spectrum, on the basis of which the so-called. quartz calibrators - when generating rectangular pulses at a frequency of, for example, 100 kHz, you can catch their descendants at units, or even tens of MHz.

E

egorinsk, 2012-10-22
@egorinsk

The harmonics are hard-wired to the fundamental frequency, meaning when it gets louder, they get louder as well, as they are generated from the same source. In general, it is impossible to distinguish between a valid fundamental frequency + a set of harmonics and a sum of 2 multiples of frequencies and their harmonics.
It's just a mathematical abstraction.

F

Fil, 2012-10-22
@Fil

Not familiar with the EEG signal. But 140 and 280 are not a consequence of 70. If the peaks stand out strongly, then your signal can be approximately represented as:
f(t) = a*cos(70*t) + b*cos(140*t) + c*cos(280 *t).
The essence of the Fourier transform is precisely in identifying such frequencies. To understand how strong a signal is generated at certain frequencies. If there were no generation at 140 and 280, then there would be no peaks.

O

Odessamarin, 2012-10-22
@Odessamarin

Thank you ... and if you leave the eeg. For example, in the amplified signal, on the same EEG, there will be interference from the electrical network. In Europe, 50 Hz, but also further down the spectrum, there is an option to see 75, 100, 125, 150 ... smaller peaks. How is it here? because in the seto 50Hz. where do the harmonics in the signal come from? After all, there are no generators of such frequencies?
In other words, when I analyze the spectrum starting from high frequencies. Having found a peak, say 210Hz, I can predict peaks at 140 and 70. with 70 being the largest. And when does such a prediction turn out to be true? What kind of magic? or all the same 210 and 140 is the harmonica 70? In this case, hormones = artifact? And there are no generats above 70? then where are the peaks from?