[Eeglablist] Central Frequency Estimation

[Martín Martínez]

----- Original Message ----- 
From: Martin Martinez 
To: eeglablist at sccn.ucsd.edu 
Sent: Friday, November 21, 2008 8:00 PM
Subject: Central Frequency Estimation


Good evening,

I am working on EEG signals in PSGs. Once i filter the EEG signal with the correct low pass filter, i take some parameters of the signal to work with. One of this parameters is dominant rythm or central frequency. To estimate this, I make the FFT of the portion of signal and then i calculate the centroid. After that, I plot the evaluation of the central frequency along the EEG signal. My problem is that I think that the central frequency is 1-2 Hz higher that I hope and my question is if anyone knows another methods or related info on the matter.

Alter that, I tried an AR-2 model of the signal and the average of central frequency decreases, but if i try to make it with the EMG signal, the behavour changes...



Can anyone help me with this 'Central Frequency Estimation' in EEGs?


Thank you very much,

Martin


P.S.: I apply before notch filters to eliminate DC signal, and contribution of low-frequency is very little (so the spectrum seems correct).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20081121/250f7b7a/attachment.html>