Needed FRequency analysis of an EEG signal

Hi everyone.i am poor about matlab. i need to decompose EEG signals to ALpha,beta,gama... by using FFT. İ do not have a EEG database neither i do not know to use to database. please show me a way. thanks

Answer 1

[Salaheldin]

to be more clear

x=load('EEG_data.txt'); y=fft(x); PS=abs(y).^2; N=length(x); fs=1000; freq=(1:N/2)*fs/N; plot(freq,PS)

Answer 2

[Kosai]

Onur i didn't see any Question in your Post, but i got that you like to extract the Band Frequancy from EEG-Data-Raw. At first the recorded EEG-Data were usually saved in Dataset not in Database, so you dont need to have a database of EEG but you need to make search in google about "EEG-DATASET". Second, The best way to extract the Band-Frequancy fromm EEG-Raw is the wavelet analysis, so if you have the wavelet-toolbox in your matlab version you can use this following code to extract the Band-Frequancy, but a very important piont is what is the sampling frequancy of your EEG-Raw ?? it is very important to determine how many Level do you have to extract to achieve the Band-Frequancy, so i will assume that your Fs = 1000Hz, so you need (8 Level wavelet decompostion):

                S = "your EEG-Data-Row";
                waveletFunction = 'db8' OR 'sym8' ;
                [C,L] = wavedec(S,8,waveletFunction);
                %%Calculation The Coificients Vectors
                cD1 = detcoef(C,L,1);                   %NOISY
                cD2 = detcoef(C,L,2);                   %NOISY
                cD3 = detcoef(C,L,3);                   %NOISY
                cD4 = detcoef(C,L,4);                   %NOISY
                cD5 = detcoef(C,L,5);                   %GAMA
                cD6 = detcoef(C,L,6);                   %BETA
                cD7 = detcoef(C,L,7);                   %ALPHA
                cD8 = detcoef(C,L,8);                   %THETA
                cA8 = appcoef(C,L,waveletFunction,8);   %DELTA

                %%%%Calculation the Details Vectors
                D1 = wrcoef('d',C,L,waveletFunction,1); %NOISY
                D2 = wrcoef('d',C,L,waveletFunction,2); %NOISY
                D3 = wrcoef('d',C,L,waveletFunction,3); %NOISY
                D4 = wrcoef('d',C,L,waveletFunction,4); %NOISY
                D5 = wrcoef('d',C,L,waveletFunction,5); %GAMMA
                D6 = wrcoef('d',C,L,waveletFunction,6); %BETA
                D7 = wrcoef('d',C,L,waveletFunction,7); %ALPHA
                D8 = wrcoef('d',C,L,waveletFunction,8); %THETA
                A8 = wrcoef('a',C,L,waveletFunction,8); %DELTA

[onur]

Thanks for your helpful answer. But yet i havent studied about wavelet. For now i just need to find frequwncy band of an EEG_DATAset by using FAst fourier Transform.?

[Srey]

And how the frequency range can be found out, corresponding to each wave ??

[Srey]

What will be the decomposition level if the sampling frequency is 128 hz?

[mamta p]

I dont have wavelet toolbox, can you please tell me how can i decompose EEG signals using FFT. can you please provide me the code.

Answer 3

[onur]

selahaddin really thnx. your answer is realy helpfull.

[Shan Ahmed]

Hi I'm trying to detect epilepsy from eeg signals, can you pls help me with the code for any feature extraction?

Answer 4

[Salaheldin]

Hi Onur,

I hope this message isn't too late...Here is what you need to do x=load('EEG_data.txt'); % assuming your data is stored in a text file called 'EEG_data.txt' y=fft(x); % this line computes the Fast Fourier Transform of your EEG data PS=abs(y).^2; this line computes the power spectrum N=length(x); % this line determines number of data points in your data for plotting fs=1000; % I am assuming your sampling frequency is 1000 Hz freq=(1:N/2)*fs/N; % this line generates a frequency vector for plotting your data...only plotting half your spectrum is sufficient..rest doesn't have unique information. plot(freq,PS) % ENJOY!!!

[Sandeep B]

Hi Salaheldin,

freq=(1:N/2)*fs/N; plot(freq,PS);To plot these two vectors they are of different length, i wonder why the frequency range is of half of the signal length and it is multiplied by fs/N. It would be greatly appreciated if you could help me. Thanks

[NavMood]

Hello Salaheldin

I have done the steps as you've mentioned. But I am having problem in plotting the graph between PS and frequency. I will be grateful if you help me

[suganya sri]

i am plting(freq,PS); ,i got error vector must be same length,pls help me how to clear

[pavan kumar]

make PS to half to plot

Answer 5

[upasana talukdar]

Hi,

My EEG Data has 250 Hz Sampling frequency, And I want to extract signals of all bands from given main signal using DWT. Please tell me that in how many levels should I decompose my Signal for Fs=250?

[suganya sri]

my EEG data has 256 Hz sampling freuency ,please tell me that in how many level should i decomposes my signal?

[Rasa]

Did you find any method to find out the number of levels needed ? Do you have matlab code to find out the frequency sub-bands of the EEG signal ?

[pavan kumar]

use max level decomposition