How to apply a linear ramp to start and end of several EEG epochs to attenuate the abrupt onset and offset?

Hey there! I have a set of auditory EEG data recorded collected with the presentation of auditory stimuli, where each stimulus consists of a 3-second speech with a 5-second inter-stimulus interval. Each stimulus onset is marked by an S1 trigger. Currently I have cut off the EEG data without audio input, but for the remain EEG data with auditory stimulation I would like to apply a linear ramp to to both sides of each trial corresponded neural response (10% at each side) to attenuate the abrupt onset and offset. Could you please help me with the code to accomplish this task? Thanks a lot！

[Elaine]

It works! Thank you very much!

[William Rose]

@Elaine, you're welcome.

Answer 1

[William Rose]

@Elaine,

I will assume you have imported the EEG data into vector x, with T*Fs rows, where T=3 s=record duration, and Fs=sampling rate. I assume you want the signal to ramp up from 0 to full scale over the first 10% of the record, and ramp down from full scale to zero in the final 10%.

Fs=256;             % sampling rate (Hz)
T=3;                % record duration (s)
N=T*Fs;             % record length (samples) 
x=rand(N,1)-0.5;    % simulated EEG signal with mean=0
t=(0:N-1)/Fs;       % time vector (s)

Make the multiplier

m=ones(N,1);
Nt=round(N/10);
m(1:Nt)=(0:Nt-1)/Nt;
m(N-Nt+1:N)=(Nt-1:-1:0)/Nt;

Adjust the EEG signal

y=x.*m;

Plot results

subplot(211), plot(t,x,'-r'); grid on; ylabel('Original'); title('EEG')
subplot(212), plot(t,y,'-r'); grid on; ylabel('Adjusted'); xlabel('Time (s)')

OK