matlab code

I. WRITE Mablab code to perform the following tasks (be careful to follow the exact sequence of steps) and COMMENT your code properly: For each file/trial (i.e., each VEP signal) do the following. Repeat 800 times automatically to process ALL the files. [Suggestion: use for loops with fgetl to read the filenames from ca800.txt.] 1. Load a VEP signal (size 64x256) from a .mat file. automatically and correctly loading the 800 VEP signals] 2. Compute the VEP signal with reference to channel CZ, which is channel 16. That is, subtract channel CZ from each channel. After that, remove channel CZ and the three reference channels, which are X (channel 32), nd (channel 63), and Y (channel 64). From now on, you have 60 channels to process in each trial. 3. Set the mean of each channel to zero (i.e., remove the mean of the signal from each channel). 4. Use PCA to reduce the noise (mainly ongoing EEG noise) from each VEP signal. Use 95% variance measure to select the appropriate number of principal components. You may use Matlab functions pca and pcares. 5. Since P300 responses are band-limited to 8 Hz, filter the VEP signal using an appropriate bandpass filter with passband from 2 Hz to 8 Hz and with stop frequencies at 0.5 Hz and 12 Hz (You should try different edge frequencies later, observe the results, and explain their effects). Set the minimum attenuation in the stopband to Rs=15 dB and passband ripple Rp=1 dB. You could use either IIR or FIR filter. If you decide to use IIR filter, you may use Matlab function buttord or ellipord to choose the filter order and filtfilt to implement the filter. If you use FIR filter, you may use Matlab functions kaiserord and fir1 to design the filter and conv to implement the filter. 6. Compute the P300 peak amplitude and the corresponding latency time for each channel. You should have one amplitude value and one latency value for each channel (you may use one matrix for amplitude and one for latency). The P300 peak can be identified as the largest positive peak (there could be several peaks) in the period of 300-600 ms (find the corresponding sample points) after the stimulus onset (the start of a trial). If there are no peaks in this time period, set the P300 peak amplitude and latency to zero. After the above loops (for all the VEP signals), do the following: 1. Calculate the average P300 amplitude and latency time of the alcoholic subjects in each channel (i.e., from the first 400 files) and the average P300 amplitude and latency time of the non-alcoholic subjects in each channel (i.e., from the last 400 files). 2. Compare the average P300 amplitudes of alcoholic subjects and non-alcoholic subjects in all the 60 channels using plot. 3. Compare the average P300 latencies of alcoholic subjects and non-alcoholic subjects in all the 60 channels using plot.