EEG Signal Processing - Science topic

I have a lot of .eas files that I need to convert to edf files or even mat files if possible. I would like to proceed with an automatic method.

How to apply K-Fold cross-validation in EEG Datasets? Because EEG datasets have train and test parts for every subject separately. So if we want to apply K-Fold cross-validation, do we have to concatenate train and test data? In this case, the train and test parts will be fully mixed. Can anybody clarify?

I am trying to export EEG data as ascii files with each trial as a series of rows. To clarify, I have 32 channels and 200 trials and I am trying to generate a data set of 6400 rows. I have used various export functions but I haven't been able to generate a file in this format (compatible for Matlab). Is this possible?

Hi,

I'm preprocessing some EEG datasets in an EEG-VR experiment with some neck movements.

In some of the datasets, some spindle-like alpha activities appear in most channels (mostly occipital, parietal, and frontal but less central channels). I'm not sure if they're some kind of noise maybe because of the VR headset on the EEG cap and therefore should be removed or not.

After doing ICA, these activities will also appear on some of the components, which ICLable identifies as either 100% or above 95% brain activity. I attached some photos of the channel activities and components. I'll appreciate it if you can help to be sure whether these are noise or not and if they are how should I remove them.

Thank you very much.

We had a question in this article. How did they calculate a single band value representing all 64 electrodes to compare between-subject groups? Usually, a specific band (i.e., alpha) should have 64 pieces of data towards corresponding 64 electrodes.

The table shows that each participant have a single rest data for one specific band, and then research use independent t-test to compare between two groups.

Can any experts help us to understand this?

Reference:

Ding, Y., Cao, Y., Qu, Q., & Duffy, V. G. (2020). An exploratory study using electroencephalography (EEG) to measure the smartphone user experience in the short term. International Journal of Human–Computer Interaction, 36(11), 1008-1021.

Zhepeng

Our research team met one question on calculating EEG relative power and absolute power at this stage.

When we integrated all negative and positive amplitude/power data in five EEG bands (delta, theta, alpha, beta, gamma), a few relative power results became huge (i.e., 440%(44.44) or even over 1000%). We thought these values were abnormal results. The reason is that the integration result of five EEG bands with negative and positive power values could be 1 or 2 as the denominator, but the numerator could be very large for the integration of one specific band(i.e., delta). The relative power calculation is (sum of spectral power in the band)/(sum of spectral in all bands)

The attached image showed some negative and positive spectral power values.

Therefore, we would like to ask whether we need first to transfer negative value to absolute value to consider relative power or absolute power. Normally, the relative power should be around 0-100%.

Can experts help us? Could experts please share some references with us?

It  is the raw data of EEG consisting of different types of artifacts.This can be used to apply different algorithms which will remove the artifacts from the signal and finally we get clean EEG

I am trying to use new emotiv epoch+ headset which was bought in 2018. I am having half of the electrode as green Like this picture. But my question is why the over all contact quality is 0% (written in Red color).

I am working on the effect of music therapy on persons with Alzheimer's Disease. I require EEG signals of healthy individuals and people with Alzheimer's Disease/Dementia that have undergone music therapy. Due to the on-going pandemic, I am unable to collect signals in real-time. It would be helpful if I could be directed to a database that I can get EEG signals from.

Thank You.

Dear community , I tried to extract features using continuos wavelets transform using python on my data , but I faced some problems ; my dataset are sleep recordings for 10 patients (physionet sleep dataset) , after selecting a patient randomly ,I kept just 2 eeg channels and dropped the other channels (eog , ecg , emg ) , I extracted the epochs (channel , time , event) , how I can do my feature extraction ?

Thank you

For my master degree thessis i need to do Two Class Motor Imagery classification analysis.

I have data sets from http://bnci-horizon-2020.eu/database/data-sets

But i can't load .mat files, and i don't know how to start?

May anyone can help me? I want to learn machine learning but i lack experience.

I have some problems to send triggers marks to the eeg recording via psychopy. Even i have searched for answers i could not be able to synhorinzation.

Best,

I'm doing my master's thesis in developing a BCI for post-stroke rehab with motor imagery. The device I'll be using only has 5 channels, so I was wondering if it is possible to train a classifier with more channels (from a dataset in physionet for example) and then use it with that device?

Thank you in advance!

Hello every one

i am using ultracortex headset cap , open bci to measure eeg signal processing ,i am using treadmill  and harvard step test open eye and closed eye , so i need to ask Is it possible to measure the perception kinesthetic as a mental process by using harvard step test closed eye? 

Thanks for attention

Any correlation between brain electrical activity signal and neurotransmitter activity?

I am required to filter out noise from EEG data using preferably Python or MATLAB. My dataset contains values for 64-electrode EEG along with their time-corresponding HEOG (horizontal eye movement), VEOG (vertical eye movement) and ECG (for heart-complex artefacts) values. I wish to implement cascading adaptive filtering technique and have gone through the paper 'Noise Removal from EEG Signals in Polisomnographic Records Applying Adaptive Filters in Cascade' by M. Agustina Garces Correa and Eric Laciar Leber () but I am not proficient enough with programming to allow me to program the logic.

Could anybody tell me what libraries and functions would be appropriate for this?

Dear All,

I want a standard data set of EEG signals for the intent of movements. I want the standard data sets for left, right, front, back, start, and stop movements of alpha, beta, and gamma signals. Please let me know, where can I find the standard data set.

Hello. I am thinking of using tDCS to stimulate the motor cortex C3/C4 or the SMA Cz to study possible effects on motor imagery. But I am still undecided which area would be better.

Thank you for any advice.

Hi,

I am new to EEG signal processing. I am now working on the DEAP dataset to classify EEG signals into different emotion catagories.

My inputs are EEG samples of shape channel*timesteps. The provider of dataset has already removed artifects.

I use convolution layers to extract features and use a fully connected layer to classify. I also use dropout.

I shuffle all trails(sessions) and split the dataset into trainning and testing sets. I get resonable accuracy on the trainning set.

However the model is anable to generalize accross trails.

It overfits. Its performance on the test set is just as bad as a random guess(around 50% accuracy for low/high valance classification).

Is there any good practice for alleviate overfitting in that senario?

Another thing bothers me is that when I search for related literature, I find many paper also give an around 50% accuracy.

Why are results from EEG emotion classifcation so bad??

I feel quite helpless now. Thank you for any suggestions and reply!

I am working on classification of schizophrenia and i see the channels placed at central sulcus give good accuracy then using all channels. I want to this automatically using some algorithm.

Here is a paper related to it,

but it seem most of them are computationally extensive or working on feature reduction leading to channel reduction.

I am looking for some direct channel selection algorithm. Any latest paper in your mind? or any idea please

I would like to use Eprime file to import maker in EEG file (Brain Vision). Data in the Eprime files correspond to good and wrong answers and I want to use it for ERPs analysis. Can someone give me some advice?

I am looking for a intracranial EEG dataset of patients with Epilepsy to validate a algorithm. In the most of papers, the researches used the Freiburg database, but now this database is part of EPILEPSIAE project. To use this database the researcher has pay 3.000 euros. Unfortunately I don’t have this money in my research.

Hello,

I am new to using Mobita in EEG recording, it was working fine with me before, but recently I found 50Hz noise in the recording (I verified the type of noise through the frequency spectrum of the signal which has a very high peak around 50Hz, and when filtering the signal with a notch filter around the 50Hz, the oscillation in the signal disappeared). They have mentioned in their website that this type of interference could be due to either: bad ground electrode contact, or bad electrodes contact, or a broken cable. I tried to make sure that the electrode in a good contact and there is enough conducting gel but I still see the problem.

I don't know how to verify that there are broken cables or electrodes since the headcap has the electrodes embedded in it. So, I would appreciate if anyone can tell me any insights about how to troubleshoot the Mobita system (maybe there is a way through polybench software or the matlab interface to get statistics about electrodes connections or something). I have attached a snapshot of the data that I am currently recording.

I appreciate your input. Thanks a lot.

I am doing research on the neural network prediction of muscle movement with EMG inputs, and I want to incorporate EEG data as well. However, I do not have the resources to collect this data myself, so I have been looking for a study or research lab that has done the data collection already.

Ideally, the data would be collected during human movement, with EEG electrodes on the scalp and EMG electrodes on the surface of whatever muscles are moving (arm trajectory, gait analysis, etc.) I have found many studies that collect either EEG data or EMG data, but not one that collects both together. I would greatly appreciate any help in this matter.

I am making a project in the neuromarketing domain (using EEG), in which we will measure attention, preference, familiarity and memory for video-based advertisements. I am having a lot of trouble selecting compatible tools for EEG pre-processing, machine learning, and creating a GUI to display results.

We will be making out stimuli in OpenVibe (most likely), and python for machine learning.

Can someone please recommend which tool to use to make a GUI that allows users to upload their EEG files, and displays the results of EEG signal processing and output of machine learning algorithms?

I would really appreciate any suggestions.

Hi all,

I am trying to extract lower alpha (8-10Hz) and upper alpha (10-13Hz) frequency from EEG signal. I have already extracted Alpha, Beta , Gamma and theta bands using Wavelet decomposition in MATLAB, but now I want to use wavelet decomposition to extract lower alpha and upper alpha. Can any body help me out?

Note: for wavelet decomposition I am using the following code to extract details coefficients;

fs=128;            %THIS IS MY SAMPLING FREQUENCY

%data is my eeg signal

wavelet = 'db2'; %wavelet function

scale = 5; %No. of wavelet level

[C,L] = wavedec(data, scale, wavelet);

%%%% Calculation the Details Vectors

D1 = wrcoef('d',C,L,wavelet,1); %NOISY

D2 = wrcoef('d',C,L,wavelet,2); %GAMMA

D3 = wrcoef('d',C,L,wavelet,3); %BETA

D4 = wrcoef('d',C,L,wavelet,4); %ALPHA

D5 = wrcoef('d',C,L,wavelet,5); %THETA

A5 = wrcoef('a',C,L,wavelet,5); %DELTA 

thank you

We want to classify motor movement tasks using 8 EEG electrodes and want to enhance our signal .Since we have 8 electrodes we are unable to use laplacian .What are methods that give comparable accuracy ?

Can anyone guide me how to do EEG signal processing using MATLAB.

Hi everybody,

I recently had some resting state EEG recordings in closed-eyes condition in normal subjects. When calculating brain connectivity using PLV measure, I couldn’t find significant amount of connectivity in different bands. For instance, in delta, theta, beta and gamma bands the values obtained are around 10% or less. For alpha band the connectivity is a bit more, somewhere around 20 – 25% in a few locations.

So I have 2 questions: First, do these values make any sense for resting state? And second, considering that applying our methods for connectivity e.g. PLV, we expect at least a little of fake connectivity in our results. So is it really safe to say 10% of connectivity is not spurious and just a miscalculation?

(As a side note, I’ve tried to diminish spurious connectivity caused by volume conduction, sample size bias and etc.)

I have heard mutual information technique can be used in signal processing, can anyone suggest with how mutual information helps in EEG signal analysis?

Most of researchers are using fMRI for analyzing dynamic functional connectivity networks, I want to know if it is possible to use EEG as well and if so, what s the advantageous and disadvantageous of using EEG in comparison to fMRI, except temporal resolution of EEG

The data is of the format with each line of the recording file represents one power spectrum from 0 to 60Hz, and each value across the row represents the power in each frequency bucket, in increments of .25Hz buckets. The first row of the file should contain the frequency labels for each column. Each row is preceded by the timestamp at which that row's power spectrum was calculated. I am not having the phase information . It is taken from Neurosky Mindwave .

I'm trying to use matlab to calculate the PLV between two channels for continuous data but when I use Equation1 below the answer comes out with all 1's for the whole time series which can't be correct. Whereas, equation two looks more appropriate but I can't find any evidence of this approach being used.

% Hilbert Transform

Hilbert_A=hilbert(filtered_data(Channel_A,:)');

Hilbert_B=hilbert(filtered_data(Channel_B,:)');

% Phase Angle

PhaseAng_A=angle(Hilbert_A);

PhaseAng_B=angle(Hilbert_B);

%PLV Equation1

PLV=abs((exp(1i*(PhaseAng_A-PhaseAng_B))));

%PLV Equation2

PLV=abs(real(exp(1i*(PhaseAng_A-PhaseAng_B))));

I need some time series of EEG visual cortex data to process in matlab in mat format for my project purpose. Kindl please help me in this asap. Many thanks

I need hand movements EEG data source for project work. I usually finding epilepsy data and similar data. What should I do to find movement data?

Thanks.

Musa Can Kavak

I am doing an EEG study for which I need to have a measure of participants' current emotional state, particularly focused on negative emotions. Any suggestions for reliable questionnaires that measure this?

I am seeking for the best signal processing package or course in python, especially for EEG/MEG signal processing, what packages are available? and which is the best one?

Good morning,how can i decompose my EEG signal into regular frames using matlab ,and after that calculate for each frame the power of wavelet coefficients in order to detect epilepsy seizure. Thank you in advance .

How can we use EEG source localization in a clinical application?

Do you know best software and package? free or other conditions

Do you have experience with BESA or Epilog?

I am comparing movement-related spectral perturbations between groups of patients with unstable ankles and matched uninjured healthy individuals. In my spectral plots, I am able to easily visualize the movement-induced alpha suppression, yet there appears to be a difference in beta activity between my groups; my uninjured group appears to have broad suppression across the beta frequencies, yet my unstable ankle group does not.

As this is a baseline-normalized dataset, I wanted to test whether or not this was due to differences between my group at baseline with respect to beta power. I plan on doing this by testing the power spectrum during both my baseline and period of interest, similar to how https://www.ncbi.nlm.nih.gov/pubmed/24457137 was reported in patients with ALS.

My question is, as EEGLAB log-transforms TF power, and my data appears to show differences through this computation method, if I am testing for differences between my groups should I compare my baseline and period of interest power spectra in log-transformed power (by setting ‘baseline’, [NaN]) or transform this output to absolute power?

Hi there,

We recently acquired an EGI MRI compatible EEG recording system of 32 channels. We have an issue concerning the ballistocardiographic artifact appearance. In all the recordings that we have performed (6) we have found a characteristic spatial distribution pattern of the artifact: The frontal electrodes are the most affected by the artifact (highest amplitude) and we observe a gradually decreasing amplitude artifact toward posterior electrodes (occipital electrodes are almost unaffected in some recordings) *See example of a recording. We have found that using post processing artifact correction methods such as OBS and ICA effectively correct the posterior electrodes (as compared to an EEG recording of the same subject outside the MR scanner), whereas the frontal and central electrodes (most affected) show significant residual artifacts following correction.

Has anyone observed this kind of pattern while recording EEG inside the MR scanner?

What are the factors related to this uneven distribution of the BCG artifact? (preasure over the electrode?, electrode position relative to the B0 magnetic field?. ¿?

Which strategies can we use to attenuate the BCG in the frontal and central electrodes and homogenize the distribution of the BCG amplitude to ensure efficient artifact removal with postprocessing methods?

Thank you for you assistance

***electrodes in the figure are represented with the EGI GSN nomenclature; E1=Fp1, E2= Fp2, E3=F3, E4=F4, E5=C3, E6=C4; E7=P3, E8=P4; E9=O1, E10=O2, E11=F7, E12=F8, E13=T3, E14=T4, E15=T5, E16=T6, E17=Fz, E19=Pz

i will work by EEG signal for diagnosis epilepsy disease and for it i need to collect EEG signal data base.please help me anyone that work in this field recently.

thankyou

I'm a master student of Computer Engineering. I'm interested in BCI (Brain Computer Interface) and I want to write a thesis on this area, but I couldn't come up with a satisfying subject! I'd be glad to hear your suggestions.

As i have 3 emotive headsets epoch, epoch + but now all shows corrosion over electrodes and their nodes after a year of use. Readings are not good now. So headsets worked fine for almost a year but now its useless and so my money spent on it. If anyone has used openBci or any other headset which can work for longer period. Recommendations are needed

I have time series data from multi channel EEG. I am looking at various symbol based complexity measures. As a preliminary step I have to convert my time series to symbols based on some logic (as simple as order dynamics or zero crossing)

I am looking for better methods/algorithms to generate symbols from EEG time series.

Hi all,

We know that any kind of perception should generate some deflection in the EEG signal

When it comes to the movement of the human upper limb or the hand some deflection will be generated when moving the arm.

Now, the questions are

1- How strong and accurate is this kind of deflection?

2- Can we really track it on a time basis (like second by second)

for example, If I move my arm from t=1 sec to t=5 sec

Can I really determine this time gap only by analyzing EEG signal deflection?

Thanks

I found RIPPLELAB, but I want to know if there are more software (or a toolbox in matlab, python, etc ) that I can use to easy detect HFO in my data.

Hi,

I am totally new to the field of EEG signal analysis, but I am exploring it to see potential processing/analysis techniques to be designed and implemented on FPGA for onsite clinical decision support. So, I have raw EEG signal in edf format which I have successfully implemented into matlab and ran the following code to calculate the relative power (based on a code I found online). However, the results do not match some done by another professional in this area (very close, but does not match). Would anybody help me improve the code below to get more accurate results or suggest other venues. Note: if you advise me to go to EEGlab, please point me to the exact steps to do there. 

Thanks in advance!

Any help highly appreciated!

% Fs is my sampling frequency, x is my EDF data imported into matlab

Fs=1000;

t=1/Fs;

S =x;

waveletFunction = 'db8';

[C,L] = wavedec(S,8,waveletFunction);

%% Calculating the coefficients 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

POWER_DELTA = (sum(A8.^2))/length(A8);

POWER_THETA = (sum(D8.^2))/length(D8);

POWER_ALPHA = (sum(D7.^2))/length(D7);

POWER_BETA = (sum(D6.^2))/length(D6);

Total=POWER_DELTA+ POWER_THETA+POWER_ALPHA+POWER_BETA;

RELATIVE_DELTA=POWER_DELTA/Total;

RELATIVE_THETA=POWER_THETA/Total;

RELATIVE_ALPHA=POWER_ALPHA/Total;

RELATIVE_BETA=POWER_BETA/Total;

%**************End of Code

Here are my results versus the results I should achieve:

                             Old results                 My results

Relative_delta      0.602487293           0.572298897

Relative_theta      0.187303504           0.196419823

Relative_alpha     0.112981588           0.132041017

Relative_beta       0.073902161           0.099240263

I have EEG signal from which i need to extract frequency information. The signal have x-axis as number of sample and y-axis as amplitude. if am applying fft on signal its giving me frequency in KHz. Can anyone help me to find frequency and what method i need to apply to get it. Thanks.

Is it necessary to perform baseline correction to the epoched EEG data before time frequency analysis?

I've been started to study about CHAOS, and my main major is EEG signal processing.

I'm working on a bigger project that is about detecting and classifying emotion, attention or memorization in EEG signals.

My main question is that, if you friends can help me find an articles in this concept; which relates "Chaos" with "Emotion or attention or memorization"?

Best recommendations would be articles with rich and available data-sets on internet.Preferably for last three years.

Even a single "key word" or "article name" will help me, although much better to have article it self. 

Don't spare your thought from me. even single words could be help full.

Hello all,

We know that action observation and motor imagery produce changes in the mu rhythm event-related desynchronization, make the mu rhythm more focal and producing a higher % decrease (see for example Naima Rüther et al 2014).

However, I am wondering if anybody is aware of studies that specifically addressed the changes in mu rhythm and its desynchronization, after a motor learning protocol with actual movements, and of various length?

Thank you!

FABIO

In ERP/P300 signal analysis, xDAWN is well-known to find the spatial filter.

I have read several reference papers about xDAWN, such as

xDAWN Algorithm to Enhance Evoked Potentials: Application to Brain–Computer Interface

A Tutorial on EEG Signal Processing Techniques for Mental State Recognition in Brain-Computer Interfaces

But I still do not know very well about xDAWN. So far, I know that the first column of D is 0 except the positions of stimuli onset, but how about the other columns? or we do not need to know the others then we can create the Toeplitz matrix?

Would you please give me an example? Where can I find the source code of xDAWN to let me study more about it?

I'm trying to use EEG signals to feed an artificial neural network in order to diagnose mood disorders (specially major depressive disorder, but it would be interesting to try with bipolar disorder, schizophrenia etc).

However, I'm having a hard time finding quality databases of EEG with annotations of mood disorders.

Anyone knows where I could find it?

Thanks in advance.

There seems to be a handful of papers on single-subject resting-state fMRI data but not for EEG data. Is this simply because it is too variable to be done? If not, what are the things to watch out for in data collection and in analysis?

I am wondering is there a precise method to get a baseline for EEG biofeedback?

In one of the lubar's article, he suggested focusing on a fixed point for 90 seconds and average the "90 seconds recorded signal" to get the baseline for theta/beta ratio.

I want to calculate baselines for Theta, Beta, and mu-wave at C4, In the beginning of each session.

In order to achieve that, what do you suggest?

Hi, I have been recording EEG data for my research for 1 months now, but I`m comfortable about the EMG artifacts in my EEG data.There are both continues and burst of EMG activity in many subjects` data even after we have shown them how muscle activity will contaminate EEG data and informed them to stay relaxed and try not to move their heads during recording.

I wish to know what are some of the tricks to set up the recording environment (like what kind of chairs should we choose, or is it good to immobilize subjects' heads) to let subjects to stay as relaxed as possible and reduce EMG contamination.

Besides, what are some good ways to remove EMG artifacts from EEG data? I don't really want to simply remove the periods of data with EMG. I have tried ICA but the effect is not satisfactory. I am currently learning CCA (canonical correlation analysis) but have not tried it yet.

By the way, we use 64-channel active electrodes from Brain Product, and we don`t have any EMG/EOG electrodes yet (but we can buy them if necessary).

Thanks in advance.

Can anyone provide me Arduino and  Open BCI code for  connecting two external button on 8 channel Open Brain computer interface system ( ADS1299). Please ?

Hi everybody,

We've recently had some EEG recordings and we have noticed of something strange.

The recorded data is completely raw (at least to extent we are aware of) and when we look at the power spectrum of raw data we see something like figure 1, which is weird!

But when we apply 1 Hz high pass filter, the outcome turns into figure 2, which is a lot more appealing.

So I have 2 questions: 1) Why we see such terrible side lobes (for lack of a better word in my mind) in the raw data, and 2) why a 1 Hz HPF can solve that??

I am working on denoising EEG signals. The signal consists of 80 epochs. The denoising parameters which give the best results differ from one epoch to another. How can I deal with this problem?

I'm looking for actual value of data generation rate that is generate by ECG sensor ( such as smartwatch ) and EEG sensor ( such as Muse ). Can anyone help me find a reference paper that refers to these values. Please help. 

we have EEG signal simultaneously recorded with FMRI in 3 conditions (outside, inside without gradient and inside with gradient). I cleaned Artifact from EEG signal.

how can I evaluate the quality of cleaned signal?

I have raw EEG dataset in .mat files ( Matlab format). I need to perform band pass filtering on the data in the certain bands between 3Hz and 30 Hz. All the EEG recordings of this dataset are sampled at 256 samples per second, at 16-bit

quantization. Most of the cases contain 23 bipolar EEG signals. 

As I am relatively new in Matlab, how can I do that ?. please help!

What companies offer an eye tracker with software to do combine it with eeg within a budget of $78,000 or ₹500000?

Whats the cheapest option available in market?

Hi , i am currently working on motor imagery BCI (Brain Computer Interface). I am completely new to this field. I had downloaded data from BCI competition IV (data set -2-b, Left hand and right hand class ). I extracted alpha(8-12)Hz and beta( 14-30) Hz signal using band pass filter for C3 and C4 electrods for different trials , then i calculated average power for all trials. But i dont know how to calculate Event related desynchronization/synchronization ( ERD/ERS) in MATLAB. i dont know ho to calculate baseline power and change in relative power of %ERD/ERS. Can any body tell me how to do in MATLAB, please ?

Using EEGlab it is possible to display channel spectra and maps that plot EEG power with frequency. However, I am unsure how to obtain the data making up these plots so I can calculate the mean power across specific frequency bands (i.e. 4-8Hz for theta). 

For example, in a relative paper, the researchers state:

"EEG power was computed on the electrode of interest (see Fig. 2). Power was calculated in 0.49-Hz frequency bins and averaged across the appropriate frequencies to obtain the power values for theta (4–8 Hz), slow alpha (8–10 Hz), fast alpha (10–12 Hz), slow beta (12–20 Hz), and fast beta (20–28 Hz)."

Has anyone come across this issue before and know how to get around it? Any advice would be appreciated!

We all know that the human eye can percept only pictures with duration higher than 13 miliseconds. I have an 8 milisecond flicker picture that repeats 12 times in a second.(I have 120 black frames in a second (so each lasts 8ms) and every 10 other of them is my stimuli frame, so I will get 12 of them in a second).

I know that the operational range of SSVEP is 3Hz to 75Hz; much like to have strong relations with that famouse 13 miliseconds(1sec/75=13 ms).

So my main question is if SSVEP can show some stimuli wich our eye cant percept it?

In my specific case I have 12hz stimuli(in ssvep operational range)  with 8 milisecond(out of perception range). I'am eager to know if I will be able to see this 12hz in my ssvep or not? 

Hi everyone,

I am trying to apply ICA to remove artifact of EEG signals. Is it possible for me to write ICA algorithm in Matlab for removing. Because I prefer manual code than toolbox so I can modify for further application.    

Based on principle of ICA theory, x = As, so if EEG recording data is considered as input, is it x or s? and what is the final output, is it clean EEG signal and what is its called?

Thanks so much. It will be helpful for my graduate project.

There are my wavelet functions are available in Matlab. How can decide the best wavelet for decomposing the EEG signal to get sub-bands such as alpha, beta, gamma, theta, and delta.

By using  "Cross-correlation method " and "ANOVA" how can we find the best mother wavelet for mining the sub-bands.

Dear All,

I am using ADJUST plugin in EEGlab for artifact removal from raw data. I follow the instruction listed in the ADJUST manual to process the data. after running ADJUST I get the artifacted ICs in a new pop-up window, I mark these ICs for rejection and then I go to tools to remove these ICs (via EEGlab GUI menu>>tools>>remove components). A new dataset is created (eegdata pruned with ICA ), according to ADJUST tutorial this is the clean eeg data, but I see from the plots that there is still artifact present in the data. When I run ICA again and then again run ADJUST I get a new pop-up window in which some other ICs are identified as artifacts. I remove these ICs again. But still, artifact are present in data. I repeat running ADUST many time and each time I get new ICs marked as an artifact.

I would be happy if you guide me with this problem.