I am using a 24-channel system with 19 active electrodes to record the brain activity during flanker task. I run ICA to remove ocular artifacts. Although ICA can identify 1 or 2 eye-blink components, time course and ERP image show there still task-related brain signal in these eye-blink components. My question, is there any tips that can help to keep the real brain signal?
I am not sure what you are referring to by 'the real brain signals'. If you have a time locked event and you have a method of identifying 'the real brain signal' then there are methods of extracting the signals from the various channels of EEG data capture; such as signal averaging. But the EEG signals from any one EEG channel is the electrical activity of not only the location immediately under the electrodes, but also the effect of 'far field potentials' from other areas. Your electrode montage, the electrical characteristics of the amplifiers, as well as the impedance of the electrodes all play a role in the signals your are capturing.
Mohamed - sometimes a current source density transform can help to better separate signals from different sources, but it is recommended to have a relatively high number of electrodes for CSD transform. You might try it though, I am not sure whether it is implemented in EEGLAB, but there are other Matlab toolboxes that can do it.
HTH, Marius
I agree with Franklin Lue. ICA "takes the signal apart" and identifies components (also artifacts embedded in data) that are independent of each other, but of course different parameters of ICA may result in variations in your data. Good news is that blinks and saccades are way stronger than "the real brain signal", so it's easier to identify them. They also influence more the frontal electrodes than the central, parietal or occipital ones, so if you're interested in components in the latter locations, it's not such a big deal. I wouldn't worry about data loss caused by rejecting [artifact] channels with ICA, it's a risk the field takes, but if you're not convinced (and of course there are valid arguments for this scepticism), dig into the literature to understand the algorithm better and make an informed decision about how you want to treat your data. A brief search led me to this article, maybe you'd find it useful:
Article Variability of ICA decomposition may impact EEG signals when...
I think that any artifact removal is changing the signal under study, manual automatic or any other specific techniques included. In our study we try to analyze the segments which are artifact-free. I agree with Franklin Lue that the real electrical activity as you call it is distinguished from artifact already, but are also composite electrical signals. Since you do not have high density EEG to be a candidate for CSD transform, maybe you can try to search for artifact free segments and analyze those.