bci2000.org BBS

Hello,

I am doing some P300 experiments running on BCI2000 plateform. Everything is working well.
I tried to go in a deeper understanding of the genesis of EEG signals because some of the expected P300 signals where negative and late (400->600ms): it does not have any effects on the classification but I would like to understand.

I find out that the polarity depends on the deep of the activating synapses, but I think a P300 is so-called because it's always positive.

By now, I pointed out two causes that could change the EPR shape:

• -the position of the electrode around the grd/ref: an positive deflexion in the front will give a negative one on the back (I can see that with EOG but I'm still surprised since my amplifier is said to use common ref)
  -the overlapping of potentials caused by differents stimuli since ISI<200ms

Do you have an idea of what could be the reason?

Thanks

Louis

Louis,

The "P300" that you will see in individual subjects will often differ from the textbook P300 response. It may contain negative components and/or have a different latency. There are probably multiple reasons for this:
1) individual differences (textbook responses are often averages from many people)
2) the fact that the P3Speller is different than a simple oddball paradigm
3) Can be attributed to the montage, as you mention.

It should not be related to the short ISI since the waveforms appear to be similar for different ISIs.

Gerv

hi every one

I am finding averages for my target data (in P300) ........ means targeted rows average and targeted columns average . The results are some thing like this ..

As we are focusing on the single letter 'P' .....so it should give Positive Peak at 300ms in column (4) and Row (3) [ refer to BCI contest November 2002 documentation for rows and columns numbers]. As i am using Brain Vision aquisition system .

Results after averaging :

[b]Column (4) has positive peak at 300ms and Row(3) has -ive peak at 300ms . [/b]

Problem:
How i conclude this ..Is this fine to have such kind of results ? or both rows and columns should be positive or negative ?

Or Is there any contribution (in problem) of machine from which we are acquiring the data i,e: Brain vision amp.


Thanks in advance.