P300 Scenario: the sum of all the potentials for evey epoch

About the GUI application to design signal processing pipelines
Post Reply
etc
Posts: 31
Joined: Sat Mar 07, 2020 4:33 pm

P300 Scenario: the sum of all the potentials for evey epoch

Post by etc »

In the P300 Scenario I would like to calculate the sum of all the values (i.e.potentials) of all the channels of all the samples of every epoch.
One only number for every epoch.
Is it possible in some manners with the available Boxes?
Thank you very much!
Etc

Thomas
Posts: 210
Joined: Wed Mar 04, 2020 3:38 pm

Re: P300 Scenario: the sum of all the potentials for evey epoch

Post by Thomas »

Hi,

I am not sure it's quite possible with the current boxes.

The closest I get is something like this:
1. Splitting channels in different signals with the channel selector (1 per channel)
2. Sending all this signals (containing 1 channel each) to a simple DSP box to sum them.
3. Send the new signal (sum of all channels) into the signal resampling box and setting Sampling Frequency the same as the incoming signal, and sample count per block set to 1. This way, each sample will come out in its own epoch.

Now the last step would be to sum N of these new epochs containing one sample, with N being the number of samples per epoch before the resampling. However, I don't see a box to achieve this. The closest is the Epoch average, which will output one epoch every N, but it will not just sum them, it will average them (hence the name!)

The easiest might be to finish the process with a python box... ?

Hope this helps a bit, and let me know if I'm missing something, or if you see a way of finishing the process with available boxes.

Cheers,
Thomas

etc
Posts: 31
Joined: Sat Mar 07, 2020 4:33 pm

Re: P300 Scenario: the sum of all the potentials for evey epoch

Post by etc »

Hi Thomas!
You were very gentle to asnswer to my question!
Sorry for my bad English and sorry for the large delay in my new answer! But I had the fever and I lost your answer. Sorry!
The your solution is good! Because the average is useful for a my attempt. But it is only an attempt!!!
In 2 words: in all the papers that I read about P300, I saw always the use of Machine Learning to undestand the row and the column target. But the row and the column target, after the average for the n ripetions, would have 2 total of the neural potentials bigger than the totals of the potential of all the non target rows and columns.
I am sure that I am wrong, because if all the scientists use Machin Learning is because it is necessary.
But I would test my idea…

Post Reply