no feet desynchronization

Forum for discussion on different brain signals
Locked
paolobernardis
Posts: 40
Joined: 02 Jul 2010, 04:51

no feet desynchronization

Post by paolobernardis » 07 Feb 2011, 10:01

Dear all,

we are still permorfing tursor task experiments in two subjects. This time we use imagined movements instead of 'normal' ones. The problem is that, while tasks performed with hands give us good results, the feet imagined movement seems not to follow a normal behaviour: we can't observe, in any frequency range, a proper desynchronization of sensorimotor rhythms. Is it normal to observe this situation in some unprospective subjects, or can it be due to a mistake made by us?

Thank you so much for your attention,

Kind regards.

mellinger
Posts: 1052
Joined: 12 Feb 2003, 11:06

Post by mellinger » 07 Feb 2011, 10:13

Dear Paolo,

you do not write where you pick up the sensorimotor rhythms for which you expect desynchronization during imagined feet movement. Sensorimotor rhythms related to feet movement are originating from the motor feet area, which are located in the space between the two hemispheres. Thus, these rhythms can be best recorded from the Cz electrode. In case you have not already done so, you should do the analysis using signals from Cz.

Best regards,
Juergen

paolobernardis
Posts: 40
Joined: 02 Jul 2010, 04:51

Post by paolobernardis » 07 Feb 2011, 12:21

mellinger wrote:Dear Paolo,

you do not write where you pick up the sensorimotor rhythms for which you expect desynchronization during imagined feet movement. Sensorimotor rhythms related to feet movement are originating from the motor feet area, which are located in the space between the two hemispheres. Thus, these rhythms can be best recorded from the Cz electrode. In case you have not already done so, you should do the analysis using signals from Cz.

Best regards,
Juergen

paolobernardis
Posts: 40
Joined: 02 Jul 2010, 04:51

Post by paolobernardis » 08 Feb 2011, 08:19

Dear Jurgen,

thanks for your response. I was referring to the lack of desynchronization of the SM rhythms in the electrode Cz (fro the motor feet area). Have you met the subjects in which there was observed no - desynchronization? Or something we're wrong?

Best

mellinger
Posts: 1052
Joined: 12 Feb 2003, 11:06

Post by mellinger » 08 Feb 2011, 08:38

Dear Paolo,

there is always a number of subjects ("BCI illiterates," reportedly up to 30% of subjects) for whom there either is no sensorimotor rhythm visible in the EEG, or no modulation of that rhythm by movement imagination.

To debug your setup and procedure, I have the following suggestions:
*You might try actual feet movement to see whether this results in a sensorimotor rhythm.
*For feet movement, you should instruct the subject to move as if he/she were able to grasp a ball using his/her feet.
*If you are using a Laplacian filter at Cz, try with CAR instead.

Best regards,
Juergen

Ryan
Posts: 29
Joined: 27 Aug 2010, 15:03

Re: no feet desynchronization

Post by Ryan » 20 Apr 2011, 12:14

Hi Juergen,
I am sorry to interrupt here, but I'm having a similar question. 30%,um, that is quite significant.

In my lab, we are using the 64-channel sensor net from EGI.
Unfortunately, Channel Cz is used as reference, so that I cannot get signal from it.
so can I use channel 4 or 34 instead (refer to the channel map below), if they have big r2 values?
One more thing I donot understand is, for 2D control, we are using C3 and REST to move cursor horizontally, using C3 and C4 together to move cursor up.
Channel C3 is used in both conditions, won't they interrup with each other? E.g. when we move the cursor up, the cursor will move to right too.

Thank you very much!
Sincerely,
Ran
Attachments
HvF.JPG
hands Vs feet
HvF.JPG (205.13 KiB) Viewed 5448 times
New Bitmap Image.JPG
channel map
New Bitmap Image.JPG (86.76 KiB) Viewed 5448 times

mellinger
Posts: 1052
Joined: 12 Feb 2003, 11:06

Re: no feet desynchronization

Post by mellinger » 20 Apr 2011, 13:09

Hi Ran,
nfortunately, Channel Cz is used as reference, so that I cannot get signal from it.
you could use the negative average of the surrounding electrodes as an approximation to a Laplace-Filtered Cz. But of course you may use neighboring channels with large r^2 values as well, provided that you obtained the r^2 values from imagined feet movements.
One more thing I donot understand is, for 2D control, we are using C3 and REST to move cursor horizontally, using C3 and C4 together to move cursor up.
It's not clear to me how you combine channels. If I understand correctly, you are using the C3 amplitude at some frequency for horizontal movement, and the sum of C3 and C4 amplitudes at some frequency for vertical movement. If the two frequencies are the same, you will indeed have a skewed cursor movement rather than a clear separation of horizontal and vertical axis.

IIRC, in their 2D studies, Wolpaw and McFarland used orthogonal combinations, such as the sum of C3 and C4 at 12Hz for one direction, and the difference between C3 and C4 at 24Hz for the other direction.

HTH,
Juergen

Ryan
Posts: 29
Joined: 27 Aug 2010, 15:03

Re: no feet desynchronization

Post by Ryan » 20 Apr 2011, 14:20

Hi Juergen,
Thank you for the reply.
Different frequencies indeed solve this problem.
Thanks a lot!

Sincerely,
Ran

Locked

Who is online

Users browsing this forum: No registered users and 1 guest