[Eeglablist] Dipolar components

[Makoto Miyakoshi]

Dear Ali,

> Yes, as you mentioned earlier, the scalp maps reflect directly the
columns of weight matrix. But this prediction of dipolar sources from the
IC scalp maps is not based on the smoothness, right? if yes, the rough
scalp maps could not be dipolar?

You can think of an counter-example of a smooth scalp map with 3 foci. This
is not single-dipolar.

> Could you please explain it more that why dipole inversion " is NOT a
cause of spatial non-stationarity"?

I already gave you the best (favorite) example. If you say dipole inversion
causes spatial non-stationarity, it is the same as saying that when you
play those dipole speakers they move around in the room.

Why do you say dipole inversion is a cause of spatial non-stationarity? Can
you tell me an example case?

By the way, when you say data non-stationarity, it usually refers to
temporal non-stationarity i.e. signal changes from block to block, for
example (from resting to task A, then to task B, etc).

Makoto



On Thu, Feb 2, 2017 at 12:04 PM, ali zahedi <ali.zahedi.bham at gmail.com>
wrote:

> Dear Makoto,
>
> Sorry for too many questions and my confusion..
>
> 1-  So does it mean that the appearance of the scalp maps does not have
> any relationship with the dipolarity? (i.e. can we judge/guess from the
> appearance of a scalp maps that it could be a dipolar source? )
>
> Dipolarity is directly determined by scalp topography. I can almost
> predict where dipoles should be fit when I see the IC scalp maps, if they
> are dipolar. It's a very simple thing.
>
> > Yes, as you mentioned earlier, the scalp maps reflect directly the
> columns of weight matrix. But this prediction of dipolar sources from the
> IC scalp maps is not based on the smoothness, right? if yes, the rough
> scalp maps could not be dipolar?
>
> 2- However, dipoles may invert when neurons fire, so the EEG sources
> can't be stationary. So how is this assumption plausible for the constitutive
> sources of EEG data?
>
> As EEG is an AC signal, by definition polarity must invert (a lot), which
> is NOT a cause of spatial non-stationarity. Do you have a speaker with a
> bass-reflex port in the back side? Or even better, do you own Magnepan,
> Martin Logan, Apogee, Quad, or old STAX? These speakers are dipolar, but
> they don't run around the room.
>
> > Could you please explain it more that why dipole inversion " is NOT a
> cause of spatial non-stationarity"?
>
> Regards,
> Ali
>
> On Thu, Feb 2, 2017 at 7:46 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> wrote:
>
>> Dear Ali,
>>
>> > So does it mean that the appearance of the scalp maps does not have any
>> relationship with the dipolarity? (i.e. can we judge/guess from the
>> appearance of a scalp maps that it could be a dipolar source? )
>>
>> Dipolarity is directly determined by scalp topography. I can almost
>> predict where dipoles should be fit when I see the IC scalp maps, if they
>> are dipolar. It's a very simple thing.
>>
>> > However, dipoles may invert when neurons fire, so the EEG sources can't
>> be stationary. So how is this assumption plausible for the constitutive
>> sources of EEG data?
>>
>> As EEG is an AC signal, by definition polarity must invert (a lot), which
>> is NOT a cause of spatial non-stationarity. Do you have a speaker with a
>> bass-reflex port in the back side? Or even better, do you own Magnepan,
>> Martin Logan, Apogee, Quad, or old STAX? These speakers are dipolar, but
>> they don't run around the room.
>>
>> Makoto
>>
>>
>>
>> On Thu, Feb 2, 2017 at 10:33 AM, ali zahedi <ali.zahedi.bham at gmail.com>
>> wrote:
>>
>>> Dear Makoto,
>>>
>>> Thank you for your explanation.
>>> So does it mean that the appearance of the scalp maps does not have any
>>> relationship with the dipolarity? (i.e. can we judge/guess from the
>>> appearance of a scalp maps that it could be a dipolar source? )
>>>
>>> Also, as it mentioned in Delorme et al. (2012) PLoS One paper "The
>>> motivation for the dipolarity is the assumption that brain and
>>> non-brain EEG sources have spatially fixed source locations and orientations".
>>> However, dipoles may invert when neurons fire, so the EEG sources can't be
>>> stationary. So how is this assumption plausible for the constitutive
>>> sources of EEG data?
>>>
>>> Regards,
>>> Ali
>>>
>>>
>>> On Thu, Feb 2, 2017 at 2:55 AM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
>>> wrote:
>>>
>>>> Dear Ali,
>>>>
>>>> The scalp maps of ICs directly reflects the columns of mixing matrix
>>>> (i.e. EEG.icawinv).
>>>> See slide 14: 'How does ICA model physiology' (from EEGLAB workshop
>>>> 2017 at Mysore)
>>>> https://sccn.ucsd.edu/mediawiki/images/7/74/IcaDecomposition
>>>> OfEegData4.pdf
>>>>
>>>> Therefore, if you see nice dipolar scalp maps, it means that ICA (which
>>>> does not know anything... it does not know channel locations, does not even
>>>> know the signal is EEG!) identified theoretical electrophysiological
>>>> property of EEG (see Nunes books about it).
>>>>
>>>> > Is it just the smoothness of the scalp maps that determines capacity
>>>> to fit dipoles?
>>>>
>>>> No, not the smoothness. It's dipolarity. See Delorme et al. (2012) PLoS
>>>> One paper.
>>>>
>>>> Makoto
>>>>
>>>> On Sat, Jan 28, 2017 at 12:00 PM, ali zahedi <ali.zahedi.bham at gmail.com
>>>> > wrote:
>>>>
>>>>> Dear all,
>>>>>
>>>>> Regarding the dipole fitting to the ICs and dipolar components, what
>>>>> are the properties of the scalp maps of the ICs that can fit with a dipole
>>>>> with less than a residual variance? Is it just the smoothness of the scalp
>>>>> maps that determines capacity to fit dipoles?
>>>>>
>>>>> I really appreciate it if you could help me with this.
>>>>>
>>>>> Regards,
>>>>> Ali
>>>>>
>>>>> _______________________________________________
>>>>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>>>>> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.uc
>>>>> sd.edu
>>>>> For digest mode, send an email with the subject "set digest mime" to
>>>>> eeglablist-request at sccn.ucsd.edu
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Makoto Miyakoshi
>>>> Swartz Center for Computational Neuroscience
>>>> Institute for Neural Computation, University of California San Diego
>>>>
>>>
>>>
>>
>>
>> --
>> Makoto Miyakoshi
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>>
>
>


-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170207/72377fff/attachment.html>