[Eeglablist] Connectivity normalization

[Makoto Miyakoshi]

Hi Michael,

The connectivity measures computed by SIFT is already smoothed in the sense
that moving-window freq-domain calculation is used, and you can specify the
type of windows. It is up to you whether you want to apply further
smoothing at the cost of time resolution. If you ask me if that's a good
practice or not, it generally indicates not so good because it indicates
your data are so noisy so that you need additional smoothing. But the
applying additional filter is not prohibited. If you want to do it, do it!

Makoto

On Wed, Jun 21, 2023 at 8:56 AM Michael Glassen - Biomedical Engineer <
MGlassen at kesslerfoundation.org> wrote:

> Hi Makato,
>
> I had a follow up, I’m doing a similar normalization to task related
> connectivity, and I was wondering if it’s good practice to also apply some
> filtering(moving average) to the connectivity data either before or after
> normalizing to help smooth out the effects in the time domain as I noticed
> they can oscillate.
>
> Best,
>
> Michael Glassen
>
>
>
> > On Jun 14, 2023, at 10:54 PM, Makoto Miyakoshi via eeglablist <
> eeglablist at sccn.ucsd.edu> wrote:
> >
> > Hello Sampath,
> >
> > You may divide the post values with the pre values. I have been using
> > subtraction to show the differences between the conditions in the past. I
> > guess both should work just fine. Or do you see any concern in using
> > division there?
> >
> > Makoto
> >
> >> On Wed, Jun 14, 2023 at 8:23 AM Sampath Thoutanahalli Kapanaiah <
> >> sampath.kapanaiah at uni-ulm.de> wrote:
> >>
> >> Dear Makoto,  Arnaud Delorme and EEGLab Users,
> >>
> >> I hope this email finds you well. I seek guidance on the appropriate
> >> method for normalizing SIFT resting state connectivity matrix data
> obtained
> >> from LFP continuous recordings in mice depth electrodes.
> >>
> >> My dataset consists of connectivity matrices calculated over 20-second
> >> intervals spanning a total duration of 60 minutes. The first 10 minutes
> of
> >> the recording represent the pre-injection baseline, while the remaining
> >> time captures the post-injection data.
> >>
> >> To check the drug's effect, I am interested in normalizing the
> >> post-injection connectivity data relative to the baseline. I have been
> >> dividing the post-injection connectivity values by the connectivity
> values
> >> obtained during the 10-minute baseline period. However, I am uncertain
> if
> >> this approach is the most appropriate and would greatly appreciate your
> >> expert advice.
> >>
> >> Considering the nature of the data and the desired normalization, could
> >> you kindly suggest the best method to normalize the connectivity matrix
> >> data to the baseline? I want to ensure that differences do not solely
> >> influence any observed changes in connectivity post-injection in the
> >> baseline values.
> >>
> >> Thank you very much for your time. I look forward to your valuable
> >> insights and suggestions.
> >>
> >> Best regards,
> >> Sampath
> >>
> >> PhD student,
> >> Ulm University
> >> Ulm, Germany
> >>
> >>
> > _______________________________________________
> > Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> > To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> > For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>