[FieldTrip] What exactly does 'minnbchan' refer just to spatial neighbors?

Eric Maris e.maris at donders.ru.nl
Thu Feb 10 11:16:05 CET 2011


Hi Michael,

> there is one additional side effect of specifying a LARGE minnbchan
> that actually leads to SMALLEr clusters in a certain sense. If I
> understand correctly, at any point in time&frequency the cluster has to
> have a spatial extension of minnbchan+1. So when you specify a large
> value the cluster may more easily be cut up into several smaller ones
> along the space&time dimension - which may or may not  be desirable.

Whether it is a side effect or not, depends on how you look at it. For me it is its intended effect. I implemented the minnbchan-parameter after discussions with PhD-students that applied the cluster-based permutation statistics on channel-time-frequency data and produced significant clusters with weird shapes. Often, these shapes contained narrow "bridges" connecting nearby blobs in channel-time-frequency space. By increasing minnbchan, these bridges were removed and only the blobs remained.

Although I understand the requests of the PhD-students that I discussed with, I find the discussions about the minnbchan-parameter not very rewarding. In fact, it is looking for isolated blobs that one would like to interpret as separate physiological entities. This contrast with the nature of the EEG/MEG data, which has poor spatial resolution (due to volume conduction and common pick-up), often poor spectral resolution (due to the short time windows of our time-resolved spectral analysis), and often poor time resolution (due to latency jitter over trials).

Best,

Eric





> 
> Michael
> 
> -----Urspr√ľngliche Nachricht-----
> Von: "David Groppe" <dgroppe at cogsci.ucsd.edu>
> Gesendet: Feb 8, 2011 7:18:59 PM
> An: "Email discussion list for the FieldTrip project"
> <fieldtrip at donders.ru.nl>
> Betreff: Re: [FieldTrip] What exactly does 'minnbchan' refer just to
> spatial neighbors?
> 
> >On Tue, Feb 8, 2011 at 11:01 AM, Eric Maris <e.maris at donders.ru.nl>
> wrote:
> >> Dear David,
> >>
> >>
> >>> Thanks very much for the email Eric, but I'm confused as to why you
> >>> answered "No" to my second comment.  From what I understand, if you
> >>> use a value of minnbchan that is greater than 0, then any
> >>> significant clusters will necessarily spread across multiple
> >>> electrodes.  For example, if you set minnbchan to 1, then all
> >>> significant clusters will have to include at least two electrodes
> so
> >>> that each significant "voxel" has at least one neighbor at another
> >>> electrode.  Thus, if your effect is truly at only a single
> >>> electrode, having a value of minnbchan greater than 0 will force
> you
> >>> to miss the effect or to think it occurs at more electrodes than it
> >>> really does.  What part of my reasoning is incorrect?
> >>
> >> Your reasoning is correct and mine was incorrect. I should read
> >> better before I write. My apologies for this.
> >
> >Got it. Thanks very much for clarifying Eric.
> >
> >>
> >> I do have to qualify that with minnbchan=0 it may also happen that
> >> your significant cluster contains more than one electrode (the
> >> electrode with the true effect plus some surrounding ones that by
> >> accident exceed the threshold that is used to identify clusters).
> >
> >That makes perfect sense.  Since our lab uses only a limited number of
> >electrodes (32), I wanted to make sure I understood the minimum number
> >of electrodes an effect needs to span to be detected.
> >      once again, I very much appreciate your answers to my questions,
> >              -David
> >
> >> Best,
> >>
> >> Eric
> >>
> >>
> >>
> >>
> >>>      again, thanks very much for the help,
> >>>             -David
> >>>
> >>>
> >>> > Best,
> >>> >
> >>> > Eric
> >>> >
> >>> >
> >>> > dr. Eric Maris
> >>> > Donders Institute for Brain, Cognition and Behavior Radboud
> >>> > University P.O. Box 9104 6500 HE Nijmegen The Netherlands
> >>> > T:+31 24 3612651
> >>> > Mobile: 06 39584581
> >>> > F:+31 24 3616066
> >>> > mailto:e.maris at donders.ru.nl
> >>> > http://www.nphyscog.com/
> >>> >
> >>> >
> >>> >
> >>> >
> >>> >
> >>> >
> >>> >
> >>> >
> >>> >>     thanks for your help,
> >>> >>         -David
> >>> >>
> >>> >>
> >>> >> --
> >>> >> David Groppe, Ph.D.
> >>> >> dgroppe at cogsci.ucsd.edu
> >>> >> http://www.cogsci.ucsd.edu/~dgroppe/
> >>> >> _______________________________________________
> >>> >> fieldtrip mailing list
> >>> >> fieldtrip at donders.ru.nl
> >>> >> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> >>> >
> >>> > _______________________________________________
> >>> > fieldtrip mailing list
> >>> > fieldtrip at donders.ru.nl
> >>> > http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> >>> >
> >>>
> >>>
> >>>
> >>> --
> >>> David Groppe, Ph.D.
> >>> Postdoctoral Researcher
> >>> Kutaslab
> >>> Dept. of Cognitive Science
> >>> University of California, San Diego
> >>> http://www.cogsci.ucsd.edu/~dgroppe/
> >>>
> >>> _______________________________________________
> >>> fieldtrip mailing list
> >>> fieldtrip at donders.ru.nl
> >>> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> >>
> >>
> >>
> >> _______________________________________________
> >> fieldtrip mailing list
> >> fieldtrip at donders.ru.nl
> >> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
> >>
> >
> >
> >
> >--
> >David Groppe, Ph.D.
> >Postdoctoral Researcher
> >Kutaslab
> >Dept. of Cognitive Science
> >University of California, San Diego
> >http://www.cogsci.ucsd.edu/~dgroppe/
> >
> >_______________________________________________
> >fieldtrip mailing list
> >fieldtrip at donders.ru.nl
> >http://mailman.science.ru.nl/mailman/listinfo/fieldtrip





More information about the fieldtrip mailing list