[FieldTrip] Question regarding nonparametric testing for coherence differences

[Jack Fogarty]

Hi Jan-Mathijs and Raghavan,

Thank you for your emails and for sharing the paper. I’ve been doing further research into this and have come across some other useful papers that may interest you.

The paper by Zalesky et al. describes a spatial pairwise clustering algorithm that provides a useful explanation of how you can analyse matrices with chan x chan x … dimensions using a cluster-based permutation method, similar to the network-based statistic. The paper by Hipp et al. also provides another example of this type of approach. There is no open-source code or toolboxes for spatial pairwise clustering that I know of, but there is code for the network-based statistic (via the brain connectivity toolbox) that may provide a starting point for developing your own code; I hope this is helpful for your own work.

Papers:

Zalesky, A., Cocchi, L., Fornito, A., Murray, M. M., & Bullmore, E. D. (2012). Connectivity differences in brain networks. Neuroimage, 60(2), 1055-1062.

Hipp, J. F., Engel, A. K., & Siegel, M. (2011). Oscillatory synchronization in large-scale cortical networks predicts perception. Neuron, 69(2), 387-396.

All the best,
Jack


From: Raghavan Gopalakrishnan <gopalar.ccf at gmail.com>
Sent: Friday, 19 March 2021 11:45 PM
To: FieldTrip discussion list <fieldtrip at science.ru.nl>
Cc: Jack Fogarty <jf752 at uowmail.edu.au>
Subject: Re: [FieldTrip] Question regarding nonparametric testing for coherence differences

Hi Jack,
I am also trying something along the same lines.
I think this paper might be of help to you. I tried contacting the authors but could not get a working code.
Phillips JM, Vinck M, Everling S, Womelsdorf T. A long-range fronto-parietal 5- to 10-Hz network predicts "top-down" controlled guidance in a task-switch paradigm. Cereb Cortex. 2014;24(8):1996-2008. doi:10.1093/cercor/bht050

Thanks,
Raghavan

On Fri, Mar 19, 2021 at 3:27 AM Schoffelen, J.M. (Jan Mathijs) <jan.schoffelen at donders.ru.nl<mailto:jan.schoffelen at donders.ru.nl>> wrote:
Hi Jack,

No, currently it is not possible to cluster data of dimensionality chan_chan_<something else>_<etc>. In older versions of the low-level code (findcluster) there were some lines of code (experimental) which have never been ‘active’ as such - i.e. they were not reachable when these functions were called from the higher level code. In the old days, if I remember well, Eric played around with channel-like clustering along more than one dimension, but the active development of this functionality was abandoned because it was hard to get it right. I have just finished a revision of the clustering code which will be released soon, where I actually haver removed those non-functional lines.

Best wishes,

Jan-Mathijs








On 19 Mar 2021, at 00:33, Jack Fogarty <jf752 at uowmail.edu.au<mailto:jf752 at uowmail.edu.au>> wrote:

Dear Prof Maris and Fieldtrip team,

Is there an established approach to clustering across channel x channel x frequency, as mentioned in this thread? I am hoping to try an exploratory between groups analysis of WPLI differences and it is unclear if the current fieldtrip functions behave when dealing with that complexity.

The ‘findcluster’ function seems well set up for matrices of chan x freq x time, but it is difficult to determine if it can handle chan x chan x freq appropriately, or if a new function is required altogether.

Any insight would be helpful thank you!

Jack Fogarty
PhD Student

T:  +61 4221 5547  |  E:  jf752 at uowmail.edu.au<mailto:jf752 at uowmail.edu.au>

Brain & Behaviour Research Institute
Psychophysiology Lab | School of Psychology | University of Wollongong
<image001.png><image002.png>

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202


--
Raghavan Gopalakrishnan,
Principal Research Engineer,
Cleveland Clinic
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20210418/f1ea17c5/attachment.htm>