Hi Jiti,
I don't think there is anything inherently wrong in treating space x
time x frequency as a 4D or 5D topological map i.e clustering across
space,time and frequency simultaneously. However, using RFT over that 5D
space doesn't seem intuitive. On the other hand, deriving the p-values
for individual clusters in 5D using non-paramteric methods such as by
scrambling trials as suggested in the following (and its references):
http://www.ncbi.nlm.nih.gov/pubmed/11595276
seems like a reasonable thing to do. These p-values can then be subject to
FDR correction as suggested in the 2 references that you have listed.
Regards,
Hari
On Thu, March 31, 2011 8:26 am, Jiti K wrote:
> Hi,
>
> This post is regarding the applicability of topological FDR method (see
> the references 1,2 below) to EEG functional connectivity analysis that I
> am interested in trying for my study.
> My measure of functional connectivity is a phase locking value - computed
> from pair-wise EEG signals that are obtained using 10-20 system. To
> compute the phase locking value, the signals were decomposed in a wavelet
> domain. My objective is to detect the electrode pairs, frequency bands,
> and time-intervals with significant phase locking value. My analysis
> accounts for possible electrode pairs in a pre-specified range of
> frequency-bands and time-intervals.
> I do not have structural MRI data for subjects.
> If it is at all possible to topologically reconstruct EEG signals without
> the structural scans, I am not sure how well the good lattice assumption
> that is required by random field theory will hold for such low-density
> EEG. Moreover, as I have understood the method can be either applied to
> topological space or over time, or over frequency but not for all the
> three dimensions simultaneously.
> If there is some way this method can be applied/extended to the kind of
> analysis I mentioned above, I would be interested in getting some inputs
> from you.
>
> 1) Chumbley J, Worsley K, Flandin G, Friston K., Topological FDR for
> neuroimaging. Neuroimage. 2010 Feb 15;49(4):3057-64
> 2) James M. Kilner and Karl J. Friston. Topological inference for EEG
> and MEG. Ann. Appl. Stat. Volume 4, Number 3 (2010), 1272-1290
>
> Thanks in advance,
>
> Jiti
>
>
>
--
Hari Bharadwaj
--
Hari Bharadwaj
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
|