<html><head></head><body><div style="font-family:Helvetica Neue, Helvetica, Arial, sans-serif;font-size:13px;"><div><div>Dear list mates:</div></div><div><br></div><div>Pascual-Marqui et al state on page 4:</div><div><div><br></div><div>"It is very important to emphasize a special property of Everson’s orthogonalization: the</div><div>orthogonal signals are unique, i.e. they are in a one-to-one correspondence with the original signals.</div><div><b>This is distinct from principle and independent components analysis (PCA and ICA), where there are </b><b>sign and permutation indeterminacies."</b></div><br></div><div><div>Pascual-Marqui, Biscay, Bosch-Bayard, Faber, Kinoshita, Kochi, Milz, Nishida, Yoshimura. Innovations orthogonalization: a solution to the major pitfalls of EEG/MEG “leakage correction”. 2017-08-20.</div><div><br></div><div>Robert</div><br></div><div><br></div><div><br></div><div id="yahoo_quoted_4643275670" class="yahoo_quoted"><div style="font-family:'Helvetica Neue', Helvetica, Arial, sans-serif;font-size:13px;color:#26282a;"><div>On Monday, August 28, 2017, 8:29:48 PM EDT, pascualmarqui@gmail.com <pascualmarqui@gmail.com> wrote:</div><div><br></div><div><br></div><div><div dir='ltr'>Dear Colleagues,<br></div><div dir='ltr'>The pre-print at:<br></div><div dir='ltr'><a href="https://doi.org/10.1101/178657" target=_blank>https://doi.org/10.1101/178657</a><br></div><div dir='ltr'>might be of interest to those working in the field of brain<br></div><div dir='ltr'>connectivity based on signals of electric neuronal activity. It is<br></div><div dir='ltr'>shown that "leakage correction" in the form of "signal<br></div><div dir='ltr'>orthogonalization" consistently produces false connectomes. More<br></div><div dir='ltr'>importantly, a new method is proposed for the resolution ("unmixing")<br></div><div dir='ltr'>of electrophysiological signals, based on "innovations<br></div><div dir='ltr'>orthogonalization".<br></div><div dir='ltr'><br></div><div dir='ltr'>--------<br></div><div dir='ltr'>Abstract : The problem of interest here is the study of brain<br></div><div dir='ltr'>functional and effective connectivity based on non-invasive EEG-MEG<br></div><div dir='ltr'>inverse solution time series. These signals generally have low spatial<br></div><div dir='ltr'>resolution, such that an estimated signal at any one site is an<br></div><div dir='ltr'>instantaneous linear mixture of the true, actual, unobserved signals<br></div><div dir='ltr'>across all cortical sites. False connectivity can result from analysis<br></div><div dir='ltr'>of these low-resolution signals. Recent efforts toward “unmixing” have<br></div><div dir='ltr'>been developed, under the name of “leakage correction”. One recent<br></div><div dir='ltr'>noteworthy approach is that by Colclough et al (2015 NeuroImage,<br></div><div dir='ltr'>117:439-448), which forces the inverse solution signals to have zero<br></div><div dir='ltr'>cross-correlation at lag zero. One goal is to show that Colclough’s<br></div><div dir='ltr'>method produces false human connectomes under very broad conditions.<br></div><div dir='ltr'>The second major goal is to develop a new solution, that appropriately<br></div><div dir='ltr'>“unmixes” the inverse solution signals, based on innovations<br></div><div dir='ltr'>orthogonalization. The new method first fits a multivariate<br></div><div dir='ltr'>autoregression to the inverse solution signals, giving the mixed<br></div><div dir='ltr'>innovations. Second, the mixed innovations are orthogonalized. Third,<br></div><div dir='ltr'>the mixed and orthogonalized innovations allow the estimation of the<br></div><div dir='ltr'>“unmixing” matrix, which is then finally used to “unmix” the inverse<br></div><div dir='ltr'>solution signals. It is shown that under very broad conditions, the<br></div><div dir='ltr'>new method produces proper human connectomes, even when the signals<br></div><div dir='ltr'>are not generated by an autoregressive model.<br></div><div dir='ltr'>--------<br></div><div dir='ltr'><br></div><div dir='ltr'>Cordially,<br></div><div dir='ltr'>Roberto<br></div><div dir='ltr'>...<br></div><div dir='ltr'>Roberto D. Pascual-Marqui, PhD, PD<br></div><div dir='ltr'>The KEY Institute for Brain-Mind Research, University of Zurich<br></div><div dir='ltr'>Visiting Professor at Neuropsychiatry, Kansai Medical University, Osaka<br></div><div dir='ltr'>[www.keyinst.uzh.ch/loreta] [scholar.google.com/citations?user=pascualmarqui]<br></div><div dir='ltr'>_______________________________________________<br></div><div dir='ltr'>Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" target=_blank>http://sccn.ucsd.edu/eeglab/eeglabmail.html</a><br></div><div dir='ltr'>To unsubscribe, send an empty email to <a ymailto="mailto:eeglablist-unsubscribe@sccn.ucsd.edu" href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu">eeglablist-unsubscribe@sccn.ucsd.edu</a><br></div><div dir='ltr'>For digest mode, send an email with the subject "set digest mime" to <a ymailto="mailto:eeglablist-request@sccn.ucsd.edu" href="mailto:eeglablist-request@sccn.ucsd.edu">eeglablist-request@sccn.ucsd.edu</a></div></div></div></div></div></body></html>