% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@INPROCEEDINGS{Ito:1009393,
author = {Ito, Junji and Gutzen, Robin and Krauße, Sven and Denker,
Michael and Grün, Sonja},
title = {{T}owards classification of spatio-temporal wave patterns
based on principal component analysis},
reportid = {FZJ-2023-02797},
year = {2023},
abstract = {Spatio-temporal oscillatory dynamics are found in a variety
of subjects in the natural sciences. [1]They are
mathematically described in terms of a complex-valued field
variable Z(r, t), from whichone can uniquely derive the
oscillation amplitude A(r, t) = |Z(r, t)| and oscillation
phase θ(r, t) = argZ(r, t), as functions of location r and
time t. In a wide class of systems, the phase variable
exhibitsspecific spatio-temporal patterns, such as planar
wave, radial wave, rotating wave, and so on.These patterns
have also been observed in the cerebral cortex of the brain
as spatio-temporal waves(STWs) of local field potential
(LFP) signals [2-5]. Previous studies have suggested that
specific phasepatterns, in particular planar waves, are
related to the coordination of spiking activity of
singleneurons, and therefore might play a fundamental role
in neuronal information processing [6-8].Studying the
implications of the STWs for brain function requires a
systematic classification methodto group given phase
patterns into distinct wave types (e.g. planar, radial, and
rotating). Thestrategies taken in previous studies rely on
defining a characteristic measure quantifying a feature
ofthe phase variable θ(r, t) for each wave type, and
setting a threshold on this measure to assign awave type to
an episode of data. An inherent shortcoming of this approach
is that it requires the adhoc and eventually arbitrary
selection of characteristic measures and thresholds.Here we
propose a method to quantify phase pattern characteristics
based on principal componentanalysis (PCA), which can be
used for a non-parametric classification of wave types. In
addition tothe standard PCA, we also employ the complex PCA,
which works on a complex-valued data matrixand decompose it
into components represented by complex-valued vectors (see
Figure). We showthat the principal components (PCs) obtained
via the complex PCA can naturally represent
phaserelationships among variables. We apply both methods to
Utah array recordings of LFPs from themacaque motor cortex,
which has been reported to exhibit various types of wave
patterns, anddiscuss the commonalities and differences
between the PCs obtained by the two methods.Furthermore, we
relate the time course of the obtained PCs to the time
course of the characteristicmeasures of wave types, which
were used in previous studies, and examine how individual
PCscorrespond to one or multiple of the characteristic
measures.We thereby employ the phase pattern quantification
with (the standard or complex) PCA as analternative method
of wave type classification. Further, decomposing the
cortical waves into“eigenmodes” and studying their
relations to neuronal and behavioral covariates would
provide apromising approach for investigating the functional
implications of the waves.References1. Winfree (1980) The
geometry of biological time. Vol. 2.2. Ermentrout et al.
(2001) Neuron 29(1):33–44. doi:
10.1016/S0896-6273(01)00178-73. Heitmann et al. (2012)
Front. Comput. Neurosci. 6:67. doi:
10.3389/fncom.2012.000674. Denker et al. (2018) Sci. Rep.
8(1):5200. doi: 10.1038/s41598-018-22990-75. Townsend et al.
(2018) PLoS CB 14(12):e1006643. doi:
10.1371/journal.pcbi.10066436. Takahashi et al. (2015) Nat.
Commun. 6(1):1–11. doi: 10.1038/ncomms81697. Vinck and
Bosman (2016) Front. Syst. Neurosci. 10:35. doi:
10.3389/fnsys.2016.000358. Davis et al. (2020) Nat. Commun.
12(1):6057. doi: 10.1038/s41467-021-26175-1},
month = {Jul},
date = {2023-07-15},
organization = {32nd Annual Computational Neuroscience
Meeting, Leipzig (Germany), 15 Jul 2023
- 19 Jul 2023},
subtyp = {After Call},
cin = {INM-6 / IAS-6 / INM-10},
cid = {I:(DE-Juel1)INM-6-20090406 / I:(DE-Juel1)IAS-6-20130828 /
I:(DE-Juel1)INM-10-20170113},
pnm = {5231 - Neuroscientific Foundations (POF4-523) / HBP SGA2 -
Human Brain Project Specific Grant Agreement 2 (785907) /
HBP SGA3 - Human Brain Project Specific Grant Agreement 3
(945539) / HAF - Helmholtz Analytics Framework (ZT-I-0003) /
JL SMHB - Joint Lab Supercomputing and Modeling for the
Human Brain (JL SMHB-2021-2027) / Algorithms of Adaptive
Behavior and their Neuronal Implementation in Health and
Disease (iBehave-20220812)},
pid = {G:(DE-HGF)POF4-5231 / G:(EU-Grant)785907 /
G:(EU-Grant)945539 / G:(DE-HGF)ZT-I-0003 / G:(DE-Juel1)JL
SMHB-2021-2027 / G:(DE-Juel-1)iBehave-20220812},
typ = {PUB:(DE-HGF)6},
url = {https://juser.fz-juelich.de/record/1009393},
}