Minding the theory or mining the data: EEG and neural oscillations: presented to Oscillatory Brain Dynamics Conference, Arizona State University, Department of Psychology Cognitive Science program and EEG Lab, 2016.

1. TO MINE OR TO MIND:
A PRIMAL VIEW ON THE SPATIOTEMPORAL DYNAMICS OF NEURAL OSCILLATIONS
Emmanuelle Tognoli
Center for Complex Systems and Brain Sciences, Florida Atlantic University
May 7th, 2016, ASU

2. DISCLAIMER
This presentation will heavily rely on normal color perception
(disorder incidence: ~1/10-1/12 male, fewer female)

3. PREAMBLE DEFINITION
METASTABILITY: a regime of coordination characterized by
tendencies to integrate (dwells, quasi-phase locking)
interspersed with tendencies to segregate (escapes, quasi-
independence)

4. Movement #1
observation
Movement #2
theory

5. Movement #1
observation
Movement #2
theory

6. A journey into oscillatory anecdotes

7. Brain and behavior, a rosetta stone?
1 hour of waking EEG(delta-to-gamma) = O(1 million patterns)
Tognoli,Benites,Kelso,submitted

8. Observing oscillations: goals
Look at oscillations visually and be able to say something
about:
-Which neural ensembles (proxy: which parts of the brain)
-How do they associate?
-How do they organize temporally
-Across frequencies?
[-For which (dys)function?]
Address all idiosyncrasies: inter-individual, intra-
individual, inter-temporal
x (multiple frequency bands)
(filtered in the “alpha” band)

9. Oscillatory patterns: a taxonomy
Tognoli, Benites,
Kelso, submitted
Truncated
dipole (A)
Full dipole (B)
Key observations:
-Relative simplicity of
patterns, lot of variance
can be explained by few
(here one) neural
ensemble(s)
-Typical duration, 1-2 cycles
(waking, active state)

10. How do they associate?
Tognoli,Benites,Kelso,submitted
Key observations:
-three types of
associations:
coexistence (C),
metastability (D,
alternation phase
escape/phase
attraction), and
phase-locking (E-
G, under various
relative phase, in
agreement with
theory of BCD)

11. Repeatability
Tognoli, Benites, Kelso, submitted
Key observations:
-neural patterns recur (e.g. 1), repetition may preserve
phase relationship between “phase aggregates”

12. Temporal organization
Tognoli,Benites,
Kelso,submitted
Key observations:
-intermittency, rather than continuous amplitude modulation
-succession often in brief, smooth switches between patterns (top), evocative
of metastability (no attractor, no need for scattering)
-sometimes phase scattering (bottom), evocative of phase locking (attractors)
Tognoli&Kelso,
ProgressNeurobiol.2009

13. Repeatability
Tognoli, unpublished
Key observations:
-duration of patterns vary but preserved successions are
sometimes observed

14. Across frequencies
Parsing all the frequencies to identify meaningful oscillations
in space and time

15. Across frequencies
Frequency scanner:
identify bands carrying largest amplitude
(least filter attenuation)

16. Across frequencies
Key observations:
In noninvasive EEG, neural ensembles are sparse,
between frequencies.
They exhibit spatiotemporal metastability(*)
Tognoli & Kelso, Neuron, 2014

17. Tie to functions
[This page left intentionally blank]
(an heterologism)

18. Movement #1
observation
Movement #2
theory

19. Movement #1
observation
Movement #2
theory

20. An equation of coordination dynamics
Haken et al., 1985; Kelso et al., 1990
Brandon Kidwell

21. Crash course on phase and relative phase
time
x
x=0
x=1.
x=1
x=0.
x=0
x=-1.
x=-1
x=0.
x=0
x=1.

22. Symmetry and symmetry breaking
Haken et al., 1985; Kelso et al., 1990
2 regimes:
Bistable | Monostable
Symmetry Broken Symmetry
3 regimes:
Bistable | Monostable | Metastable
Inphase, antiphase, “out of phase”

23. Weak & Broken
Haken et al., 1985; Kelso et al., 1990
Metastability arises for:
-weak coupling -broken symmetry

24. Spatiotemporally
Kuramoto & Battoghotk, 2002
Tognoli & Kelso, Neuron, 2014
SpatioTemporal
metastability
reveals neural
ensembles that:
T-last longer or
shorter periods
depending on
ensemble Size
and at the same time
S-involve more or
fewers oscillators
over Time

25. Wrapping it up

26. Evidence of Metastability: Broken
Tognoli & Kelso, Arxiv, 2013
With oscillations at many
frequencies in the brain,
examples of coordination with
broken symmetry abound.

27. Metastability creates macroscopic “intermittency”
Metastability
transforms a
continuous
(micro- meso-
scopic) dynamics
into intermittent
(meso- macro-
scopically)
Tognoli & Kelso, Neuron, 2014

28. Electromagnetic fields across scales
Key control parameter: phase alignment (in our theory: a spatial
gain parameter, wherein phase locking plays a role of integrator)
Tognoli&Kelso,Neuron,2014

29. Evidence of Metastability: Weak
Model
Experimental data
- Weak coupling – spontaneous emergence of coupling and
independence at multiple frequencies

30. Conclusion

31. In between
Key model of brain function
in cellular neuroscience
Important model of brain function
In system neuroscience (EEG)

32. Perspective
Tognoli & Kelso, Frontiers in System
Neurosciences, 2015
Flows of information
&
Mass action

33. Acknowledgments
Scott Kelso
Daniela Benites
Rodrigo Calderon
Julien Lagarde
Gonzalo de Guzman
Armin Fuchs
Slava Murzin

34. Understanding the workings of the brain: EPs cons
Con#1: the average is not a good reflection of the particular
Con#2: bigger are winners
Con#3: physically unsound
Con#4: conveys a certain idea of information processing (*)

35. Space
x,y,(z)
Wave
amplitu
de
v
Time
t
Colorimetric approach to see pattern in 4D

36. Colorimetric approach to see pattern in 4D
v
t
x
yspace
wave
time
x
y

abscis
sa
“wave
length”
(colorimetric
model)
ordinates

37. m
f
(m)
(f)
Relative