Critical Clinical

REVIEW
published: 26 August 2020

doi: 10.3389/fncir.2020.00054
Why Brain Criticality Is Clinically

Relevant: A Scoping Review
Vincent Zimmern*
Division of Child Neurology, The University of Texas Southwestern Medical Center, Dallas, TX, United States
The past 25 years have seen a strong increase in the number of publications related
to criticality in different areas of neuroscience. The potential of criticality to explain
various brain properties, including optimal information processing, has made it an
increasingly exciting area of investigation for neuroscientists. Recent reviews on this
topic, sometimes termed brain criticality, make brief mention of clinical applications of
these findings to several neurological disorders such as epilepsy, neurodegenerative
disease, and neonatal hypoxia. Other clinicallyrelevant domains – including anesthesia,
sleep medicine, developmental-behavioral pediatrics, and psychiatry – are seldom
discussed in review papers of brain criticality. Thorough assessments of these
application areas and their relevance for clinicians have also yet to be published. In
this scoping review, studies of brain criticality involving human data of all ages are
evaluated for their current and future clinical relevance. To make the results of these
studies understandable to a more clinical audience, a review of the key concepts
behind criticality (e.g., phase transitions, long-range temporal correlation, self-organized
criticality, power laws, branching processes) precedes the discussion of human clinical
studies. Open questions and forthcoming areas of investigation are also considered.
Edited by:
Ioanna Sandvig, Keywords: criticality, brain, long-range temporal correlation, neurodevelopment, neurodegeneration, sleep,
Norwegian University of Science epilepsy, anesthesia
and Technology, Norway
Reviewed by:
Adriano Barra, INTRODUCTION
University of Salento, Italy
Stephen M. Johnson, The brain criticality hypothesis suggests that neural networks and thus, many aspects of brain
University of Wisconsin-Madison, activity self-organize into a critical state (Wilting and Priesemann, 2019). Critical states are unique
United States configurations of physical systems that have been a central focus of statistical physics for more
*Correspondence: than a century. Criticality, which is a synonymous term for “critical phenomena” or “critical states,”
Vincent Zimmern marks the transition between ordered and disordered states. The theory of critical phenomena
vincent.zimmern@utsouthwestern.edu has found applications in many scientific fields, including neuroscience and clinical neurology
(Sornette, 2004; Cocchi et al., 2017). In the neurosciences, criticality is appealing because theory
Received: 24 May 2020
and modeling suggest that neural networks at criticality exhibit optimal processing and computing
Accepted: 23 July 2020
properties (Beggs, 2008; Shew and Plenz, 2013). These properties include information transmission,
Published: 26 August 2020
information storage, dynamic range, metastable states, and computational power (Maass et al.,
Citation:
2002; Bertschinger and Natschläger, 2004; Latham and Nirenberg, 2004; Haldeman and Beggs,
Zimmern V (2020) Why Brain
Criticality Is Clinically Relevant:
2005; Kinouchi and Copelli, 2006; Tanaka et al., 2009; Boedecker et al., 2012; Shriki et al., 2013;
A Scoping Review. Gautam et al., 2015; Shriki and Yellin, 2016; Hoffmann and Payton, 2018). The term “brain
Front. Neural Circuits 14:54. criticality” is used in this review as a catch-all term for the various manifestations of critical
doi: 10.3389/fncir.2020.00054 phenomena in human neuroscience.
Frontiers in Neural Circuits | www.frontiersin.org 1 August 2020 | Volume 14 | Article 54

Zimmern Clinical Relevance of Brain Criticality
This scoping review of the brain criticality literature will start phase transitions occur over a parameter range called a Griffiths
with a section that reviews the physics-based ideas behind brain phase rather than at a single critical point and are sometimes
criticality, including a discussion of the Ising model. It will then referred to as quasi-critical states (Moretti and Munoz, 2013).
review the brain criticality literature in each of seven domains Quasi-critical states obey many of the properties of a critical
(i.e., anesthesia, epilepsy, neurodegeneration, neurodevelopment, system but are not entirely critical.
cognition, sleep medicine, and psychiatry), focusing on the These concepts can be illustrated with the help of the two-
clinical applications. This review should serve as an entry-point dimensional Ising model, which is a classic example of a critical
for clinicians and translational researchers interested in using transition in ferromagnetism (Beggs and Timme, 2012). The
this conceptual framework and its associated tools to advance Ising model consists of a lattice in a piece of iron, with each
patient care. Understandably, it is not possible to summarize site of the lattice corresponding to a dipole moment (i.e., an
nearly a century of developments in statistical physics − not to up or down spin) (Figure 1B). Each dipole moment operates
mention 30 years of applications of these physical concepts to like a bar magnet and can influence its nearest neighbors to
neuroscience−in a single review paper. An excellent introductory align in the same direction. At a low temperature, nearest-
paper for newcomers to criticality is Beggs and Timme, 2012. neighbor effects will dominate the system. At a fixed temperature
References are available for those interested in diving into the below the critical temperature, a cluster of aligned spins will
technical details of critical phenomena (Nishimori and Ortiz, get larger and larger, and, with time, will take over the entire
2011; Tomko et al., 2018), including two books on criticality in lattice to make a uniform dipole moment (Figure 1A). Thus,
neural dynamics (Plenz and Niebur, 2014; Tomen et al., 2019). at this low temperature, the piece of iron will behave like a
In this section, criticality is introduced from the perspective of magnet because the spins will align throughout the lattice to
phase transitions and is illustrated using the two-dimensional yield a strong net magnetization (i.e., order parameter). However,
Ising model (see Figure 1). This is followed by a discussion as temperature (i.e., control parameter) increases, the energy
of the various characteristics of critical phenomena and a brief from heat begins to jostle the spins. Past a critical temperature
discussion of self-organized criticality. called the Curie temperature (Tc), the disordered spins from the
added heat will overwhelm the ordering effect of the nearest-
neighbor interaction, leading to a loss of the magnetization
CRITICALITY AND ASSOCIATED (Figures 1A–C). Heat, therefore, takes the system from a
CONCEPTS subcritical, magnetic phase through a critical phase transition and
on to a supercritical, non-magnetic and disordered phase. At the
In statistical physics, criticality refers to the behavior that is critical temperature, a critical phase emerges where order and
seen when a physical system undergoes a specific kind of phase disorder are evenly matched (Nishimori and Ortiz, 2011; Beggs
transition. Typically, a macroscopic property of the system called and Timme, 2012). The correlation length (i.e., how far a single
the order parameter changes as a function of an underlying spin change can propagate through the system) is maximized
feature called the control parameter. In the example of a vapor- in this phase (Figure 1D), and in the infinitely large system,
to-water phase transition, the order parameter is the macroscopic goes to infinity at the critical point. The order parameter (i.e.,
appearance reflective of the phase’s entropy (i.e., water or vapor), magnetization)−along with other observables like magnetization
and the control parameter is the temperature (Hesse and Gross, domain size and magnetic susceptibility−become power-law
2014). Generally, gradual changes in the control parameter lead distributed with unique power-law exponents (Figure 1F). Power
to similarly gradual changes in the order parameter. However, at laws refer to a probability density function of the form of p(x) =
specific points, the order parameter changes abruptly. On a plot C x−α for some x > xo and with αcorresponding to the power-law
of the order parameter on the y-axis and the control parameter exponent. Power laws exhibit scale invariance and are therefore
on the x-axis (i.e., a phase diagram), there is either a jump (i.e., a called scale-free. A function f (x) is scale invariant if f (cx) ∝
discontinuity) or a sharp corner (i.e., a non-differentiable point) f (x), where α signifies “proportional to.” In other words, scaling
at the transition point of a phase transition. If the change in the the argument of the function is equivalent to a proportional
order parameter is a jump, the phase transition is known as first- scaling of the function itself. In the case of the power-law,
order or discontinuous. If the change in the order parameter is f (cx) = (cx) −α
= c−α x−α =c−α f(x) ∝ f (x). Moreover, because
log f (x) = log ∝ (x−α ) = −αlog (x), a log-log plot of a power-

a sharp corner, the phase transition is known as second-order or
continuous. The second-order phase transition allows a system law distributed dataset should produce a straight line with slope
to be at the exact transition point at the interface between two −α. Caveats on using this log-log plot technique to extract the
very different states, with usually one state being more disordered power-law exponent are addressed in a subsequent section.
than the other. The system is said to be “at criticality” or in the The Ising model can lead to very complex behavior patterns
critical state (Nishimori and Ortiz, 2011; Hesse and Gross, 2014). and has been used to model neural networks (Fraiman et al.,
The phase in which the control parameter is below the critical 2009; Beggs and Timme, 2012; Deco et al., 2012; Marinazzo
value is called the subcritical phase, while the phase in which et al., 2013; Stramaglia et al., 2017). The Ising model and
the control parameter is above the critical value is called the several other well-characterized models have led to a better
supercritical phase. It is important to note that the theory of phase understanding of how critical systems behave. Nevertheless,
transitions usually involves systems at the thermodynamic limit, critical systems remain difficult to identify because the relevant
i.e. infinitelylarge volume. When not at the thermodynamic limit, order and control parameters may not always be readily available

FIGURE 1 | Continued

FIGURE 1 | Illustration of critical phase transition using the two-dimensional (2D) Ising model. For more information on the Ising model, see Nishimori and Ortiz
(2011), Beggs and Timme (2012). (A) Simulation of 2D Ising model with length = 256 in subcritical, critical, and super-critical states as temperature increases from
left to right. Black and white areas represent magnetization domains with differing spins. Simulations were generated using open-source code from Matt Bierbaum
(mattbierbaum.github.io/ising.js) using the Metropolis algorithm. (B) Illustration of 2D Ising spin lattice to show differing spin states during phase transition. Image
adapted with permission from Beggs and Timme (2012). The organized spins in the subcritical state give way to random spin arrangements in the supercritical state,
passing through an intermediate critical state with a complex arrangement of spins. (C) The order parameter of the system decreases smoothly as the control
parameter increases, until it abruptly changes at the critical temperature (Tc or the Curie temperature). (D) The correlation length is maximized at the critical point.
(E) A branching parameter of 1 allows an aggregate one-to-one transmission of neural signals. Branching parameter greater than 1 leads to supercritical, run-away
neuronal excitation. Branching parameter less than 1 leads to subcritical dying-off of the neuronal transmission. (F) At the critical point, observables of the system,
including the order parameter, obey a power-law or scale-free distribution, seen as a straight-line on a log-log plot. The order parameter remains constant in the
subcritical case, while it drops off exponentially in the supercritical state.
for experimentation, making it difficult to construct a complete et al., 2012). This phenomenon is not to be confused with flicker
phase diagram. In the absence of a complete phase diagram, one noise or 1/f noise, which is discussed next.
can use several known markers of criticality, as follows:
The Emergence of Power-Law (1/f) Noise
Branching Parameter and Power-Law Observables
A branching parameter σ, in the setting of brain criticality, is Critical systems, when perturbed by weak inputs, will exhibit
the ratio of downstream activated neurons to upstream activated superposed geometric responses to the inputs, which yields
neurons (Harris, 1963). In other words, as in Figure 1E, a 1/f noise, also called pink noise, power-law noise, or flicker
branching parameter of 1 means that every activated neuron noise (see Figure 3C). Many authors employ these terms
on average fires or activates one other downstream neuron synonymously with long-range dependence or long-memory, as
(Hobbs et al., 2010). Branching parameter less than 1 indicates these are identical phenomena. The term 1/f noise refers to
a subcritical phase that evolves with time to a quiescent, the phenomenon in which the power spectrum S(f ) of a time
inactive state. On the other hand, a branching parameter series obeys a power law of the formS(f ) = α f −β . Historically,
greater than 1 indicates a supercritical phase of increasing the cases of β = 0, β = 1, β = 2 are referred to as “white”
activity. Of course, branching parameters are variable and noise, “pink” noise, and “brown” noise, respectively (Li et al.,
dynamic. Caution is needed in the interpretation of the 2005). The range 0.5 < β < 1.5 is commonly accepted as 1/f
branching parameter, however, because a branching parameter noise. While all critical systems should exhibit 1/f noise, not
of 1 can also be observed in certain supercritical states all 1/f noise is indicative of criticality (Bédard et al., 2006;
(Hesse and Gross, 2014). Hesse and Gross, 2014).
As mentioned earlier, multiple observables will follow a
power-law distribution when in a critical state. However, power
Long-Range Temporal Correlation, laws are necessary but not sufficient to prove the existence
Critical Slowing, and Flickering of criticality – many other processes can generate a power-
In critical systems, the response of the system to external law distribution. For more on this topic, see the following
stimuli – called the dynamic range or dynamic correlation – is references (Mitzenmacher, 2004; Newman, 2005; Touboul and
maximized. Small perturbations of the system at criticality lead Destexhe, 2010). Demonstrating the existence of a power law
to geometric (rather than exponential) returns to the steady- is not straightforward. As a rule, power laws take the form of
state (Hesse and Gross, 2014), leading to long-range temporal Y = C x−α for some x>xo . For many years, it was common
correlation (LRTC or long-memory). One can measure LRTC practice to plot y against x (from the previous formula) in
in multiple ways. Popular methods include the Hurst exponent log-log coordinates – if a straight line emerged, the slope of
(through various estimators) and detrended fluctuation analysis that line was interpreted as the power-law exponent α (see
(DFA) (Peng et al., 1995a,b; Simonsen et al., 1998; Hardstone Figure 3B). This approach led to many false claims of power-
et al., 2012). DFA produces a scaling exponent over a defined laws since many other finite datasets (e.g., from a log-normal
time period (see Figure 2 for an illustration of DFA). If that distribution) can also approximate a straight line on a log-
scaling exponent is between 0.5 and 1, with a good fit (see log plot. Since Clauset et al. (2009) and subsequent work,
Figure 2), one can conclude that the time series exhibits LRTC researchers have had access to more sophisticated statistical
over that time period. methodology to argue that their variable obeys a power law, as
This geometric rate of return to steady-state is also called opposed to other heavy-tailed distributions. While definitions
critical slowing down (Scheffer et al., 2009; Van De Leemput vary, heavy-tail distributions are probability distributions whose
et al., 2014). More generally, at the critical point, the dynamic tails are “heavier” than the exponential distribution, of which
correlation of the system diverges such that avalanches (i.e., the Gaussian is a sub-type. Examples include the Fisher-Tippett
network activity) occur at all scales of the system (Hesse and (double-exponential) distribution, the log-normal distribution,
Gross, 2014). Another phenomenon seen at the critical transition the Weibull distribution, among many others. An exhaustive
is called flickering, which emerges when noise allows a system review of several heavy-tailed distributions and their role in
to migrate back-and-forth between two attractor basins (Wang neuroscience can be found in Roberts et al. (2015). In this

FIGURE 2 | Figure legend taken from Moran et al. (2019) and reproduced with permission. This figure depicts the steps of the detrended fluctuation analysis (DFA).
(A) 200 s EEG recording that has been band-pass filtered (13–25 Hz, channel 01 = Cz) in a patient with schizophrenia. (B) A close-up view (9 s) of the band-pass
filtered EEG along with the amplitude envelope derived from the Hilbert transformation (in green), which is used for estimating the scaling exponent. (C–D) Two
examples of the amplitude envelope with two window sizes. In the left graph, no detrending (i.e., removal of the trend line) has been applied whereas in the right
graph, detrending has been applied. For each time window size, the fluctuation of the detrended signal is calculated as the mean standard deviation over all identical
sized segments (E) The DFA scaling exponent is given by the slope of the log-log plot between fluctuation F and window size. A DFA scaling exponent between 0.5
and 1 indicates the presence of LRTC – in this case, the correlation extends up to 20 s. In this example, the DFA scaling exponent for the beta-band-oscillations in
this channel is α = 0.628. It is common to report the R2 value of the linear regression of the log-log plot as a goodness-of-fit score.
review, publications that claim a power-law distribution based on exponents of these power laws are inter-related. The
current practices are contrasted with publications that rely on the details of these mathematical relationships are beyond
previously accepted “log-log approach.” the scope of this review but represent a fascinating topic
in their own right. To give but one example, avalanche
Relationship of Power-Law Exponents to size distribution with power exponent α and avalanche
Each Other lifetime distribution with power exponent β should be
In systems at criticality, many observable variables (e.g., related to avalanche lifetime γ by the following equation:
β−1
correlation, size distribution) obey power laws. The different γ = α−1 . This relationship has been experimentally

FIGURE 3 | (A) The sandpile model is a classic model of self-organized criticality (SOC), derived as a though experiment (Bak et al., 1988) and proven later with rice
piles (Frette et al., 1996). Sand is dropped continuously at a fixed rate onto a flat surface. A sandpile forms. As the sand is added, the slope of the pile increases and
avalanches of sand grains begin to occur. If the slope exceeds a “critical slope,” small and large avalanches will decrease the slope back closer to the critical slope.
(B) At the critical slope, the distribution of avalanche sizes (i.e., number of sand grains) is power-law distributed. On a log-log plot, the slope of a
power-law-distributed variable gives the power-law exponent. Most avalanches are small but a non-negligible number are quite large, up to nearly the size of the
entire system. (A,B) reused with permission from Hesse and Gross (2014). (C) Neurophysiologic measurements such as voltage from an EEG recording can be
displayed as time series that exhibit characteristic spectral densities S(f). (C) Figure reused with permission from Scholarpedia and E. Izhikevich. Plot of different
color-coded time series (left) and a log-log plot of their respective spectral densities (right). In Gaussian “white” noise, each frequency has equal energy, leading to a
constant spectral density. “Pink” noise, which is characteristic of SOC but arises in many other settings, has a spectral density given by 1/f, thus the name “1/f
noise.” “Brown” noise refers to Brownian or random motion, whose spectral density is typically given by 1/f2 .

FIGURE 4 | Avalanches recorded from neural tissue. (A) Micrograph of a cortical slice on an electrode array, represented by a black rectangle. (B) Electrical potential
recorded from an electrode. Arrow points to a spike from an individual neuron. The spike is expanded in panel (C) and marked by a black dot. (D) Raster plot of
spike times (dots) from multiple neurons over time. (E) Expanding the view of the raster plot reveals an avalanche. Each frame represents the electrode array during a
single 5 ms period. Black dots are spikes detected on the array. An avalanche is made up of consecutively active frames, beginning and ending with inactive frames.
(F) Plotting the number of spikes in each frame versus time produces an avalanche shape. (G) Avalanche shapes are produced by averaging the temporal profiles of
all avalanches of a particular duration from an experimental dataset close to criticality. Different colors here represent different durations. (H) The collapses are plotted
by rescaling the horizontal and vertical axes. The tight overlap after rescaling is indicative of criticality. Images reproduced with permission Beggs and Timme (2012)
and Friedman et al. (2012).
validated in individual neurons (Beggs and Timme, 2012; observed in many critical systems (Perković et al., 1995; Mehta
Friedman et al., 2012). et al., 2006; Papanikolaou et al., 2011).
There are various theories as to how physical systems can
Scaling Function bring themselves to criticality. In the case of the brain, one of
Because of the self-similar or fractal nature of the avalanches of the popular theories is called self-organized criticality or SOC
activity in critical systems, the “shape” of avalanche activity is also for short (Bak et al., 1987; Christensen et al., 1992; Hesse
expected to behave as a fractal (see Figures 4A–G). Therefore, and Gross, 2014). It is also referred to as self-organized quasi-
all cascades of activity at criticality ought to be re-scalable to criticality (Bonachela and Muñoz, 2009) since, as mentioned
a unique shape, as a function of time (or duration) and power earlier, true criticality occurs only in infinitelysized systems.
exponents (see Figure 4H). This phenomenon of critical systems Introduced initially by Bak et al. (1987), the idea of SOC is that
allows a “data collapse” or “shape collapse” of all activity onto the control parameter is constantly being adjusted to the critical
a single unique shape (Friedman et al., 2012). In some neural value by a decentralized feedback mechanism. In other words, the
avalanches, this shape takes the form of an inverted parabola control parameter spontaneously decreases when the system is in
(Beggs and Timme, 2012). This kind of shape collapse has been a supercritical phase and increases when in a subcritical phase.

FIGURE 5 | PRISMA-ScR flow chart.
The use of the term “control parameter” is maintained, even if though the exact details remain unclear. A recent experiment,
the parameter is not being controlled externally but rather by the however, has cast some light on this question. Rat cortical
system itself (Hesse and Gross, 2014). Figure 3A illustrates one of networks exhibiting criticality in controlled circadian conditions
the earliest models of SOC, called the sandpile model (Bak et al., lost their network criticality when deprived of visual inputs.
1988; Frette et al., 1996). Imagine that sand is accumulating at a However, signs of criticality resumed in under 48 h with neuronal
certain rate on a flat surface. As the sandpile rises, the slope of the firing rates being maximally inhibited. This finding suggests that
sandpile increases. At a critical point, the distribution of sand- homeostatic regulation of inhibition plays an important role in
pile avalanche sizes (i.e., a few grains of sand or the entire pile) generating criticality (Ma et al., 2019).
obeys a power-law. With a power-law or scale-free distribution,
avalanches of all scales can occur, but most avalanches will be
small with a few being very large (Figure 3B). As sand continues ARTICLE SELECTION
to be added at the same rate, larger avalanches take place and
the slope decreases back to the “critical slope.” The criticality in The preceding review was conducted to set the stage for a scoping
this system is therefore self-organized and the control parameter review of the literature (Colquhoun et al., 2014) using PRISMA-
does not have to be tuned externally, as it does for example in ScR methodology (Moher et al., 2009). A scoping review
the Ising model. format was chosen since the nature of this literature is large,
The internal “tuning” of SOC models for the brain is heterogeneous, and not amenable to a more precise systematic
reminiscent of the concept of homeostatic plasticity. Homeostatic review. See Figure 5 for a schematic of the article selection
plasticity refers to the capacity of neurons to regulate their process. PubMed and Web of Science databases were searched
excitability based on network activity (Turrigiano and Nelson, from their inception until March 2020. The search terms were:
2004). Criticality seems to be connected to homeostatic plasticity “criticality anesthesia,” “criticality brain,” “criticality epilepsy,”

neurodevelopment, cognition, psychiatry, and sleep medicine.

Figure 6 summarizes the changes in LRTC along a subcriticality–
supercriticality spectrum that are thought to occur in disease
states belonging to these seven clinical categories.
CLINICAL APPLICATIONS OF BRAIN

CRITICALITY
Anesthesia
The literature on criticality in anesthesia is limited but has already
shown signs of significant clinical promise. From this small
body of research, one can extract two areas of clinical relevance:
markers for the depth of anesthesia and predictors of recovery
from a persistent comatose state. This literature harmonizes well
overall with that of sleep medicine (see Sleep Medicine below).
Three studies have reported on criticality-based markers
of the anesthesia depth (see Table 1). An electrocorticogram
(ECoG) study of patients undergoing anesthesia for surgical
removal of epileptic foci showed that the critical eigenvalues of
matrices obtained from a vector auto-regressive (VAR) model
could potentially serve as a marker of the depth of anesthesia
(Alonso et al., 2014). A scalp electroencephalogram (EEG) study
of patients under anesthesia found that LRTC, in combination
with oscillation amplitude, could help differentiate between
consciousness and anesthesia-induced unconsciousness (Thiery
et al., 2018). The authors found that, under anesthesia, LRTC
increases in the beta frequency range in frontocentral channels.
This increase may reflect decreased neuronal excitability, leading
to signal persistence and a resulting limitation on cognitive
FIGURE 6 | These plots illustrate long-range temporal correlation (LRTC) in processes (He, 2014; Thiery et al., 2018). The results of
different clinical areas as a function of criticality. Figure adapted from Colombo this study contradict a previous study that had found a
et al. (2016) with permission. The black parabolic curve represents the decrease in LRTC during anesthesia (Krzemiński et al., 2017).
dynamics of long-range temporal correlation from the subcritical state,
However, the contradiction may be secondary to significant
reaching a theoretical maximum at the critical point, and decreasing in a
supercritical state. The gray blurred rectangle represents the physiological differences between the experiments. Thiery et al. (2018)
range of brain dynamics, which is thought to be near or below the critical examined human subjects undergoing light anesthesia with
point. Black arrows (excluding the x- and y-axes) represent deviations toward sevoflurane, while Krzemiński et al. (2017) examined macaques
subcriticality (to the left) or toward supercriticality (to the right) with associated undergoing deep anesthesia with different anesthetic agents.
decreases in correlation along the black parabolic curve. Double arrows imply
that evidence for both increases and decreases (or absence) in LRTC have
These differences suggest that future research should focus on
been reported in that disease process. See the relevant tables for further identifying criticality-based metrics (e.g., power-law estimation,
detail. VAR eigenvalues, LRTC) that are specific to each anesthetic agent
at different depths of anesthesia.
Another area of ongoing research is the use of criticality
“criticality neural,” “criticality neurology,” “criticality seizure,” signatures to distinguish persistent coma from other forms of
and “criticality sleep.” A snowball approach identified additional unconsciousness, including deep sleep and anesthesia-induced
studies that were not captured by the search terms. The reviewer unconsciousness. A function magnetic resonance imaging
screened titles and abstracts for English language original (fMRI) study compared both blood oxygen level-dependent
articles with full-text availability. The following manuscripts were (BOLD) signals and brain networks between patients undergoing
excluded: (1) reviews of a general nature, conference proceedings, anesthesia and patients in unresponsive wakefulness syndrome
and publications offering new theories but without empirical (UWS or persistent coma) (Liu et al., 2014). While the node
evidence, (2) papers based on non-human data, on connectome size in these brain networks was power-law distributed in
data, or mainly on modeling or simulation, (3) duplicates and both patient groups, the node degree distribution (i.e., the
studies not relevant to brain criticality. The remaining articles number of connections between nodes in the brain network) was
were then evaluated for eligibility based on clinical relevance. only power-law distributed in the healthy patients undergoing
Ultimately, seventy-eight studies met study criteria for the anesthesia. The authors offer a helpful analogy from the world of
scoping review. These articles were then analyzed according to airports and air traffic control. In patients under anesthesia, both
seven major categories: anesthesia, epilepsy, neurodegeneration, the airport size and the air traffic follow a power-law distribution.

TABLE 1 | Summary of anesthesia-related criticality literature.
Study Study population Modality Analysis Main findings
Alonso 3 adult subjects ECoG Vector • Eigenvalues of VAR matrices change significantly as anesthesia is induced.
et al., 2014 with intractable under auto-regressive This finding is robust to changes in how data is normalized and could be
epilepsy propofol (VAR) model; critical used as a metric for depth of anesthesia.
undergoing surgical anesthesia eigenvalues • As anesthesia is induced, high frequency modes are damped, suggesting
removal of an that cognitive processes associated with higher frequencies are being
epileptic focus tuned out while lower frequency processes are associated with maintaining
the patient alive during anesthesia.
• Self-organized criticality (SOC) could be result of synaptic adaptation.
Disrupting synaptic adaptation should lead to loss of SOC.
Liu et al., 8 healthy adults fMRI Power-law • Node degree distribution was power-law distributed for healthy participants
2014 receiving propofol estimation throughout all phases of anesthesia, but was never power-law distributed
infusion; 5 adults for patients with UWS, regardless of spatial scale. Node size distribution
with unresponsive was power-law distributed for both. Study did not meet (Clauset et al.,
wakefulness 2009)criteria for power law.
syndrome (UWS) • Criticality would not be needed for wakefulness alone but would underlie the
brain’s ability to recover from anesthesia or deep sleep. Future research
should investigate whether power laws or other markers of SOC are helpful
predictors of recovery from coma or other minimally conscious states
Thiery 7 adults receiving EEG DFA • Unconsciousness under sevoflurane was associated with increases in LRTC
et al., 2018 sevoflurane in beta amplitude over frontocentral channels and decrease in alpha
anesthesia amplitude over occipito-parietal channels.
• LRTC and oscillation amplitude may reflect different properties of the brain
that are impacted during anesthesia.
ECoG, electrocorticogram; fMRI, functional magnetic resonance imaging; EEG, electroencephalogram.
But in patients with UWS, the airport size remains power-law to administer anti-epileptic medications before seizure onset,
distributed while the air traffic loses its power-law distribution. either by themselves or through automated implanted devices
This loss may reflect changes in underlying neuronal network (Cook et al., 2013). To this end, several studies have identified
topology that give UWS such a poor prognosis. However, more criticality-based signatures that may help predict the onset of a
research in this area is required since the node degree distribution seizure. For example, critical systems near a phase transition will
in this study does not meet the power-law criteria of Clauset exhibit signs of “critical slowing.” A combined EEG and ECoG
et al. (2009) and, therefore, may follow a different distribution. study of children with epilepsy found evidence of such critical
In fact, another study found an exponentially truncated power slowing in synchronous fluctuations up to 1 h before seizure
law for both patients groups (anesthetized and UWS patients), onset (Cerf et al., 2004). Other studies have taken advantage of an
which suggests the absence of a distinctive signature for the uncanny similarity between seizures and earthquakes. Seizures,
UWS group (Achard et al., 2012). Significant technical differences like earthquakes, exhibit several properties that are characteristic
between Liu et al. (2014) and Achard et al. (2012) make a of SOC (Bak and Tang, 1989; Bak et al., 2002). For example,
direct comparison difficult. Nevertheless, these studies raise the both seizures and earthquakes cluster temporally, such that
possibility that SOC underlies the brain’s ability to rebound the likelihood of the next seizure or earthquake decreases the
quickly from anesthesia or deep sleep, but not from UWS or longer the seizure-free or earthquake-free interval (Omori, 1895;
other major brain insults. Future research in this area will benefit Sornette and Sornette, 1989; Osorio et al., 2009, 2010). Studies
intensivists and neurologists looking for prognostic markers of have identified other SOC properties in seizures. These include
irreversible brain damage. a power-law distribution of inter-seizure intervals (Osorio et al.,
2009, 2010; Cook et al., 2014) and of time intervals between
Epilepsy non-ictal epileptiform discharges, including burst suppression in
Epilepsy, a disorder characterized by multiple seizures and neonatal hypoxia (Worrell et al., 2002; Roberts et al., 2014a). By
affecting 1% of the world’s population, represents a significant better characterizing these properties, one could hope to identify
clinical challenge (Fiest et al., 2017). There is a robust body parameters that can help predict a future seizure−much the same
of work examining the applications of critical phenomena to way that seismologists would like to predict the next earthquake
epilepsy, as summarized in Table 2. From this literature, the (Meisel and Loddenkemper, 2019).
following four topics emerge as clinically relevant areas of The predictability of these seizures may be related to increased
ongoing research: seizure prediction, seizure localization, seizure temporal correlation or long memory. A study of long-term
characterization, and quantitative analysis of seizure genesis ECoG recordings found that epileptiform discharges and seizures
and termination. in some patients were consistent with long-memory processes,
Seizure prediction is an important goal in epileptology with signal correlations going as far back as 40 days before
(Mormann et al., 2007). Efficient seizure prediction, even if only seizure onset (Cook et al., 2014). Long memory suggests a
by a few seconds to minutes, may allow patients enough time decrease in signal complexity and, thus, more predictability.

TABLE 2 | Summary of epilepsy-related criticality literature.
Worrell 7 adult patients Interictal ECoG Power-law • Study found evidence of SOC in interictal epileptiform discharges and
et al., 2002 with medication- estimation suggested SOC-based method for identifying seizure focus.
resistant temporal
lobe epilepsy
Cerf et al., 6 pediatric patients Pre-ictal and Critical • Study found evidence of criticality in synchronous fluctuations up to 1 h
2004 with presurgical inter-ictal ECoG, slowing-down of prior to seizure onset.
epilepsy evaluation scalp EEG pre-ictal amplitude • Root mean square amplitude or excess energy content were suggested as
possible criticality order parameters.
Parish 5 adult patients Wake, sleep, DFA • LRTC in energy fluctuations over seconds to minutes was seen in both
et al., 2004 with unilateral pre-ictal and epileptogenic and non-epileptogenic hippocampus.
mesial temporal ictal/post-ictal • DFA exponents for non-epileptogenic regions were smaller compared to
lobe epilepsy ECoG epileptogenic regions, but no difference in DFA exponents was noted
between pre-ictal and baseline state.
Monto 5 adult patients Inter-ictal ECoG DFA • LRTC was present near the seizure focus, and seen prominently in beta
et al., 2007 with medication- band (14−30 Hz). Lorazepam decreased beta-band LRTC near the focus
resistant and increased LRTC in other cortical areas.
neocortical • Anti-epileptic mechanism of benzodiazepines may be related to
epilepsy; 2 patients normalization or reduction of LRTC in epileptic focus and may serve as
received lorazepam biomarker during presurgical localization of epileptic foci.
Osorio 60 adult patients Pre-ictal and ictal Power-law • Study finds evidence of power law distribution for seizure energy and
et al., 2009 with mesial ECoG estimation inter-seizure interval time. Moreover, study found that seizures tend to occur
temporal and in clusters, obeying an Omori-type law in which the likelihood of next
frontal lobe seizure decreases the longer the seizure-free interval. These insights
medication- suggested a strong analogy between seizures and earthquakes, which
resistant behave in a self-organized critical way.
epilepsy
Osorio 60 adult patients Pre-ictal and ictal Power-law • Five statistics from seismology (including energy, inter-event waiting time,
et al., 2010 with mesial ECoG estimation direct and inverse Omori law, time to next earthquake) were compared to
temporal and analogous statistics in seizures. Insights from SOC in earthquakes were
frontal lobe applied to seizures.
medication-
resistant
epilepsy
Hobbs 6 pediatric epilepsy LFPs from removed Branching • Some epileptic brain tissue exhibited prolonged hyperactivity. Study found
et al., 2010 patients epileptic brain parameter, positive correlation between firing rate and critical branching parameter
tissue correlation, during this prolonged hyperactivity, suggesting possible existence of
power-law positive feedback loop in some forms of epilepsy.
estimation
Meisel 8 adult patients Pre-ictal, ictal, Power-law • Study found a robust power law distribution of phase-locking intervals and
et al., 2012 with focal epilepsy post-ictal ECoG estimation* saw this as evidence of SOC. Variations in goodness-of-fit suggested that
not all brain regions are tuned to criticality at the same time.
• Significant deviation from power law during seizure suggested departure
from critical state, in part due to excessive synchronization.
Meisel and 8 adult patients Pre-ictal, ictal Variance of signal • The inverse of the signal variance followed a scaling law and decreased as a
Kuehn, with intractable ECoG amplitude seizure approached.
2012 epilepsy • Oscillations in variance leading toward seizure onset were suggestive of
critical transition characterized by a Hopf bifurcation.
Kramer 19 adult patients ictal and post-ictal Critical slowing • Multi-scale analysis suggested seizure termination happens through a
et al., 2012 with different from scalp EEG, down, temporal critical transition, modeled by a discontinuous fold bifurcation.
epilepsy etiologies ECoG, LFP and and spatial • Status epilepticus may represent a system’s inability to cross a critical
MUA correlations, transition, instead reverberating between ictal and post-ictal attractors.
flickering
Roberts 13 term neonates Non-ictal EEG of Power-law • Power law relationship was seen between burst size and duration. The
et al., with either birth burst suppression estimation*, scaling scaling exponent of that relationship increased as burst suppression gave
2014a hypoxia or seen within 18 h of relations, burst way to normal EEG activity.
circulatory collapse birth shape analysis • Shape analysis revealed leftward skewness, also seen in crackling noise like
(skewness, Barkhausen noise. Skewness resolved as burst suppression gave way to
kurtosis) normal EEG activity. Leftward skewness may be related to state-dependent
metabolite depletion.
• Other signs of criticality include evidence of shape scaling function and
inter-relationship of power-law exponents for burst area, duration, and
area-duration relationship. There is evidence of critical phase transition from
burst suppression to resumption of normal EEG.
(Continued)

TABLE 2 | Continued
Cook et al., 15 adult patients ECoG recorded Power-law • Study found evidence of a power law for inter-seizure interval in large
2014 with refractory over 0.5−1.8 years estimation; Hurst human dataset, with scaling exponent −1.5 consistent with previous
epilepsy exponent studies. Hurst analysis was consistent with a long-memory process in most
subjects, with memory ranging from 3 to 40 days.
• The presence of long-memory implies a less complex and more predictable
system−epilepsy prediction may depend on the existence (or not) of
long-memory in different types of epilepsy.
Minadakis 2 adult patients Pre-ictal and ictal Q-parameter, Tsallis • While seizures showed consistently elevated q-parameter in range 1.6−1.8,
et al., 2014 (iEEG); 6 adult ECoG and scalp entropy, volumetric inter-ictal and pre-ictal EEG could not be readily distinguished by
patients (scalp EEG energy density q-parameter alone.
EEG) • There was evidence of intermittent criticality (IC), which may generalize to
SOC on larger time scales. Tsallis entropy did not change significantly from
pre-ictal to ictal, suggesting other forms of complexity may be involved in
the ictal period. The concept of “fractures and faults in the brain,” a
continuation of the earthquake-seizure analogy, may be a fruitful framework
for advancing seizure prediction.
Yan et al., 3 adult patients Pre-ictal, ictal, Power-law of • The pre-ictal to ictal transition was characterized by transition from
2016 with refractory post-ictal ECoG wavelet spectral anti-correlation to correlation as given by Hurst exponent and fractional
temporal lobe density, Hurst Brownian motion (fBm) model.
epilepsy exponent, linear • Hurst exponent changes in ictogenesis happened throughout the brain and
correlation not just at epileptic foci. In posti-ical state, high Hurst exponents were seen
coefficient throughout the brain, suggesting seizure was result of breakdown of global
neuronal network.
• Wavelet-based spectral density approach in setting of fBm model may be
helpful tool for seizure prediction.
Arviv et al., 12 adult and 8 Inter-ictal MEG Power-law • Patients with drug-resistant epilepsy showed deviations from expected
2016 pediatric patients estimation*, branching parameter at criticality, especially at interictal epileptiform
with refractory branching discharges.
epilepsy; 18 parameter, • Quantitative analysis of MEG using criticality-related parameters may allow
age-matched avalanche shape better evaluation of excitation-inhibition balance in sleep-related disorders
healthy controls analysis and in epilepsy.
Witton 2 pediatric and 1 MEG, beamformer Hurst exponent, • Hurst exponent analysis, kurtosis, and rogue waves could serve as
et al., 2019 adult patients with source models, rogue wave important parameters in automatic classifiers for epilepsy detection, as well
medication- volumetric maps analysis, kurtosis of as for patients referred for pre-surgical MEG evaluation who do not have
resistant inter-ictal spikes interictal spikes.
epilepsy • Epileptiform activity was strongly persistent, suggesting that Grainger
causality is not suitable for epilepsy data.
Asterisk (*) represents power-law estimations that meet criteria equivalent to or more stringent than Clauset et al. (2009). MEG, magneto-encephalography; ECoG,
electrocorticogram; LFP, local field potential; MUA, multi-unit activity; DFA, detrended fluctuation analysis.
These studies open up the possibility of using LRTC and a power-law-like behavior of seizure energy and inter-seizure
other correlation measures to predict impending seizures. While intervals (Worrell et al., 2002). An ECoG study of epileptic
there is no gold-standard approach for seizure prediction at patients found that both non-epileptic and epileptic foci in the
this time, the brain criticality framework offers new insights hippocampus exhibited LRTC (Parish et al., 2004). However,
that will hopefully produce several candidates for effective the epileptic foci had larger scaling exponents compared to
seizure prediction. Some of these candidates include excess the non-epileptic foci. Another ECoG study found a similar
energy content of EEG signals (Cerf et al., 2004), heavy-tailed result, namely stronger LRTC near the seizure focus (Monto
distributions of inter-ictal discharges (Osorio et al., 2010), signal et al., 2007). More recently, a magnetoencephalography (MEG)
variance (Meisel and Kuehn, 2012), Hurst exponent analysis study found that the Hurst exponent – a metric of LRTC –
(Cook et al., 2014; Yan et al., 2016), q-parameter, and volumetric improved the detection of seizure foci (Witton et al., 2019). These
energy density from Tsallis non-extensive statistical mechanics studies globally suggest that observables of critical phenomena,
(Minadakis et al., 2014). like power-law regimes and increased correlation, can improve
The surgical removal of epileptic foci for treating refractory seizure localization techniques.
epilepsy requires adequate localization of the source of The lens of criticality is also casting new light on the
epileptiform discharges (Mu et al., 2014). Localizing the epileptic characteristics and dynamics of seizures themselves. Several
focus remains a challenging part of this process. Epileptic foci publications have argued that seizures and inter-ictal epileptiform
exhibit many critical features, like power-law distributions and discharges, including burst suppression, represent a critical
LRTC, that are useful for localization. In an ECoG study of phenomenon that is power-law distributed (Worrell et al., 2002;
patients with temporal lobe epilepsy, the epileptic foci produced Cerf et al., 2004; Osorio et al., 2009, 2010; Roberts et al., 2014a). If

this is correct and the reported power-law exponents for seizure from particular experimental conditions (including digitization
energy are between 2 and 3 (as many are), then it follows that, at rates of instruments). But this range may also reflect the reality
least mathematically, seizures ought to have a finite mean (energy, that neurons, unlike earthquakes and sand-piles, learn and adapt
size, duration) but infinite (energy, size, duration) variance. Some (Bonachela et al., 2010). The existence of characteristic scales
have argued that this may account for status epilepticus, the (“bumps”) on log-log plots, which perturb the expected scale-
phenomenon of prolonged seizures lasting hours to days. The free distribution, may also be due to characteristic scales from
theoretically infinite variance of seizure energy would lead to rare neurological events such as dragon-kings (Pisarenko and
prolonged seizures that would ultimately resolve because of the Sornette, 2012; Sachs et al., 2012). Finally, since many control
finite metabolic supply available to neurons (Osorio et al., 2010; parameters may be involved in governing these systems, the
Roberts et al., 2014b). Moreover, if seizures represent a power- possibility of “double criticality” whereby critical regimes coexist
law distributed critical phenomenon, they cannot be described by with different order and control parameters may also be at work
their mean values, since, in a scale-free distribution, there is no in this apparent disagreement (Hesse and Gross, 2014). Resolving
“typical” value or mean. Thus, there may be no point in reporting this disagreement on both theoretical and experimental grounds
mean seizure duration or energy in clinical publications. Rather, will be an important area of future research.
in future clinical and epidemiological studies, it may be more Brain criticality also offers insights into seizure initiation.
pertinent to report power-law exponents, which best characterize Epileptic foci removed from pediatric epilepsy patients exhibited
this distribution (Osorio et al., 2009). neuronal hyperactivity, whose increased firing rate correlated
However, other publications have argued that the normal with an increased branching parameter (Hobbs et al., 2010). This
brain at rest is in a critical state. Therefore, seizures and finding suggests that in some epileptic syndromes, a positive
interictal epileptiform activity should represent a departure feedback loop between firing rate and branching parameter may
from criticality. This departure from criticality perhaps arises be responsible for generating seizures as a super-critical state.
from synchronization effects and characteristic scales present An ECoG study found oscillations in signal variance in the
in seizures that become dominant, thus diminishing the scale- lead-up to a seizure (Meisel and Kuehn, 2012). These pre-ictal
free distribution (Osorio et al., 2009; Meisel et al., 2012; Arviv oscillations were suggestive of a critical transition, characterized
et al., 2016). This departure from a scale-free distribution has mathematically by a Hopf bifurcation. Despite the small number
been confirmed visually with a “knee,” “shoulder,” or “bump” – of studies in this area, research on seizure generation using
different words for the same anomalous deviation−in the log- criticality is promising.
log plot of several probability density functions, including that Several studies suggest that seizure termination may also
of seizure energy (Osorio et al., 2009), phase-locking intervals involve a critical transition. In one study by Kramer et al.
(Meisel et al., 2012), neuronal avalanche size (Arviv et al., 2016) (2012) the brain’s inability to complete a critical transition results
and burst area in burst suppression (Roberts et al., 2014a). in status epilepticus, in which the brain dynamics constantly
The evidence suggesting that the resting brain is in a critical reverberate between the ictal and post-ictal state (i.e., attractor),
state is strong (Kitzbichler et al., 2009; He, 2011; Tagliazucchi without ever crossing the threshold that effectively ends a seizure.
et al., 2012; Daffertshofer et al., 2018), even if some of that In a study of neonates with birth hypoxia, researchers found
evidence (Kitzbichler et al., 2009; Meisel et al., 2012) has been evidence of a critical phase transition in the shift from burst
challenged (Botcharova et al., 2012). Seizures would logically suppression to the resumption of normal EEG patterns (Roberts
seem to represent a departure from the resting state of the et al., 2014a). The role of benzodiazepines in seizure termination
brain and thus from a critical state. How then can one reconcile may also be related to criticality. In a small ECoG study of
this with all the evidence suggesting that seizures behave like a patients with epilepsy, study authors found that a decrease in
critical phenomenon? LRTC in the ictal focus accompanied the clinical resolution
Perhaps one way to reconcile this information is to look more of a seizure after benzodiazepine administration (Monto et al.,
closely at the types of variables (Milton, 2012). In cases that have 2007). These studies all suggest that criticality plays a role in
identified power-laws in the resting brain (with departures during seizure termination.
seizures), the variables studied were usually neuronal avalanche
size and duration, in what could be called an “avalanche Neurodegeneration
approach.” On the other hand, studies that found power-law A small number of studies (see Table 3) have examined the
behaviors of seizures took more of an “earthquake approach” in role of critical phenomena in neurodegenerative diseases, like
which the variables were usually seizure energy and inter-seizure Alzheimer’s disease (AD) and Parkinson’s disease (PD). These
interval (Worrell et al., 2002; Osorio et al., 2009). Since different studies reveal new insights about the pathophysiology of these
properties, or “laminar phases,” are being examined in each case, diseases and suggest novel markers for disease monitoring.
it may not be reasonable to compare their power-law exponents Cognition requires production and subsequent decay of
(Milton, 2012). synchronization in neural networks (Breakspear and Terry,
Moreover, the range of power-law exponents found in both the 2002). Moreover, in healthy adults, spontaneous fluctuations in
avalanche and earthquake approaches is broad and overlapping. synchronization are known to follow a power-law distribution,
The exponents encompass the range of –3/2, which is expected suggestive of an underlying SOC state (Stam and de Bruin,
for avalanche size in SOC, up to −5/3, characteristic in turbulent 2004). Does AD represent a departure from critical dynamics?
dynamical systems (Milton, 2012). This broad range may result Is synchronization perhaps a control parameter given its

TABLE 3 | Summary of neurodegeneration-related criticality literature.
Study Study Population Modality Analysis Main Findings
Stam et al., 24 adults with AD, EEG during DFA • Study examined mean synchronization in different frequency bands. Mean
2005 19 non-demented eyes-closed resting EEG synchronization and spontaneous fluctuations of synchronization were
adults with state lower in AD in upper alpha and beta bands compared to non-AD patients.
subjective memory Mean synchronization level and DFA exponents were correlated to MMSE
complaints score. Both patients and controls showed scale-free patterns of
synchronization fluctuations, extending to up to 10 s.
• AD patients may have brain electrical pattern consistent with SOC but
exhibit decreased processing speed from decreased fluctuations of
synchronization.
Montez 19 adults with MEG during DFA, burst statistics • Using criticality-based “avalanche analysis,” study found that AD patients
et al., 2009 early-stage AD, 16 eyes-closed resting had a strongly reduced incidence of alpha-band oscillation bursts over
age-matched state temporo-parietal regions and markedly weaker autocorrelations on long
controls time scales (1–25 s).
• Study suggested that criticality-related measurements of amplitude
dynamics of oscillations may prove useful as neuroimaging biomarkers of
early-stage AD.
Hohlefeld 10 adults with LFP from bilaterally DFA • Study examined LRTC of the amplitude envelope of LFPs recorded from
et al., 2012 idiopathic PD implanted subthalamic nucleus, both on and off of levodopa. “On levodopa” state was
electrodes from characterized by stronger LRTC (up to 50 s) than the “off” state in beta and
STN DBS high-frequency oscillations.
• Weaker LRTC in off state might indicate limited information processing in
dopamine-depleted basal ganglia. Study suggests LRTC may serve as
possible biomarker of pathological neuronal processes in PD.
Ruiz et al., 1 adult with severe Inter-onset interval DFA, spectral • Study investigated temporal deviations during skilled piano performance of
2014 idiopathic PD, (time between note density a non-professional pianist with severe PD treated with STN DBS. In
treated with STN onset of two tremor-affected right hand, timing fluctuations of the performance exhibited
DBS subsequent notes) random correlations while off DBS. When DBS was on, LRTC increased
while playing piano, along with general motor improvement.
with STN DBS both • The authors remark that the presence of LRTC and 1/f laws in performance
on and off (improved by DBS) can be related to the brain operating near criticality.
Vyšata 110 adults with EEG during Power-law • Study evaluated power-law exponents for power-law distribution of EEG
et al., 2014 moderate-to-severe resting-state estimation, spectral spectrogram from patients with AD compared to healthy controls.
AD, 110 healthy density Power-law exponent was found to be a specific marker of AD in the frontal
controls EEG channels. Authors suggest that loss of functional connectivity may
explain these differences in power-law exponents. Clinical utility of
power-law exponent of spectrogram would require repeating the study on
patients with mild cognitive impairment or early stages of AD.
West et al., 12 adults with PD LFP while on and Spectral density, • Study examined LFPs from PD patients undergoing STN DBS surgery, on
2016 who received off dopaminergic signal coherence, and off of dopaminergic medications. Authors demonstrated up-modulation
bilateral STN DBS medications DFA of alpha-theta (5−12 Hz) band power with L-DOPA treatment, whilst low
beta band power (15−20 Hz) band-power was suppressed. Using DFA
adapted to phase synchrony (DFA-PS), study found LRTC in phase
dynamics of coupled left and right STN region for low beta band. Low beta
band DFA-PS scaling exponent magnitude for interhemispheric pairs was
positively correlated with PD symptom severity in the off-medication state.
Findings suggested that the more severe the motor impairment, the closer
the subthalamic network was to onset of synchronization, implying shift of
network toward supercritical regime.
AD, Alzheimer’s disease; PD, Parkinson’s disease; STN DBS, subthalamic nucleus deep brain stimulation; MMSE, Mini-Mental Status Examination; LFP, local field potential.
importance in cognition? To begin to answer these questions, ability to generate and destroy synchronized neural networks.
researchers compared various measures of synchronization on The authors go on to speculate that perhaps synchronization
EEG between patients with AD and non-demented patients loss in the upper alpha and beta band is one of the first
with subjective memory problems (Stam et al., 2005). While quantifiable changes in AD since it is statistically different
both cohorts maintained a scale-free distribution of spontaneous from non-demented patients who report memory impairment.
fluctuations of synchronization, the mean synchronization and Moreover, the mean synchronization level and the DFA exponent
its fluctuations were both decreased in the upper alpha and of synchronization fluctuations were both correlated to the Mini-
beta frequency range in the AD patients compared to the non- Mental State Examination (MMSE) score. Synchronization-
demented patients. This finding is consistent with the view based metrics may, therefore, prove helpful for diagnosis and
that AD patients maintain a SOC state but with decreased monitoring of early-onset AD.

Other metrics inspired by brain criticality show promise rhythmic stimulation (Hove et al., 2012). In a unique study, a
for the diagnostic evaluation of early-onset AD. A MEG study single patient with idiopathic PD and right-handed tremor, who
discovered a decreased incidence of alpha-frequency oscillation happened to be an accomplished pianist, received DBS and was
bursts and weaker auto-correlations in patients with early- subsequently asked to perform works of piano both with and
onset AD compared to controls (Montez et al., 2009). Its without active DBS (Ruiz et al., 2014). Without DBS, correlations
authors concluded that oscillation amplitude dynamics may be in the inter-onset interval (i.e., the time between note onset of
beneficial for early-onset AD detection. Operating from the subsequent piano notes) were random. But with active DBS, long-
framework of SOC, a large, resting-state EEG study of patients range correlations of inter-onset interval emerged, along with
with moderate-to-severe AD compared to healthy controls also general motor improvement of the affected right hand on the
identified a possible marker for early AD detection (Vyšata Unified Parkinson’s Disease Rating Scale (UPDRS) - III scale.
et al., 2014). The power-law exponents for spectral densities The authors suspected that DBS provides a similar stimulus to
(per brain region) were compared between AD and healthy the non-invasive rhythmic stimulation of Hove et al.’s (2012)
patients. A statistically significant difference in the power-law experiment, which restores the 1/f noise of gait, presumably by
exponent in the frontal and pre-frontal lobes was noted. This restoring a critical state in the motor basal ganglia circuit (Ruiz
result is perhaps not so surprising given that frontal lobe et al., 2014). Future research should focus on establishing whether
atrophy typically accompanies AD dementia. Interestingly, the restoration of 1/f noise is necessary for motor improvement in
most highlyspecific and predictive area of the brain for AD PD and whether gait-based metrics can serve a clinical purpose
in this study was the temporal region (Vyšata et al., 2014). in either diagnosing or monitoring PD.
Future studies using this approach and focusing on the temporal
region may be able to validate this power-law approach as a Neurodevelopment
diagnostic metric in patients with early manifestations of AD. The brain criticality hypothesis also applies to the newborn brain
To this point, a recent fMRI-connectome study showed that that matures and ages across an average human lifespan. As it
a combination of criticality-based metrics can help distinguish grows, the brain’s electrical signature undergoes changes that are
neurotypical adults from those with mild cognitive impairment helpful markers of typical development. Research in this area
or AD (Jiang et al., 2018). has taken shape around two central themes (see Table 4). The
Parkinson’s disease (PD) may be a case of how deviation first is the study of brain oscillations in premature infants or
from a critical state in crucial motor circuits leads to motor infants with birth asphyxia (hypoxemic-ischemic encephalopathy
symptoms like tremors, bradykinesia, and rigidity. Researchers or HIE). The other is the study of brain oscillations in children,
have identified LRTC from the subthalamic nuclei (STN) of adolescents, and adults to characterize the electrical patterns that
patients with PD undergoing deep brain stimulation (DBS) correlate with structural and anatomic changes of aging.
(Hohlefeld et al., 2012). These correlations increased with the If the mature human brain exhibits signs of criticality, it
administration of levodopa, one of the common medications for seems reasonable to ask whether those signs are also present
treating PD. In a rat model of PD, LRTC also increased following in the term or pre-term neonate. Studying burst activity on the
administration of apomorphine (Cruz et al., 2009), which EEG of pre-term infants, Hartley et al. identified LRTC and
suggests that restoration of LRTC may be related to symptomatic dynamics that were suggestive of a phase transition (Hartley et al.,
improvement in PD. A study of PD patients who underwent DBS 2012). Despite several infants having intracranial hemorrhages,
surgery found LRTC in the dynamics of the bilateral STN, both the Hurst exponents describing the LRTC were similar for
on and off medications (West et al., 2016). Using an adaptation of infants with and without bleeding. This finding suggested that
DFA to study synchronization called DFA-PS (Botcharova et al., the brain maintains temporal complexity despite this vascular
2015), the authors found that the DFA-PS exponent was positively insult. Another EEG study showed that power-law exponents of
correlated with motor symptom severity when patients were not electrical bursts were predictive of neurodevelopmental sequelae
receiving dopaminergic medications. Therefore, these authors in term infants with HIE (Roberts et al., 2014a). This fascinating
suggested that patients with more severe motor symptoms are discovery led to a similar EEG study in extremely preterm infants,
closer to the onset of pathological synchronization, which may searching for criticality-based metrics that could predict long-
reduce effective information transfer in these important neural term sequelae. A careful analysis of several parameters from
circuits (Hanslmayr et al., 2012; West et al., 2016). In this regard, shape analysis led to the identification of the slope of burst
PD may represent a situation of departure from a critical state shape, among other parameters, that could serve as a sensitive
toward perhaps a hyper-synchronized supercritical state. and specific predictor of neurocognitive and motor sequelae in
Obtaining recordings from deep brain structures like the basal this unique population (Wikstro et al., 2015). In sum, shape
ganglia is not usually possible outside of DBS. Since DBS is not analysis and power-law exponents predict clinical outcomes in
the first-line therapy for PD, it seems unlikely that LRTC will preterm and term infants, respectively. Regardless of hemorrhage
develop into a helpful marker of PD onset and progression. Gait status, the brains of preterm infants exhibit LRTC. These findings
analysis and related behavioral metrics, on the other hand, may suggest that criticality plays a vital role in the dynamics of the
offer a more convenient way to follow the clinical evolution of preterm and term infant brain.
PD. Healthy human gait is characterized by 1/f noise, which is How criticality or near-criticality emerges and evolves in the
known to be a feature of SOC systems (Hausdorff, 2009). This 1/f infant’s brain through adolescence into adulthood is another
noise disappears in PD but resumes with non-invasive auditory crucial area of investigation. Jannesari et al. (2020) studied term

TABLE 4 | Summary of neurodevelopment-related criticality literature.
Suckling 22 healthy adults Task and Hurst exponent, • Previous research had shown that healthy aging and cholinergic blockade
et al., 2008 (11 in the age range resting-state fMRI; singularity with scopolamine were associated with increase in Hurst exponent,
20−25 years old, double-blind, spectrum using implying a marker of suboptimal neurophysiological dynamics (Wink et al.,
11 in the age range randomized wavelet transform 2006). However, previous research had also shown that faster processing
60−70 years old), administration of maximum modulus speed in certain tasks also led to increased Hurst exponent (Wink et al.,
matched for subcutaneous method, multifractal 2008). This study used multifractal approach to tease apart the discrepancy
education. scopolamine or parameters and used (Castaing et al., 1990) algorithm to identify the role of turbulence.
saline (placebo) Authors conclude that turbulence has limited validity, while invariance of
energy dissipation is better explained by critical phenomena.
Thatcher 458 healthy resting-state EEG Mean phase shift • Study explored development of SOC as measured by EEG phase reset−a
et al., 2009 pediatric subjects duration, combination of phase shift followed by phase stability (or phase locking) –
(age 2 months to phase-locking from infancy to adolescence. Mean duration of phase locking (150−450 s)
16 years old) intervals, power-law and phase shift (45−67 s) increased as a function of age. Development and
estimation, spectral number of synaptic connections may be a possible order parameter for
density analysis SOC during human brain maturation.
Berthouze 36 healthy subjects EEG during Spectral density • In physical systems, SOC states take time to develop. Study found that
et al., 2010 (age 0–55 years wrist-extension analysis, DFA there is a scale-free nature to EEG LRTCs from early childhood through to
old) task maturity but that the magnitude of these effects changed with age.
Smit et al., 1433 healthy Resting-state EEG DFA, spectral • Study observed significant increases in LRTC from childhood to
2011 subjects (age density, principal adolescence and into early adulthood. PCA of the spatial distribution of
5–71 years old) component LRTC showed functional-anatomic segregation between frontal,
analysis occipito-temporal, and central regions that became more integrated with
development. DFA scaling analysis may be useful as a biomarker of
pathophysiology in neurodevelopmental disorders like ADHD and
schizophrenia.
Hartley 11 pre-term EEG Hurst exponent • LRTC were identified in very pre-term infants through two estimate of Hurst
et al., 2012 newborns (Whittle estimator exponents of EEG bursts. The study found no difference in Hurst exponents
(23−30 weeks and DFA) between subjects with and without brain hemorrhages, indicating that
gestation) despite lower burst event frequency for newborns with hemorrhages, signal
complexity was maintained. Overall EEG pattern was suggestive of
relaxation dynamics as can be seen near a phase transition.
Mares 17,722 healthy Resting-state EEG Spectral density, • Study investigated parameters of colored noise in EEG in healthy adults.
et al., 2013 adults (ages power-law Absolute value of power spectra exponent decreased significantly with age,
18−70 years old) estimation perhaps indicative of age-related changes in self-organization of brain
activity due to brain atrophy. Globally, there was a trend from pink noise to
white noise with age that was seen consistently in beta and delta bands.
Fransson fMRI: 18 term EEG in stage 2 power-law • Study found that newborn brain dynamics follow apparently scale-free
et al., 2013 newborns and 17 sleep, fMRI estimation frequency power distribution across several orders of magnitude in both
healthy adults (ages fMRI and EEG signals. In newborns, primary sensory areas exhibit larger
22−41 years old); power-law exponents than higher associative cortical areas, in contrast with
EEG: 15 term or the adult brain.
post-term • High power-law exponents in newborns were likely due to spontaneous
newborns, 7 activity transients (SATs) or bursts that seem to underlie brain activity in the
healthy adults (ages first neuronal networks in the human brain (Vanhatalo and Kaila, 2006).
14−53 years old)
Thatcher 70 healthy subjects LORETA (EEG) of Phase shift • Study found no significant correlation between age and phase shift and
et al., 2014 (age 13−20 years the Brodmann duration, phase phase lock duration from EEG of the default mode network. Study findings
old) areas of the default lock duration were globally consistent with SOC.
mode network in
the delta frequency
band
Frohlich 39 preschool-age EEG Frequency • Study quantified variance of rate of change of signal phase (i.e. frequency
et al., 2015 healthy subjects variance, variance) as a proxy for phase reset (or signal stability). Frequency variance
power-law increased with age in preschool age children. This method is helpful in
estimation pediatric studies because it does not require long recordings. Authors
suggest that phase resets are critical fluctuations driven by SOC.
Iyer et al., 43 preterm Resting-state Power-law • Study found scale-free properties of EEG bursts in extremely preterm
2015 neonates single-channel EEG estimation*, burst infants as soon as 12 h after birth. Metrics of burst shape were predictive of
(23−28 weeks recorded at 12, 24, shape analysis, neurodevelopmental outcomes using Bayley scales. Specifically, symmetric
gestation) 48, and 72 h of life generalized linear bursts that are relatively flat at long time scales suggested a favorable
model neurodevelopmental outcome. Conversely, skewed and highlykurtotic
bursts in neonates shortly after birth were suggestive of long-term
(Continued)

TABLE 4 | Continued
disability. Low burst slope values, moderated by effect of gestational age,

correlated with poor scores on the Bayley scales or early death.
Padilla 33 children born fMRI and diffuse Ignition analysis, • Study compared 10 year-old children who were either born extremely
et al., 2020 extremely MRI at 10 years old structural and pre-term (EPT) or born at term, using fMRI with ignition analysis. Intrinsic
prematurely and 29 connectivity ignition events allow propagation of neuronal activity to other regions over
children born term matrices, time which drives global integration. Extremely pre-term children had
whole-brain Hopf reduced intrinsic ignition events, consistent with previous study that had
model shown reduced spontaneous neuromagnetic activity in pre-term children.
Study found that the hierarchy of information processing based on the
variability of intrinsic ignition events was predominantly driven by visual and
sensory region in EPT children compared to the higher-order processing
areas like the fronto-temporal region and the associative area in term
children.
Jannesari 19 term infants High-density EEG Power-law • Study evaluated infants at 6 and 12 months of age during auditory odd-ball
et al., 2020 during an oddball estimation*, DFA task to see if the bursting, scale-free activity of pre-term infants continues
auditory task as scale-free avalanche activity outside the newborn period. Suprathreshold
events organized as spatiotemporal clusters whose size and duration were
power-law distributed while time series of these events showed significant
LRTCs. Power law was a better distribution fit than log-normal and
exponential. No significant differences were noted between 6 and
12 months, suggesting stability of avalanche dynamics and LRTCs in the
first year after birth.
Asterisk (*) represents power-law estimations that meet criteria equivalent to or more stringent than Clauset et al. (2009). LORETA, low-resolution electromagnetic
tomography; PCA, principal component analysis; ADHD, attention deficit hypersensitivity disorder; MDI, motor development index.
infants at 6 and 12 months of life using high-density EEG to scale-free power (frequency) distributions across several orders of
see if scale-free brain activity that was already known to exist magnitude, with larger power-law exponents in primary sensory
at birth (Fransson et al., 2013) continued in the first year of areas compared to associative cortical areas (Fransson et al.,
life (Jannesari et al., 2020). At both 6 and 12 months, the 2013). In contrast, power-law exponents for adult brain dynamics
EEG organized spatiotemporally according to a power-law with were highest in associative cortices. While both adults and term
significant LRTC. The authors found no significant differences infants exhibit scale-free dynamics, the increased exponents in
between the 6- and 12-month data, suggesting a degree of stability different brain areas suggest that development and aging control,
in neuronal avalanche dynamics after the newborn period. But to some degree, the cartography of near-criticality in the brain.
while there do not seem to be significant differences in LRTC As for aging, the development of criticality and its suspected
and power-law distributions during the first year of life in term decay with age have been the subjects of several studies (see
babies, a significant difference does exist between pre-term babies Table 4). Phase reset (PR), a combination of phase-shift followed
and term babies, even a decade after birth. Using functional and by phase-locking of EEG signals, is a powerful marker of SOC
diffusion MRI, researchers examined the brain dynamics of 10- used in many age-based studies of criticality. In a study of
year-old children, roughly half of whom were born extremely PR in close to 460 subjects in the age range of 2 months to
premature, and the other half at term (Padilla et al., 2020). 16 years old, PR followed a 1/f distribution, with a longer mean
Extremely premature (EP) children generated fewer electrical duration of both phase-shift and phase-locking as a function
events (i.e., decreased ignition) compared to the term children – of age (Thatcher et al., 2009). The presence of this distribution
suggesting decreased global integration from decreased firing argues in favor of scale-invariant fluctuations in PR, consistent
to other important brain regions over time (Deco et al., 2017). with SOC (Thatcher et al., 2009, 2014). In a study of pre-
EP children also exhibited abnormal hierarchical organization, school children, EEG frequency variance, a proxy for PR, was
which autistic children are also known to exhibit (Parr et al., also power-law distributed and increased with age (Frohlich
2018), further linking prematurity and autism (Padilla et al., et al., 2015). Like PR, frequency variance may reflect critical
2017). EP children’s brain dynamics showed decreased synchrony fluctuations driven by SOC.
and sub-criticality compared to term children, mainly in brain LRTC has also been found to correlate with aging. A study
areas with rich-club architecture (Ball et al., 2014). This difference of about 1430 subjects from ages 5 to 71 revealed significant
in synchronization perhaps occurs due to abnormal development increases in LRTC from childhood through adolescence into early
of white matter in EP children (Uhlhaas et al., 2010). Yet, whereas adulthood (about age 25), after which LRTC stabilized (Smit
extreme prematurity seems to predispose to some degree of sub- et al., 2011). Scale-free modulations of resting-state oscillations,
critical brain dynamics, it seems from other studies that term therefore, seem to reflect brain maturation. Moreover, principal
infants go on to exhibit similar power-law-like brain dynamics as component analysis (PCA) showed progressive integration
adults. In an EEG-fMRI study, newborn brain dynamics followed of segregated functional-anatomic brain regions with age –

consistent with increased spatial correlation from critical Another study of healthy adults undergoing a cognitive task
dynamics. In a smaller study of subjects ages 0 to 55, EEG also showed a power-law scaling of response time fluctuations
LRTC was present from early childhood into adulthood but (Simola et al., 2017). Furthermore, Simola et al. (2017) found
with magnitudes that changed differently depending on age, autocorrelations and LRTC of response time fluctuations, with
EEG electrode, and frequency band (Berthouze et al., 2010). the LRTC scaling exponents correlating negatively with error
For example, LRTC magnitude increased with age in the beta rates and positively with executive function testing scores. As
band in central and parietal electrodes but decreased with age the authors pointed out, LRTC can originate in the setting
in the theta frequency range. But temporal correlation is not of criticality but also exist in systems with a slow decay of
the only observable that appears to decrease with age in some memory. The negative correlation of LRTC scaling exponent
frequency bands. A study of resting-state EEG in nearly 18,000 with error rates suggests increased cognitive flexibility, which
individuals ages 18 to 70 found a decrease in the power-law is more consistent with criticality than with a system that
exponent of spectral density with age (Vyšata et al., 2014). operates with slowlydecaying memory. Without criticality, LRTC
The authors theorized that brain atrophy with age might lead would decrease cognitive flexibility because long-memory would
to loss of neuronal connections (Morrison and Hof, 1997), not allow reconfigurations. Patients with ADHD are known to
which would shift the neural networks away from their scale- show increased response time variability in these kinds of tasks.
free topology (Barabási and Albert, 1999) and thus away from However, in this study, the LRTC scaling exponents were not
power-law dynamics. correlated to response time variability or mean response time,
While there may be some appearance of contradiction in implying that another brain process may be involved in ADHD.
these results (i.e., increased LRTC in one frequency range versus Studies of behavioral dynamics (e.g., response times on tasks)
another with aging), there is precedent for contradictory results in ADHD patients compared to neurotypical patients may grant
finding resolution when the right tool is applied. For example, further insights into the role of scale-free, critical dynamics
both healthy aging and anticholinergic medications increase the in ADHD.
Hurst exponent (Wink et al., 2006). Yet, faster processing speeds, But while the behavioral response (i.e., response times) to
which are not characteristic of aging, also lead to increases increased cognitive load may be power-law-like, studies of
in the Hurst exponent (Wink et al., 2008). Suckling et al. the electrical brain response to increased cognitive load show
(2008) reconciled these results with the introduction of a multi- more mixed results. A MEG study of neurotypical children
fractal approach and demonstrated that criticality offered a better and children with high-functioning autism undergoing executive
explanation than turbulence for these brain dynamics. This function tasks demonstrated a decrease in power-law scaling of
example highlights how the right methodology and tools can help phase synchrony as cognitive tasks became more difficult, i.e.,
make sense of disparate results. Methodological developments, increased cognitive load (Tinker and Velazquez, 2014). Power-
like the index of functional criticality (Jiang et al., 2019), will law distributions were uncommon in this study in both autistic
hopefully generate more useful clinical results in this area. The and neurotypical children and were more likely to co-exist with
applications of criticality to aging and geriatric medicine are still other distributions (e.g., exponential). These results suggest that
in their infancy but deserve more emphasis in a society with an metastable, rather than purely critical, dynamics are at play in
increasingly elderly population. the brain’s response to increased cognitive loads (Bressler and
Kelso, 2001; Deco and Jirsa, 2012). Along similar lines, an EEG
study of healthy adults during a working memory task found an
Cognition, Attention, Learning, and inverse correlation between LRTC in the theta frequency range
Autism and memory performance (Euler et al., 2016). This result is
In computational and theoretical models, criticality optimizes perhaps the opposite of what one might have predicted from work
certain features of learning, including optimal information that had found indications of criticality in behavioral dynamics
capacity and transmission (Shew et al., 2011; Shew and Plenz, (Altamura et al., 2012; Simola et al., 2017).
2013; Del Papa et al., 2017). Important aspects of human learning, But not all brain studies of human cognition have argued
including cognition and attention, have been analyzed through against the existence of critical dynamics. An EEG study of
the lens of criticality (see Table 5). Studies on attention-deficit 210 neurotypical adults undergoing an object recognition task
hypersensitivity disorder (ADHD) and autism, in which features showed that variation in 1/f noise was a robust predictor of
of neurotypical cognition and attention are disrupted, are few in cognitive processing speed (Ouyang et al., 2020). Moreover, the
number but appear promising. power-law exponent of the 1/f noise was most predictive of
In the area of cognition, researchers have wondered whether person-to-person processing speeds. While there are many non-
criticality plays a role in the brain’s response to increasing critical sources of 1/f noise, this study harmonizes well with the
cognitive loads. In a study of healthy adults undergoing working theoretical literature on criticality’s optimization of information
memory and response tasks, Altamura et al. (2012) found power- transmission. In an EEG study of healthy adults undergoing an
law-like behavior in the upper tails of the cumulative distribution action adjustment task, midfrontal theta – an electrical signature
of response times. One possible interpretation of this result is known to correlate with real-time error correction (Cavanagh
the emergence of scale-free behavior in response to increased and Frank, 2014) – displayed scale-free-like fluctuations over
cognitive loads. In other words, increased cognitive load shifts durations of up to tens of seconds (Cohen, 2016). The author
random behavior toward scale-free behavior near a critical point. suggested that fluctuations may modulate the midfrontal theta

TABLE 5 | Summary of cognition-related criticality literature.
Lai et al., 30 male adult fMRI during Hurst exponent • Study examined complexity of endogenous, low-frequency
2010 subjects with resting-state neurophysiological processes in patients with ASD compared to control
autism or patients. Study confirmed that spontaneous BOLD signal fluctuations in the
Asperger’s brain, specifically in regions implicated as atypical in previous autism
syndrome, 33 age- neuroimaging studies, had small but significant decrease in Hurst
and IQ-matched exponents in the autistic compared with neurotypical group. This finding
male adult controls indicated a shift-to-randomness of brain oscillations in the autistic brain.
• Though the meaning of the Hurst exponent ins limited by our understanding
of neuronal and blood-supply sources to the measured BOLD signal,
nevertheless fractal scaling may serve as indicator of organizational
properties of local neural circuits.
Altamura 12 healthy adults Response times to Probability density • Power law-like behavior was noted in the upper tails of the CDF of response
et al., 2012 working memory functions, times for working memory tasks. This finding possibly reflects emergence of
and response tasks power-law scale-free behavior in time series as an adaptation to increased cognitive
estimation requirements. Increasing cognitive load could shift random behavior to
scale-free behavior near a critical point.
Dimitriadis 23 children with MEG during DFA • Study examined MEG in children with reading difficulties compared to
et al., 2013 reading difficulties, resting-state children without reading difficulties. Children with reading difficulties had
27 age- and decreased overall network organization across all frequency bands (global
IQ-matched efficiency decrease) and a decrease in temporal correlations between
children sensors covering the left temporoparietal region. Study suggested that the
specific parameters of SOC vary systematically in presence of reading
difficulties. Both groups exhibited scale-free global network connectivity
dynamics.
Tinker and 15 children with MEG during two power-law • Study examined scaling of phase synchrony in MEG in patients with ASD
Velazquez, high-functioning executive function estimation compared to controls. Power-law scaling of phase synchrony was not
2014 autism, 16 tasks common in either group. Its frequency of occurrence diminished with
neurotypical increased cognitive load/effort as children performed more difficult tasks.
children (ages Power law distribution coexisted with other distributions (e.g., exponential)
7−16 years old) suggesting a sign of the metastability of brain dynamics.
Fagerholm 18 healthy adults EEG-fMRI during Power-law • Study examined combined EEG and fMRI in healthy volunteers during rest
et al., 2015 rest and a estimation*, shape and cognitive task. Resting-state EEG cascades were associated with
visuomotor analysis approximate power-law distribution, while task state was associated with
cognitive task subcritical dynamics. Decreased response times during the cognitive task
were associated with better approximation of a power-law form of cascade
distribution. Findings suggest that resting-state was associated with
near-critical dynamics while focused cognitive state induced subcritical
dynamics with a lower dynamic range to reduce interference with task (i.e.
promoting task performance).
Euler et al., 54 healthy adults EEG at rest and DFA • Study finds evidence of inverse relation between theta band LRTC and
2016 during a working working memory performance−higher scaling exponent was related to
memory task poorer cognitive performance. Authors suggest that since elevated LRTC
have been noted in epilepsy, increases in LRTC are not always beneficial.
Cohen, 21 healthy adults EEG during action Demeaned • Real-time error correction has been correlated to an idiosyncratic
2016 adjustment task fluctuation analysis electrophysiological signature called midfrontal theta. This study found that
(DMA) midfrontal theta is a transient but non-phase-locked response modulated
by task performance over three time scales, including scale-free-like
fluctuations over many 10 s. The phasic midfrontal theta brain response to
errors or error corrections is modulated by slow fluctuations in criticality.
Simola 27 healthy adults Response times Autocorrelation, • Response time fluctuations in the Go/NoGO task exhibited a power law
et al., 2017 during a Go/NoGo spectral density, frequency scaling, autocorrelations and LRTCs, with LRTC scaling
task DFA exponents negatively correlated with the commission error rates. Finding
suggested that LRTCs co-exist with cognitive flexibility which is in line with
the criticality hypothesis. LRTC scaling exponents were uncorrelated to the
mean response time (MRT), suggesting that performance variables derive
from distinct processes than brain criticality. Understanding the individual
variation in scale-free behavioral dynamics may improve utility of
neuropsychiatric assessment in ADHD.
(Continued)

TABLE 5 | Continued
Irrmischer 28 EEG during Spectral density, • Study evaluated EEG from meditation practitioners and meditation-naive
et al., meditation-trained eyes-closed rest DFA participants from independent labs. In practitioners, but not in controls,
2018a healthy adults, 21 and meditation meditation strongly suppressed LRTC of oscillations relative to eyes-closed
meditation-naïve rest, across all frequency bands and scalp locations. Sustained practice led
healthy adults to reduction in LRTC during meditation after 1 year of additional training.
Practice also impacted normal waking brain dynamics as reflected in
increased LRTC during eyes-closed rest state, indicating an alteration
beyond merely meditation. Authors suggested that meditation-induced
release of GABA may lead to subcritical regime.
Irrmischer 57 healthy adults EEG during Spectral density, • High levels of alpha band LRTC in sensorimotor region during rest predicted
et al., eyes-open, DFA good reaction-time performance in attention task. During task execution,
2018b eyes-closed, and fast reaction times were associated with high-amplitude beta and gamma
temporal oscillations with low LRTC.
expectancy task • Authors hypothesize that focus and attention move the neural system from
near-criticality optimized for environmental and internal demands, to a state
of reduced input propagation but increased attentional stability, leading to
suppression of LRTC.
Jia et al., 35 children with Functional DFA • The hemoglobin concentration signals (i.e., oxy-Hb and deoxy-Hb) of young
2018 ASD (ages near-infrared children with ASD and normal children were recorded via fNIRS while
4−9 years old), 31 spectroscopy watching a cartoon. DFA exponents of young children with ASD were
age- and (fNIRS) while significantly smaller over left temporal region for oxy-Hb signal, and over
gender-matched watching a cartoon bilateral temporo-occipital regions for deoxy-Hb signals, indicating a
neurotypical shift-to-random-ness of brain oscillations in children with ASD. Testing the
children relationship between age and DFA exponents revealed that this association
could be modulated by autism.
• Studying the temporal structure of brain activity via fNIRS technique may
provide physiological indicators for autism. Authors speculated about a
connection with SOC, though functional significance of DFA exponent is
unclear. LRTC could play a role in evaluating disease progression in ASD.
Bongers 22 healthy EEG during Irregular • Study identified power law exponent of fractal signal during continuous EEG
et al., 2019 university students computerized Resampling of computerized chemistry learning. Mixed power increase of broadband
learning and Auto-spectral frequencies, which reflected an overall increase in fractal power, was seen
resting-state Analysis (IRASA), during learning. A low power law exponent with increased band power of
power-law the fractal component seemed to correlate to high learning gains.
estimation
Kwok et al., 41 healthy children High-density EEG Spectral density, • Study used resting state EEG of children with typical development to
2019 (ages 4−6) during eyes-open DFA explore relation between alpha (7−10 hz) oscillations and oral language
and eyes-closed ability. Higher language scores were correlated with lower alpha power and
increased temporal correlations. Findings further demonstrated existence of
critical state dynamics as important for language acquisition.
Ezaki et al., 138 healthy adults Resting-state fMRI; Correlation • Study added support to criticality hypothesis by showing moderate but
2020 IQ scores between spin-glass clear correlation between IQ scores and distance from criticality at an
susceptibility and individual level using dynamic fMRI signals. A model of criticality using spin
performance IQ glasses was compared to data from healthy adults with a range of fluid
score intelligence IQs. Human fMRI data was found to be within paramagnetic
phase close to the boundary with the spin-glass (SG) phase if using the
framework of the Ising model. High fluid intelligence was associated with
proximity to boundary between paramagnetic and SG phases. SG phase
yields chaotic dynamics in spin systems, consistent with idea of enhanced
computational performance “at the edge of chaos.”
Ouyang 210 healthy adults EEG during Structural equation • The goal of this study was to investigate how individual differences in
et al., 2020 eyes-open, modeling; fitting cognitive processing speeds could be predicted by the power spectrum of
eyes-closed oscillations and resting-state EEG signals. Alpha oscillations were not significantly
resting-state, and one-over-f (FOOOF) associated with cognitive processing speed once the 1/f noise was
object recognition methodology; eliminated by SEM. Variation in 1/f was revealed to robustly predict
task IRASA cognitive processing speed in eyes open and eyes closed. Slope of the
power law decaying function was most predictive of between-person
processing speed.
Asterisk (*) represents power-law estimations that meet criteria equivalent to or more stringent than Clauset et al. (2009). ASD, autism spectrum disorder; BOLD, blood
oxygenation level-dependent; CDF, cumulative distribution function.

brain response to errors at a critical point (Cohen, 2016). In reading difficulties and dyslexia. A high-density EEG study of
summary, evidence from both brain dynamics and behavioral neurotypical children found that lower alpha frequency power
dynamics points to some aspects of criticality in human cognition and increased LRTC correlated positively with language scores
(Altamura et al., 2012; Cohen, 2016; Simola et al., 2017; Ouyang (Kwok et al., 2019). This study confirmed the findings of
et al., 2020) while other studies do not (Tinker and Velazquez, Dimitriadis et al. (2013) that critical state dynamics are important
2014; Euler et al., 2016). for language acquisition.
Studies of human attention have centered on the roles of A recent resting-state fMRI study of neurotypical adults with a
criticality in meditation and its possible applications to ADHD, range of IQ scores found that high fluid intelligence is associated
as mentioned earlier. There are many forms of meditation, with proximity to a critical phase transition in a spin-glass model
and a commonly practiced form, called focused attention (Ezaki et al., 2020). This finding was consistent with previous
(FA) meditation, helps its practitioners improve their ability work suggesting near-criticality as perhaps optimal for learning
to focus (e.g., on their breath, on their bodily sensations). (Gisiger et al., 2014). From an EEG study of healthy university
An EEG study of FA meditation-trained subjects compared students learning organic chemistry, it is also known that a lower
to meditation-naïve subjects found that FA meditation in power-law exponent of the fractal component of EEG signals
practitioners led to strong suppression of LRTC in oscillations correlates with higher learning gains during a computerized
across all frequency bands and electrodes, compared to the eyes- learning task under EEG (Bongers et al., 2019). While this latter
closed state (Irrmischer et al., 2018a). Meditation-naïve subjects study makes no claims regarding criticality, one can infer that
did not show this suppression of LRTC. A year’s worth of the scale-free behavior of neurons matters for the acquisition of
meditation practice led to more permanent changes since FA complex new knowledge.
meditation practitioners had increased LRTC during eyes-closed Only a few studies have examined the role of criticality-related
rest state when they underwent EEG a year later. The authors markers in autism. A resting-state fMRI study of adult males with
hypothesized that meditation-related release of GABA might be autism spectrum disorders, including high-functioning autism,
contributing to excess inhibition and a subcritical regime as detected a small but significant decrease in the Hurst exponent
seen by decreased DFA exponent (i.e., LRTC suppression). In compared to controls in brain regions implicated in autism by
a similar EEG study of healthy subjects, high levels of alpha previous neuroimaging studies (Lai et al., 2010). This decrease
frequency LRTC in the sensorimotor region of the brain during in the Hurst exponent indicates a shift toward randomness of
rest predicted strong reaction-time performance in an attention fluctuations in blood oxygen level-dependent (BOLD) signals
task (Irrmischer et al., 2018b). During the execution of the in the brain impacted by autism. Though the interpretation of
task, on the other hand, suppressed LRTC in beta and gamma the Hurst exponent in BOLD signals is challenging because of
frequencies was associated with fast reaction-time performance. the unclear roles of blood supply and neural circuits among
This study complements the results obtained in an earlier many others, the use of fractal scaling can serve as a barometer
EEG-fMRI study of healthy adults during a visuomotor task of underlying coordination and organization of neural circuits.
and at rest. It found that the resting-state is associated with In this study, the shift toward randomness in BOLD signals
near-critical dynamics, while focused cognitive tasks induce may indicate decreased coordination of small-scale circuits,
subcritical dynamics that may reduce interference with the to the disadvantage of larger brain circuitry. This hypothesis
task at hand (Fagerholm et al., 2015). The upshot of these harmonizes well with one of many prevailing theories about
three studies is that focused attention – whether meditation- the origins of autism, namely the local overconnectivity theory
related (Irrmischer et al., 2018a) or visual attention (Fagerholm (Belmonte et al., 2004; Baron-Cohen and Belmonte, 2005).
et al., 2015; Irrmischer et al., 2018b) – is associated with a A more recent study of children with autism spectrum disorders
reduction in criticality fluctuations. Generalized human attention (ASD) used functional near-infrared spectroscopy (fNIRS) to
at rest, on the other hand, balances a certain degree of compare LRTC with age and gender-matched neurotypical
focus with the ability to respond quickly to both internal and children (Jia et al., 2018). Consistent with the findings of
external stimuli. That balance should theoretically be optimal Lai et al. (2010) this study found that DFA exponents (i.e.,
near criticality. LRTC scaling exponents) were significantly smaller over the
A handful of studies have examined the role of criticality in left temporal region for the oxy-hemoglobin (oxy-Hb) signal in
human learning, both in children and in adults. A MEG study children with ASD compared to controls. DFA exponents were
of children with reading difficulties showed decreased temporal also significantly smaller over both temporo-occipital regions
correlations in the left temporoparietal region compared to age for deoxy-hemoglobin (deoxy-Hb) signals in children with ASD
and IQ matched children without reading difficulties (Dimitriadis compared to controls. DFA exponents correlated well with
et al., 2013). While this study examined subjects at rest age in neurotypical children, consistent with findings discussed
instead of in a reading activity, its results dovetail well with previously (Smit et al., 2011). However, DFA exponents of
previous studies that showed aberrant cortical activation in oxy-Hb in the left temporal region correlated negatively with
left posterior and temporal regions in children with severe autism symptom severity on a parental questionnaire. This result
reading difficulty during a reading assignment (Hoeft et al., dovetails nicely with a previouslyidentified negative correlation
2007; Cao et al., 2008). Both groups of children exhibited between cerebral blood flow to the left superior temporal gyrus
scale-free global network connectivity, suggesting that local, and Autism Diagnostic Interview-Revised scores (Meresse et al.,
rather than global, decreases in LRTC may be involved in 2005). While the authors of this fNIRS study suspect that

autism may represent a departure from SOC, it seems further in the strength of LRTC. This latter result suggests a new
neuroimaging studies using criticality-based approaches will be approach for examining the physiological mechanism by which
needed before one can safely reach that conclusion. psychotherapeutic interventions improve depressive symptoms.
Given the inconsistencies in findings, more extensive studies
Psychiatry will help clarify the nature of the LRTC in the various
Researchers have turned to criticality-based tools to improve frequency bands and the multiple stages of depression (acute,
their understanding of common psychiatric conditions like treated, etc. . .). Future studies could assess whether LRTC
depression, schizophrenia, anxiety, post-traumatic stress disorder changes significantly after initiation of first-line pharmacological
(PTSD) (see Table 6). Insights from criticality theory have intervention for MDD. Other criticality-based metrics could be
also helped describe the psychological effects of neurofeedback useful to characterize the nature of brain dynamics in MDD as
and psychedelics. either sub-critical or, less likely, super-critical.
Studies of major depressive disorder (MDD) or unipolar Criticality-based studies of schizophrenia are few but have
depression have mostly relied on the search for temporal opened up new horizons. An EEG study of around 30 patients
correlations (LRTC) using DFA. A small MEG study of patients with schizophrenia and schizoaffective disorders used a fractal
with MDD and healthy controls found absent LRTC in the theta analysis of 1/f EEG rhythm fluctuations to try to distinguish
frequency band in patients with MDD compared to controls the brain dynamics of schizophrenia from that of healthy
(Linkenkaer-Hansen et al., 2005). The study authors suggested controls (Slezin et al., 2007). A frontal electrode (F4) exhibited
that abnormal temporal structure of theta oscillations could increased instability and randomness in the alpha band for
reflect an underlying defect in limbic-cortical networks identified patients with schizophrenia and schizoaffective disorders. The
in anatomic-functional studies of MDD. Another small EEG theta band, on the other hand, exhibited increased stability and
study of patients with MDD and healthy controls did not decreased complexity in patients with schizophrenia compared
reproduce this finding of absent theta LRTC (Lee et al., 2007). to healthy controls. A subsequent fMRI study of patients with
Instead, the study authors found that increased LRTC scaling schizophrenia compared to healthy controls (Radulescu et al.,
exponents (i.e., slower decay of LRTC) correlated positively with 2012) also identified decreased complexity of brain dynamics, but
the severity of depression over most EEG channels. This finding different from that of Slezin et al’s. (2007)study. In Radulescu
led the authors to speculate that rumination and psychomotor et al’s. (2012) study, a sophisticated analysis pipeline, including
retardation – typical features of MDD – may be responsible for power spectrum scale invariance (PSSI) and Poincare maps
this persistence of LRTC. (a measure of signal variance), found that signals coming
Other studies have examined whether depression leads to from Brodmann area 10 (BA10) exhibited lower power-law
alterations in LRTC during sleep. A small sleep EEG study of exponents in schizophrenic patients than in healthy controls. In
patients with untreated, acute episodes of MDD found a decrease other words, schizophrenic patients displayed white (Gaussian
in LRTC scaling exponents in NREM2, NREM3, and NREM4 or random) noise in this brain region, compared to pink
compared to healthy controls (Leistedt et al., 2007b). However, (1/f) noise in healthy patients. This finding is consistent with
this decrease was not statistically significant. A similar study by previous research documenting the role of BA10 in executive
the same research team, this time with patients in remission function, working memory, and emotion regulation (Radulescu
from MDD, found no differences in LRTC through the stages of et al., 2012). This transition from pink noise to white noise in
sleep (Leistedt et al., 2007a). In addition to demonstrating the schizophrenia suggests, from the perspective of criticality, a loss
existence of LRTC during sleep (see Sleep Medicine section for of responsiveness to stimuli.
more details), these results argue against the “depressive scar” While the studies mentioned above took a fractal and power-
hypothesis, according to which depression leads to permanent law approach to signal analysis, others have taken the route
residual defects. of studying LRTC in schizophrenia. Their results complement
Researchers have also taken an interest in negative emotional and refine those obtained using fractal approaches. A high-
regulation as a precursor for MDD. A large EEG study of density EEG study of LRTC in adults with schizophrenia and
non-depressed undergraduate students found a positive linear schizoaffective disorders found strong attenuation of LRTC
correlation between LRTC scaling exponents of theta and scaling exponents in alpha and beta frequency bands in patients
broad-band frequencies and negative emotional regulation on compared to healthy controls (Nikulin et al., 2012). While Slezin
various questionnaires (Bornas et al., 2013). Their research et al. (2007) did not study the beta frequency range, they did
suggested that negative emotional control may anticipate full- observe increased complexity in the alpha range, which is globally
blown MDD by several years. While this study broadly agreed consistent with the findings in this study. These attenuated
with the findings of Lee et al. (2007), it disagreed with scaling exponents indicate decreased temporal correlation and
Linkenkaer-Hansen et al.’s (2005) conclusions with regards to a decrease in temporal precision of neuronal firing compared
the magnitude of scaling exponents with worsening rumination. to healthy subjects. This interpretation fits well with at least
A more recent, large EEG study found that, at baseline, two theories for the origins of schizophrenia, namely that of
patients with MDD have higher LRTC scaling exponents in excessive neuronal noise and variability (Rolls et al., 2008) and
the theta frequency range than healthy controls (Gärtner et al., the “disconnection hypothesis” of dysfunctional relationships
2017). After intervention with either mindfulness training or between neural networks (Friston et al., 2016). A follow-up study
stress reduction training, both groups experienced a reduction found similar results (Moran et al., 2019). In this high-density

TABLE 6 | Summary of psychiatry-related criticality literature.
Linkenkaer- 12 adults with MEG during DFA, linear • This study recorded MEG data from patients with MDD compared to
Hansen MDD, 10 eyes-closed resting correlation of DFA healthy controls during eyes-closed wakeful rest and quantified LRTC in
et al., 2005 age-matched state exponent to amplitude fluctuations of different frequency bands. Temporal correlations in
healthy controls Hamilton theta band were absent in the 5−100 s time range in patients but were
Depression Rating prominent in controls. The magnitude of temporal correlations over left
Scale temporo-central region predicted severity of depression in patients. LRTCs
in theta oscillations are a salient characteristic of healthy human brain and
may have diagnostic potential in psychiatric disorders.
Lee et al., 11 unmedicated EEG during resting DFA • Study compared LRTC in depressed subjects compared to controls. Study
2007 adults with MDD state found a significant linear correlation between severity of depression and
per DSM-IV, 11 scaling exponent over most channels. There was slower decay of LRTC
non-depressed and persistence of LRTC in depressed patients associated with severity of
age-matched depression over most cortical areas.
controls
Slezin 33 patients with EEG during resting Spectral density, • Study applied multifractal analysis of 1/f EEG rhythm fluctuations in patients
et al., 2007 paranoid state power-law with schizophrenia-related disorders compared to controls. Study found
schizophrenia, estimation increased instability and randomness of alpha rhythm in the F4 electrodes
schizotypal in schizophrenia-related disorders. Theta rhythm, on the contrary, showed
disorder, increased stability, regularity, and decreased complexity compared to
schizoaffective normal/healthy controls in the same disorders.
disorder per
ICD-10, 23 healthy
controls
Leistedt 10 untreated EEG during sleep spectral density, • Major depressive episodes are characterized by modification in correlation
et al., inpatients with an DFA structure of sleep EEG time series. Power law exponents were lower but
2007b acute episode of not statistically significant in stage 2 and NREM3-4. These changes in
MDD per DSM-IV, scaling behavior could provide an explanation for why patients with acute
14 healthy controls depression have sleep fragmentation.
Leistedt 10 untreated men EEG during sleep DFA • Goal of the study was to investigate the scaling properties of the sleep EEG
et al., in full or partial in remitted depressed men and to see whether history of MDD could
2007a remission from significantly alter dynamics of sleep EEG as a “scar marker.” No significant
MDD per DSM-IV; differences were noted between the two groups during sleep. There were
14 healthy controls no functional sequelae of past history of one or more unipolar MDD
episodes on fluctuation properties of sleep EEG. Study argued against
"depressive scar hypothesis" in which some permanent residual defect is
created by depressive disease. Study also confirmed LRTC in human sleep
EEG.
Tolkunov 50 healthy adults fMRI during visual power spectrum • Study examined whether patients with higher levels of trait anxiety would
et al., 2010 without psychiatric stimulus testing scale invariance show less efficient regulation of limbic responses using a visual stimulus
history (PSSI) during fMRI. Significant positive correlations were found between beta
frequency for limbic control circuit and trait anxiety. Dysregulated outputs
from limbic system in trait anxiety also led to dysregulated inputs to the
autonomic nervous system.
Radulescu 9 adults with fMRI scan during Power spectrum • Study hypothesized that paranoid schizophrenia might be result of
et al., 2012 schizophrenia per affect-valent scale invariance optimization abnormalities in the prefrontal-limbic circuit that regulates
DSM-IV, 26 healthy stimulus (PSSI), Poincare emotion. Patients and controls showed distinct PSSI in the
controls maps orbitofrontal/medial prefrontal cortex (Brodmann 10). Poincare maps
showed less variability in patients compared to controls. PSSI may be
useful for psychiatric diagnoses, partly for the spatial localization it affords.
Nikulin 18 adults with high-density EEG DFA, • Study evaluated LRTC in schizophrenia-related disorders compared to
et al., 2012 schizophrenia, 3 during resting state cross-frequency healthy controls. LRTCs were significantly decreased in patients with
with schizoaffective correlations schizophrenia in both alpha and beta frequency ranges. Authors
disorder per hypothesize that decrease in LRTC arises from increased variability in
DSM-IV, 28 healthy neuronal activity in patients with schizophrenia. Haloperidol dosing and
age- and scaling exponent did not correlate in electrodes or frequency bands, which
gender-matched argues against the effects of anti-psychotics on the noted differences in
controls LRTCs.
Bornas 56 healthy EEG during DFA, linear • Goal of study was to look for possible differences in LRTC in brain signals
et al., 2013 undergraduate eyes-open, correlation between from people with different negative emotion regulation strategies, including
students eyes-closed resting questionnaire rumination, that lead to a depressed lifestyle. Study identified linear positive
state; Beck scores and DFA correlations between the scaling exponents of both broad band and theta
Depression exponents band oscillations and negative emotion regulation strategies and depression
(Continued)

TABLE 6 | Continued
inventory and scores. Authors suggested that differences may exist between depressed
emotion regulation and non-depressed even before depression manifests, though depressed
questionnaires mental state clearly impacts the degree of correlation.
Tagliazucchi 15 healthy adult fMRI before, during, Variance and total • Study aimed to quantify the repertoire of neural states under the influence of
et al., 2014 subjects and after IV spectral power psychedelics like psilocybin. Changes in spectral scaling exponents and
psilocybin and variance of BOLD signals exclusively affected higher brain systems. Authors
placebo infusions found that psilocybin resulted in a larger repertoire of connectivity states at
rest than in control conditions, consistent with brain criticality.
Zhigalov 9 healthy adult EEG while resting Spectral density, • Study hypothesized that LRTC could be manipulated by closed-loop NFB
et al., 2016 subjects (ages with eyes-closed DFA stimulation. Over multiple sessions, there emerged a significant difference in
18−23 years old) during closed-loop LRTCs of α-band oscillations, with LRTCs stronger during NFB than sham.
and sham NFB Study served as a proof-of-concept that EEG LRTCs, and thus critical brain
sessions dynamics, could be modulated with closed-loop stimulation.
Ros et al., 40 healthy adults; fMRI scan before DFA • Study aimed to evaluate possibility of manipulating LRTC in patients
2014 21 adults with NFB, EEG during suffering from PTSD. Brain areas with low LRTCs in PTSD subjects
PTSD per DSM-IV, NFB or sham-NFB, normalized toward healthier population levels with application of
with 30 age- and and fMRI scan after neurofeedback compared to sham. Authors suggest that LRTC changes
gender-matched NFB seen with NFB are due to fluctuations in excitation-inhibition balance.
healthy controls
Gärtner 71 depressed adult EEG before and DFA • Study sought to understand whether neural dynamics improved in patients
et al., 2017 patients per after mindfulness with MDD after psychological treatment. Depressed subjects exhibited
DSM-IV, 25 healthy training or stress stronger LRTC in theta oscillations (4–7 Hz) at baseline compared to
controls reduction training controls. Following the psychological interventions both groups exhibited
decreased LRTC in the theta band, with marginal numerical differences
between the groups. Future of this research area will involve uncovering
how psychological interventions effectively reduce LRTCs.
Moran 23 patients with EEG with DFA, LORETA • Patients with schizophrenia showed area of significantly reduced beta-band
et al., 2019 schizophrenia per eyes-open LRTC over bilateral posterior regions compared to controls. Absence of
DSM-IV, 24 alpha band differences (contrary to Nikulin et al., 2012) could be related to
education-, eyes-open EEG used in this study.
handedness-, age-,
and
gender-matched
healthy controls
DSM-IV, Diagnostic and Statistical Manual, 4th edition; MDD, major depressive disorder; NFB, neurofeedback.
EEG and standardized low-resolution brain electromagnetic the excitation-inhibition balance in psychiatric disorders with
tomography (sLORETA) study of patients with schizophrenia decreased LRTC. A larger EEG and fMRI study of healthy
compared to healthy controls, there was a significant reduction patients confirmed this improvement in LRTC after closed-loop
in LRTC in the beta-band over the bilateral posterior regions NFB compared to sham. The study also involved a population
in patients compared to healthy controls. This study found no of patients with PTSD, whose LRTC improved with closed-
differences in the alpha frequency between patients and controls. loop NFB (Ros et al., 2014). The study authors speculated that
However, the study authors suggested that the eyes-open state of NFB leads to stronger excitation with an associated increase
their study (compared to the eyes-closed, resting state of Nikulin in temporal correlation and symptomatic improvement. While
et al., 2012) may have eliminated an underlying difference in it is unclear whether NFB can help in disorders characterized
that frequency range. More importantly, both studies confirm by super-criticality (e.g., seizures), NFB has already shown
the importance of LRTC as markers of network instability benefit in several disorders characterized to some degree by sub-
in schizophrenia. criticality, including schizophrenia (Surmeli et al., 2012) and
A much smaller number of studies have examined the role major depression (Escolano et al., 2014).
of neurofeedback (NFB) as a treatment modality for psychiatric A few criticality-based studies have taken an interest in the
disorders. A small study of healthy adults randomized to role of psychedelics in treating mental illnesses. Researchers
either closed-loop NFB or sham NFB found that, after three have focused on psilocybin, the active ingredient in “magic
sessions, LRTC in the alpha frequency band was stronger mushrooms,” and lysergic acid diethylamine (LSD). In an fMRI
in the NFB group compared to the sham (Zhigalov et al., study of healthy patients receiving psilocybin versus a placebo,
2016). This difference was statistically significant and did not study authors found an increased variance in BOLD signals in
involve any statistically significant topographical changes in the hippocampi of psychedelic recipients (Tagliazucchi et al.,
alpha power. The study provided proof-of-concept that closed- 2014). The increased variance implies increased synchronization,
loop NFB can restore critical brain dynamics by altering consistent with a super-critical state of brain dynamics. This

TABLE 7 | Summary of sleep-related criticality literature.
Nikulin and 12 healthy adults EEG with DFA • LRTC in alpha oscillations were not changed significantly by wakefulness
Brismar, eyes-open and level while beta oscillation scaling exponent significantly increased in the
2004 eyes-closed closed eye condition.
Increased synaptic activity associated with arousal/wakefulness may
interfere with dynamics of LRTC. Study confirmed existence of LRTC in
both awake and closed eyes but more consistently in closed eye state and
may be reflective of underlying SOC.
Weiss 22 healthy adults EEG during REM, Hurst exponent, • Study assessed various metrics of sleep EEG including monofractal,
et al., 2011 NREM2, NREM4 power spectral multifractal, and spectral power measures. Sleep stage discrimination with
sleep measures multifractal measure was superior to relative band powers, spectral edge
frequency, or Hurst exponent.
• Study found higher H exponent, DFA exponent, and fractal exponent in
deep sleep, while multifractal measure was decreased. These findings
indicate a decrease of multifractality and an increase in long memory in
deep sleep.
Dehghani 2 adult patients iEEG from temporal Power-law • Study investigated power-law distribution of neuronal avalanches (spikes)
et al., 2012 with intractable gyrus in awake estimation* from iEEG data. Neuronal avalanches (spikes) did not clearly follow
epilepsy state, REM, and power-law in awake, SWS, or REM states and instead followed closer to
slow-wave sleep exponential distribution. Positive and negative LFPs followed apparent
power laws with log-log analysis but closer examination with CDF-based
testing did not confirm power law and favored double exponential
distribution. In cases where power laws were seen with log-log analysis,
exponents were too high for SOC systems.
• These results contradict those of prior studies (Petermann et al., 2009;
Ribeiro et al., 2010) and perhaps could be harmonized with prior results by
taking into account recording methods or volume conduction effects.
Meisel 8 healthy adults EEG during 40 h of Power-law • Study evaluated evolution of criticality parameters during prolonged
et al., 2013 sustained estimation*, wakefulness. At the start of sleep deprivation, coherence potentials were
wakefulness branching organized as neuronal avalanches in space and time with power law −3/2
parameter, spectral and branching parameter 1.17, both of which suggest a system near
density criticality. With increased duration of wakefulness, size distributions of
coherence potentials and PLIs developed larger tails, an increase in
branching parameter, and an increase in mean synchronization while
variability of synchronization decreased.
• These findings suggested that, during sustained wakefulness, the neural
networks move from a critical to a supercritical state, perhaps as a result of
increased excitation and decreased inhibition (Shew et al., 2009; Yang
et al., 2012). Sleep might serve to reorganize network dynamics to critical
state in order to assure optimal computational capabilities while awake.
Priesemann 5 adults with iEEG (depth Power-law • Neuronal avalanches were recorded from intracranial depth electrodes in 5
et al., 2013 refractory epilepsy electrode) estimation*, epilepsy patients over two nights through all sleep stages. Avalanches were
branching described by power laws in all cases but with different dynamics depending
parameter on sleep stages. SWS showed the largest avalanches, wakefulness showed
intermediate-sized, while REM showed smallest. Differences in avalanche
distributions implied that not all vigilance states could be derived from SOC.
• Modeling suggested that human brain operates within subcritical regime,
near criticality where differences between vigilance states can be mediated
by small changes in effective synaptic strength which allow the brain to tune
closer to criticality (SWS) or farther away (REM). SWS showed increased
correlations between cortical areas due to increased criticality, while REM
sleep showed more fragmented dynamics than SWS and wakefulness.
Lo et al., 48 healthy adults Polysomnogram Probability transfer • Study found a power law distribution of wake and arousal durations in sleep
2013 and 48 recorded for 2 matrix, power-law using log-log analysis. Power-law exponents were different between
age-matched consecutive nights estimation patients with OSA and healthy controls.
adults with • Using novel probability transfer matrix and SOC, authors revealed sleep
obstructive sleep transition pathways that could be reduced to two basic and independent
apnea (OSA) transition paths. Study also found that sleep micro-architecture at scales
from seconds to minutes exhibits a non-equilibrium behavior reminiscent of
critical systems.
Tagliazucchi 63 healthy adults fMRI and EEG DFA • Study hypothesized that breakdown of LRTC would occur during descent
et al., 2013 during NREM sleep into deep sleep. Authors found that Hurst exponent decreased during N2
(Continued)

TABLE 7 | Continued
sleep confined to DMN and attention networks. Study also discovered that
autocorrelation in fronto-parietal areas diminish from wakefulness to deep
sleep.
Allegrini 29 healthy adults Polysomnogram Random walks, • Study hypothesized that a renewal point process describing fractal
et al., 2013 DFA, fractal intermittency could be a correlate of consciousness. Fractal intermittency
intermittency can be seen in EEG data by sequence of global rapid transition processes
(RTP) with power law distribution of waiting times. During sleep, Hurst
exponent switched from 0.75 in wake and REM phases to 0.5 in deep
sleep, suggesting fractal intermittency in wake and REM but short-time
correlations in SWS.
Allegrini 29 healthy adults Polysomnogram Random walks, • Study evaluated fractal intermittency (see Allegrini et al., 2013) during sleep.
et al., 2015 DFA, fractal While critical avalanches remained unchanged, there was a breakdown in
intermittency intermittency and functional connectivity during shallow and deep NREM
sleep. The authors provided a theory for fragmentation-induced
intermittency breakdown. The possible role of critical avalanches in
dreamless sleep is to provide rapid recovery of consciousness if stimuli
arouse the person out of sleep.
Colombo 52 adults with High-density EEG DFA • There were no differences in DFA exponents between ID and controls in any
et al., 2016 insomnia disorder with eyes-open frequency bands during EO or EC. However, during EO, individuals with
(ID), 42 age- and (EO) and worse sleep quality had stronger LRTC, suggesting that subjective insomnia
sex-matched eyes-closed (EC) complaints involve distinct processes in people with ID and controls.
controls However, the measurement of insomnia severity was based on subjective
report, not polysomnography. Future studies should examine
polysomnographic data as well as examine frequency (rather than
amplitude) fluctuations.
Meisel 8 healthy adults EEG during 40 h of DFA, • Study evaluated LRTCs in resting state human EEG during 40-h sleep
et al., 2017 sustained autocorrelation deprivation experiment. LRTCs declined as sleep deprivation progressed,
wakefulness function, spectral even when taking into account changes in signal power. LRTCs naturally
density emerged in vicinity of critical state. Authors argued that the increased
LRTCs seen in insomnia patients (Colombo et al., 2016) could be due to
signal power changes associated with worse sleep quality.
Bocaccio 18 healthy adults fMRI and EEG Power-law • Study observed scale-free hierarchy of co-activated connected clusters
et al., 2019 during wakefulness estimation* using point-process transformation of fMRI data recorded during wake and
and all sleep stages NREM sleep. Sleep stage had significant impact on scaling parameter of
power law, which was robust to spatial coarse-graining, alternative
statistical models, and disappearing with phase shuffling of fMRI time
series. These findings suggest the existence of larger clusters or avalanches
during N2 sleep. Criticality may help with the “pretty hard problem of
consciousness” by offering metrics that behave one way in conscious
states and differently in another.
Asterisk (*) represents power-law estimations that meet criteria equivalent to or more stringent than Clauset et al. (2009). SWS, slow-wave sleep; REM, rapid eye-
movement; NREM, non-rapid eye-movement; PLI, phase-locking interval; DMN, default mode network.
super-critical state could account for the hyper-associative There are multiple stages of sleep, including rapid eye movement
cognition characteristic of psychedelics (Carhart-Harris et al., (REM), non-rapid eye movement (NREM), which is further
2014; Carhart-Harris, 2018). While this review limits itself divided into light sleep (NREM1 and NREM2) and deep
to non-connectomic data, the works of Atasoy et al. (2017, sleep (NREM3 and NREM4). A typical night’s sleep runs
2018) deserve mention here. Using a unique connectome- through several cycles of each of these sleep stages. The
harmonic decomposition approach, her research team has found brain criticality literature centers on two key areas of sleep
in several studies that infusions of psychedelics “push” brain medicine – characterization of sleep stages and the physiology of
dynamics out of baseline sub-criticality toward criticality (Atasoy sleeping disorders.
et al., 2017, 2018). More and more studies are finding that Researchers have tried improving the classification of sleep
psychedelics are helpful adjuncts to psychotherapy (Atasoy et al., stages with the help of criticality-based metrics. The results of
2017). A psychedelic-induced transition out of a sub-critical these studies seem, for now, largely inconsistent. An EEG study
disorder like depression and closer to criticality may account of healthy patients found LRTC in brain oscillations in both
for this benefit. the open-eye and closed-eye states (Nikulin and Brismar, 2004).
LRTC was more consistent, however, in the closed-eye state, with
Sleep Medicine a significant difference in the scaling exponent between closed
Sleep medicine deals with human sleep and its associated and open eye conditions in the beta frequency range. Building
disorders, including insomnia, sleep apnea, and narcolepsy. on this initial result for LRTC, an EEG study of healthy patients

examined the ability of various fractal and spectral metrics to regime to avoid the risk of run-away excitation in a supercritical
distinguish REM, NREM2, and NREM4 (Weiss et al., 2011). state (Pearlmutter and Houghton, 2009, 2013).
The authors found that the multifractal metric was superior to An ECoG study of patients during sleep revealed power-
other metrics for sleep stage classification. From wakefulness to law distributions of neuronal avalanches in local field potentials
deep sleep, multifractality decreased while monofractality (e.g., (LFPs), but with different exponents depending on the sleep
the Hurst exponent) increased. The loss of multifractality and stage (Priesemann et al., 2013). After making adjustments for
disruptions in the 1/f noise in the deep sleep stages suggested the fact that LFP amplitude naturally increases from wakefulness
that sleep-specific brain rhythms (e.g., sleep spindles) disrupt the to deep sleep, the study authors found that slow-wave sleep
day-time self-similarity. Previous studies had shown an increase, (SWS or NREM3) displayed large avalanches. Wakefulness
not a decrease, in multifractality in deepening sleep (Ma et al., and REM sleep, on the other hand, displayed intermediate
2006). These differing results suggest that additional studies of and small avalanches, respectively. Because the study found
multifractality are needed. a branching parameter less than one during wakefulness, the
The increase in mono-fractality that accompanied deep sleep authors concluded that the awake human brain at rest is in
in the study by Weiss et al. (2011) suggested increased long- a subcritical state rather than in a critical state as had been
memory (LRTC) in sleep, similar to the increase in scaling previously argued (Meisel et al., 2013). The authors suggest that
exponent noted in the closed eye state by Nikulin and Brismar brain dynamics shift closer to criticality during SWS and then
(2004). Subsequent studies, however, have produced results that back into sub-criticality during REM sleep (Priesemann et al.,
seem to contradict these findings. Using EEG-fMRI data from 2013). Their data and additional computational modeling found
healthy adults in NREM sleep, Tagliazucchi et al. (2013) found that small changes in effective synaptic strength could tune the
that a breakdown in LRTC happened as patients moved into the brain closer to a critical state during SWS or closer to a subcritical
deeper sleep stages. Specifically, the Hurst exponent decreased state with fragmented dynamics during REM. Of note, this study
in the default-mode network (DMN) and attention centers of did not find power-law distributions under the more stringent
the brain during NREM2 (Tagliazucchi et al., 2013). Two studies criteria of Clauset et al. (2009). Moreover, a small study of
of polysomnogram data from a different cohort of healthy power-law distributions of cortical spikes and LFPs under awake,
adults found a similar result (Allegrini et al., 2013, 2015). The SWS, and REM sleep conditions could not confirm a power-law
monofractal metric (i.e., the Hurst exponent) decreased from distribution of those brain signals using the Clauset et al. (2009)
∼0.75 during wakefulness and REM, to ∼0.5 in NREM3 and criteria (Dehghani et al., 2012).
NREM4. Recall that a Hurst exponent of 0.5 indicates random, While the study of Dehghani et al. (2012) casts doubt on
white noise, while 0.75 suggests a moderate amount of positive these power-law findings, a recent study of EEG-fMRI data in
correlation. This finding suggested the existence of a breakdown healthy patients during sleep suggests that critical dynamics may
in LRTC when entering deep sleep. Complicating matters further be involved nonetheless (Bocaccio et al., 2019). A power-law
is that studies showing decreases in LRTC during sleep did not distribution of co-activated connected clusters (voxel groups)
control for changes in signal power that naturally occur during from fMRI data during sleep was confirmed using the Clauset
sleep (Meisel et al., 2017). More studies on the existence and et al. (2009) criteria. As in Priesemann et al. (2013), this study
changes of LRTC during sleep, taking into account fluctuations found larger neuronal avalanches in NREM compared to REM or
in signal power, are needed if these are to become useful in sleep wakefulness. Moreover, each particular sleep stage impacted the
stage classification or in understanding sleep microarchitecture. power-law exponent significantly – a result that was independent
There are also conflicting results in the literature when it of spatial coarse-graining of the fMRI data and which could not
comes to power-law distributions of various observables during be accounted for by the presence of evoked neural bistabilities
sleep. In an EEG study, researchers noted that subjects in (e.g., K complexes).
normal states of wakefulness exhibited a power-law distribution All in all, while there are discrepancies in the results obtained
of coherence potentials and phase-locking intervals (PLI)(Meisel from studies in this area, one can be optimistic about the
et al., 2013). Also, their EEG signals produced a branching role of criticality-based metrics in refining the understanding
parameter near one, which suggests a system at or near criticality of sleep stages and their classification. Research into the role
(Yang et al., 2012; Meisel et al., 2013). As patients were kept awake of LRTC (with adjustments made for signal power during
longer and deprived of sleep, the distributions of coherence sleep), branching parameters, power-law distributions of various
potentials and PLIs developed larger tails along with an increase metrics (coherence potential, PLI, LFP), mean synchronization,
in the branching parameter, a decrease in the variability of and other criticality parameters should lead to robust classifiers
synchronization, and an increase in mean synchronization – all of sleep stages.
suggestive of a supercritical state. The authors suggested that The conceptual framework of criticality is also helping
prolonged wakefulness leads to a supercritical state through researchers better understand various sleep pathologies (Iyer,
excess excitability of neurons (Shew et al., 2009), as is also thought 2018). Many animal and computational models show that critical
to occur in epilepsy. Meisel et al. (2013) proposed that, during dynamics play a central role in sleep stage transitions (Comte
restorative sleep, a decrease in synaptic strength should lead to et al., 2006; Wang et al., 2019; Lombardi et al., 2020). Human
a shift away from super-criticality, back toward a critical regime. studies are also beginning to capture this essential role by
This view is in line with a similar theory that sleep is designed studying various sleep disorders. In the previouslymentioned
to restore brain dynamics to a slightly subcritical or near-critical EEG study of sleep deprivation (Meisel et al., 2013), there was

a progressive decline in markers of criticality and an increase 2019). These include criticality between ordered and
in markers of supercriticality as the sleep deprivation worsened. chaotic phases called the “edge of chaos” (Boedecker
This finding supports the idea that sleep restores healthy brain et al., 2012), criticality between synchrony and asynchrony
function by bringing dynamics back toward criticality. This study (Botcharova et al., 2014), and multiple paradigms for
was followed up by an evaluation of LRTC in the same cohort the time-evolution of a critical phase transition as
under similar conditions of sleep deprivation (Meisel et al., 2017). in extended criticality, intermittent criticality, and self-
After controlling for signal amplitude changes, LRTC strength organized criticality (Saleur et al., 1996; Huang et al.,
declined as sleep deprivation progressed, consistent with the view 1998; Sammis and Smith, 1999; Bowman and Sammis,
that sleep restores brain criticality. 2004). These forms of dynamical criticality are also distinct
In contrast with forced sleep deprivation, insomnia disorder from statistical criticality (Mora and Bialek, 2011; Tkačik
(ID) leads to chronic sleep deprivation despite the patient’s et al., 2013). How these all inter-connect is a topic of
attempts to fall asleep. In a study of patients with ID, there were ongoing research. Rigorously defining these states and
no statistically significant differences in the LRTC during eyes- enforcing a more consistent vocabulary would allow for
open and eyes-closed testing between patients and their age and better study comparison.
sex-matched controls (Colombo et al., 2016). However, patients 2. Proving the existence of a control parameter has been
who subjectively reported lower quality sleep had increased difficult. Candidates have included synchronization,
LRTC during the eyes-open state. The study authors suggested excitation-inhibition balance, and synaptic conductance
that an increased excitation-inhibition ratio during wakefulness (Beggs and Timme, 2012). Feedback processes between
may translate into similar excitation during sleep, leading to these properties may make it difficult to prove that they
lower sleep quality. Similar processes seem to be at work in behave as genuine control parameters in isolation.
obstructive sleep apnea (OSA), one of the most common sleep 3. Many of the publications that lay the groundwork
disorders. A polysomnogram study of patients with OSA and for brain criticality did not have the benefit of a
healthy controls found a power-law distribution of wakefulness reliable statistical framework for verifying a power-law
durations during sleep in both groups (Lo et al., 2013). But distribution. While publications since Clauset et al. (2009)
the study did find a statistically significant difference between have mostly implemented strong statistical testing of power
the power-law exponents. This finding reinforces the prospect laws, many of the initial papers and even some more
that power-law exponents, perhaps in combination with other recent publications suffer from this deficit. Establishing
established metrics like the apnea-hypopnea index, could prove a standard pipeline of statistical analysis for common
to be a helpful marker of diagnosis and monitoring of OSA. In the datasets (e.g., EEG, MEG, LFP) – a kind of statistical “best
future, a study of LRTC in patients with OSA and ID compared practices”−will eliminate a source of confusion and help
to healthy patients in sleep could help determine whether the clarify the clinical situations in which power-law regimes
prevalent theory of sleep as a safety margin from supercriticality do and do not exist. Moreover, arguing for a mechanism
is correct. More broadly, the theory of brain criticality is likely that generates a power-law – whether criticality or another
to offer a different but complementary perspective on many mechanism−may be as important as the discovery of the
sleep disorders (including narcolepsy, restless legs syndrome, and power-law distribution itself (Stumpf and Porter, 2012).
circadian disorders). See Table 7 for a summary of findings from 4. In many publications, the objects of study are defined
the sleep-related criticality literature. differently from one paper to the next. Such is the
case, for example, with neuronal avalanches (Wilting and
Priesemann, 2019) and seizure energy (Worrell et al., 2002;
CONCLUSION Osorio et al., 2009). While flexibility in definitions may
help with making discoveries or adjusting for specific
This scoping review surveyed the brain criticality literature, experimental situations, frequent changes in definition
focusing on seven major domains of clinical application. make it difficult to compare results, which slows down the
Wherever possible, an effort was made to emphasize areas progress of the field as a whole.
of future research for those interested in pursuing a “critical 5. A common strategy, especially since the invention of
approach” to these clinical questions. In this concluding section, DFA, is to look for LRTC as an indication of criticality.
controversies that continue to be problematic for the field of brain However, recent research suggests there may be different
criticality as a whole are addressed. types of LRTC at play in these various systems and not
Brain criticality is both an established area of neuroscience distinguishing them carefully may be leading to false
research and yet remains controversial in several regards conclusions about criticality. For example, the discovery
(Wilting and Priesemann, 2019). of crucial events in the area of turbulence led to the
identification of “crucial event LRTC” or CELRTC (Bohara
1. Diverse and inconsistent uses of the terms “critical” et al., 2018; Culbreth et al., 2019). CELRTC emerges
and “criticality” have led to confusion. In this review, in critical systems, specifically self-organized temporal
criticality has mostly referred to avalanche dynamics that criticality (SOTC) and is based on a slowly-decaying, non-
behave at the limit between stability and instability. But stationary correlation function (Mahmoodi et al., 2017).
other variants of criticality exist (Wilting and Priesemann, CELRTC is distinct from Hurst exponent LRTC (HLRTC)

which is based on a slowlydecaying but stationary MEG, and fMRI, several of these techniques may transition to
correlation function, and which may not be indicative of the bedside if they prove helpful in the diagnosis, prognosis, or
underlying criticality. Future papers relying on detection treatment of diseases. In addition to more detailed studies in the
of LRTC should incorporate this methodology in order to clinical areas mentioned in this review, one can expect the next
clarify the origin of the long-range correlation. decade to see innovative studies, anchored in criticality, in areas
6. DFA was introduced for the study of non-stationary like addiction medicine, stroke, neuro-immunology, traumatic
datasets. Yet there is evidence that this approach may not brain injury, and headache medicine.
be adequate for that purpose (Bryce and Sprague, 2012). There are inevitable hurdles when concepts from one field (i.e.,
The introduction of a finite-size effect by DFA, leading statistical physics) are translated into another (i.e., biomedicine).
to artifact, may have led to spurious results in many Still, there can be no doubt that these hurdles are worthwhile if
publications mentioned in this review. Future research they open up new horizons for both the understanding and the
should clarify the relevance of DFA-based methods for the treatment of brain-related diseases.
purpose of LRTC detection.
Despite these challenges and controversies, there is also room

AUTHOR CONTRIBUTIONS
for optimism. There has been rapid growth in the number of new VZ evaluated the literature and wrote the manuscript.
articles published in this area. In fact, approximately two-thirds
of the clinical articles discussed in this review were published in
the last 8 years. The range of tools and concepts available from FUNDING
statistical physics and complexity science with which to tackle
these problems is staggeringly broad and continually expanding. This study was supported by the Department of Pediatrics at
Determining how these tools and concepts inter-relate and in the University of Texas Southwestern, which funded open access
which research situation to use them should be an ongoing focus publication fees. No other funding was received for this research.
of research. Insights from other sciences−including geophysics,
finance, applied physics, and signal processing, where these ideas
are also commonly circulated – will undoubtedly shape the future ACKNOWLEDGMENTS
of this already multidisciplinary field.
The future of criticality in the clinical arena will depend The author would like to thank Lina Chalak, MD and Jeffrey
in large part on the formation of multidisciplinary teams in McKinney, MD, Ph.D. for proposing the idea of this review.
which physicists, mathematicians, data scientists, and clinicians The author would also like to thank Mario Altamura, Ph.D. for
collaborate to better answer a clinical question through the lens kindly sending his manuscript as well as Didier Sornette, Ph.D.,
of criticality. While none of the applications mentioned in this Gregory Worrell, MD, Ph.D., Luca Cocchi, Ph.D., John Beggs,
review has yet to become mainstream or routine, it is conceivable Ph.D., Paolo Grigolini, Ph.D., and Michele Colombo, Ph.D. for
that, under the broad umbrella of quantitative analysis of EEG, sharing their insights.
REFERENCES Arviv, O., Medvedovsky, M., Sheintuch, L., Goldstein, A., and Shriki, O.
(2016). Deviations from critical dynamics in interictal epileptiform activity.
Achard, S., Delon-Martin, C., Vértes, P. E., Renard, F., Schenck, M., Schneider, F., J. Neurosci.36, 12276–12292. doi: 10.1523/JNEUROSCI.0809-16.2016
et al. (2012). Hubs of brain functional networks are radically reorganized in Atasoy, S., Roseman, L., Kaelen, M., Kringelbach, M. L., Deco, G., and
comatose patients. Proc. Natl. Acad. Sci. U.S.A.109, 20608–20613. doi: 10.1073/ Carhart-Harris, R. L. (2017). Connectome-harmonic decomposition of human
pnas.1208933109 brain activity reveals dynamical repertoire re-organization under LSD. Sci.
Allegrini, P., Paradisi, P., Menicucci, D., Laurino, M., Bedini, R., Piarulli, A., et al. Rep.7:17661. doi: 10.1038/s41598-017-17546-0
(2013). Sleep unconsciousness and breakdown of serial critical intermittency: Atasoy, S., Vohryzek, J., Deco, G., Carhart-Harris, R. L., and Kringelbach,
new vistas on the global workspace. Chaos Solit. Fract.55, 32–43. doi: 10.1016/j. M. L. (2018). Common neural signatures of psychedelics: frequency-specific
chaos.2013.05.019 energy changes and repertoire expansion revealed using connectome-harmonic
Allegrini, P., Paradisi, P., Menicucci, D., Laurino, M., Piarulli, A., and Gemignani, decomposition. Prog. Brain Res.242, 97–120. doi: 10.1016/bs.pbr.2018.
A. (2015). Self-organized dynamical complexity in human wakefulness and 08.009
sleep: different critical brain-activity feedback for conscious and unconscious Bak, P., Christensen, K., Danon, L., and Scanlon, T. (2002). Unified scaling law
states. Phys. Rev. E Stat. Nonlinear Soft Matt. Phys.92:032808. doi: 10.1103/ for earthquakes. Phys. Rev. Lett.99, 2509–2513. doi: 10.1103/PhysRevLett.88.
PhysRevE.92.032808 178501
Alonso, L. M., Proekt, A., Schwartz, T. H., Pryor, K. O., Cecchi, G. A., and Bak, P., and Tang, C. (1989). Earthquakes as a self-organized critical phenomenon.
Magnasco, M. O. (2014). Dynamical criticality during induction of anesthesia in J. Geophys. Res.94, 15635–15637. doi: 10.1029/jb094ib11p15635
human ECoG recordings. Front. Neural Circuits.8:20. doi: 10.3389/fncir.2014. Bak, P., Tang, C., and Wiesenfeld, K. (1987). Self-organized criticality: an
00020 explanation of the 1/f noise. Phys. Rev. Lett.59, 381–384. doi: 10.1103/
Altamura, M., Elvevag, B., Campi, G., De Salvia, M., Marasco, D., Ricci, A., PhysRevLett.59.381
et al. (2012). Toward scale-free like behavior under increasing cognitive load. Bak, P., Tang, C., and Wiesenfeld, K. (1988). Self-organized criticality. Phys. Rev.
Complexity18, 38–43. doi: 10.1002/cplx.21407 A38, 364–374. doi: 10.1103/PhysRevA.38.364

Ball, G., Aljabar, P., Zebari, S., Tusor, N., Arichi, T., Merchant, N., et al. (2014). Bressler, S. L., and Kelso, J. A. S. (2001). Cortical coordination dynamics and
Rich-club organization of the newborn human brain. Proc. Natl. Acad. Sci. cognition. Trends Cogn. Sci.5, 26–36. doi: 10.1016/S1364-6613(00)01564-3
U.S.A.111, 7456–7461. doi: 10.1073/pnas.1324118111 Bryce, R. M., and Sprague, K. B. (2012). Revisiting detrended fluctuation analysis.
Barabási, A. L., and Albert, R. (1999). Emergence of scaling in random networks. Sci. Rep.2:345. doi: 10.1038/srep00315
Science286, 509–512. doi: 10.1126/science.286.5439.509 Cao, F., Bitan, T., and Booth, J. R. (2008). Effective brain connectivity in children
Baron-Cohen, S., and Belmonte, M. K. (2005). AUTISM: a window onto the with reading difficulties during phonological processing. Brain Lang.107, 91–
development of the social and the analytic brain. Annu. Rev. Neurosci.28, 101. doi: 10.1016/j.bandl.2007.12.009
109–126. doi: 10.1146/annurev.neuro.27.070203.144137 Carhart-Harris, R. L. (2018). The entropic brain – revisited.
Bédard, C., Kröger, H., and Destexhe, A. (2006). Does the 1/f frequency scaling Neuropharmacology142, 167–178. doi: 10.1016/j.neuropharm.2018.03.010
of brain signals reflect self-organized critical states?Phys. Rev. Lett.97:118102. Carhart-Harris, R. L., Leech, R., Hellyer, P. J., Shanahan, M., Feilding, A.,
doi: 10.1103/PhysRevLett.97.118102 Tagliazucchi, E., et al. (2014). The entropic brain: a theory of conscious
Beggs, J. M. (2008). The criticality hypothesis: how local cortical networks might states informed by neuroimaging research with psychedelic drugs. Front. Hum.
optimize information processing. Philos. Trans. A Math. Phys. Eng. Sci.366, Neurosci.8:20. doi: 10.3389/fnhum.2014.00020
329–343. doi: 10.1098/rsta.2007.2092 Castaing, B., Gagne, Y., and Hopfinger, E. J. (1990). Velocity probability density
Beggs, J. M., and Timme, N. (2012). Being critical of criticality in the brain. Front. functions of high Reynolds number turbulence. Phys. D Nonlinear Phenom.46,
Physiol.3:163. doi: 10.3389/fphys.2012.00163 177–200. doi: 10.1016/0167-2789(90)90035-N
Belmonte, M. K., Allen, G., Beckel-Mitchener, A., Boulanger, L. M., Carper, Cavanagh, J. F., and Frank, M. J. (2014). Frontal theta as a mechanism for cognitive
R. A., and Webb, S. J. (2004). Autism and abnormal development of brain control. Trends Cogn. Sci.18, 414–421. doi: 10.1016/j.tics.2014.04.012
connectivity. J. Neurosci.24, 9228–9231. doi: 10.1523/JNEUROSCI.3340-04. Cerf, R., El Ouasdad, E. H., and Kahane, P. (2004). Criticality and synchrony
2004 of fluctuations in rhythmical brain activity: pretransitional effects in epileptic
Berthouze, L., James, L. M., and Farmer, S. F. (2010). Human EEG shows long- patients. Biol. Cybern.90, 239–255. doi: 10.1007/s00422-004-0463-9
range temporal correlations of oscillation amplitude in Theta, Alpha and Beta Christensen, K., Olami, Z., and Bak, P. (1992). Deterministic 1/f noise in
bands across a wide age range. Clin. Neurophysiol.121, 1187–1197. doi: 10.1016/ nonconserative models of self-organized criticality. Phys. Rev. Lett.68, 2417–
j.clinph.2010.02.163 2420. doi: 10.1103/PhysRevLett.68.2417
Bertschinger, N., and Natschläger, T. (2004). Real-time computation at the edge Clauset, A., Shalizi, C. R., andNewman, M. E. J (2009). Power-law distributions in
of chaos in recurrent neural networks. Neural Comput.16, 1413–1436. doi: empirical data. SIAM Rev.51, 661–703. doi: 10.1137/070710111
10.1162/089976604323057443 Cocchi, L., Gollo, L. L., Zalesky, A., and Breakspear, M. (2017). Criticality in the
Bocaccio, H., Pallavicini, C., Castro, M. N., Sanchez, S. M., De Pino, G., Laufs, brain: a synthesis of neurobiology, models and cognition. Prog. Neurobiol.158,
H., et al. (2019). The avalanche-like behaviour of large-scale haemodynamic 132–152. doi: 10.1016/j.pneurobio.2017.07.002
activity from wakefulness to deep sleep. J. R. Soc. Interface16:20190262. doi: Cohen, M. X. (2016). Midfrontal theta tracks action monitoring over multiple
10.1098/rsif.2019.0262 interactive time scales. Neuroimage141, 262–272. doi: 10.1016/j.neuroimage.
Boedecker, J., Obst, O., Lizier, J. T., Mayer, N. M., and Asada, M. (2012). 2016.07.054
Information processing in echo state networks at the edge of chaos. Theory Colombo, M. A., Wei, Y., Ramautar, J. R., Linkenkaer-Hansen, K., Tagliazucchi,
Biosci.131, 205–213. doi: 10.1007/s12064-011-0146-8 E., and Van Someren, E. J. W. (2016). More severe insomnia complaints in
Bohara, G., West, B. J., and Grigolini, P. (2018). Bridging waves and crucial events people with stronger long-range temporal correlations in wake resting-state
in the dynamics of the brain. Front. Physiol.9:1174. doi: 10.3389/fphys.2018. EEG. Front. Physiol.7:576. doi: 10.3389/fphys.2016.00576
01174 Colquhoun, H. L., Levac, D., O’Brien, K. K., Straus, S., Tricco, A. C., Perrier, L., et al.
Bonachela, J. A., De Franciscis, S., Torres, J. J., and Muoz, M. A. (2010). Self- (2014). Scoping reviews: time for clarity in definition, methods, and reporting.
organization without conservation: are neuronal avalanches generically critical? J. Clin. Epidemiol.67, 1291–1294. doi: 10.1016/j.jclinepi.2014.03.013
J. Stat. Mech. Theory Exp. 2010. doi: 10.1088/1742-5468/2010/02/P02015 Comte, J. C., Ravassard, P., and Salin, P. A. (2006). Sleep dynamics: a self-organized
Bonachela, J. A., and Muñoz, M. A. (2009). Self-organization without conservation: critical system. Phys. Rev. E73:056127. doi: 10.1103/PhysRevE.73.056127
true or just apparent scale-invariance?J. Stat. Mech. Theory Exp. 2009. doi: Cook, M. J., O’Brien, T. J., Berkovic, S. F., Murphy, M., Morokoff, A., Fabinyi,
10.1088/1742-5468/2009/09/P09009 G., et al. (2013). Prediction of seizure likelihood with a long-term, implanted
Bongers, A., Flynn, A. B., and Northoff, G. (2019). Is learning scale-free? Chemistry seizure advisory system in patients with drug-resistant epilepsy: a first-in-man
learning increases EEG fractal power and changes the power law exponent. study. Lancet Neurol.12, 563–571. doi: 10.1016/S1474-4422(13)70075-9
Neurosci. Res. doi: 10.1016/j.neures.2019.10.011 [Epub ahead of print] Cook, M. J., Varsavsky, A., Himes, D., Leyde, K., Berkovic, S. F., O’Brien, T., et al.
Bornas, X., Noguera, M., Balle, M., Morillas-Romero, A., Aguayo-Siquier, B., (2014). The dynamics of the epileptic brain reveal long-memory processes.
Tortella-Feliu, M., et al. (2013). Long-range temporal correlations in resting Front. Neurol.5:217. doi: 10.3389/fneur.2014.00217
EEG: its associations with depression-related emotion regulation strategies. Cruz, A. V., Mallet, N., Magill, P. J., Brown, P., and Averbeck, B. B. (2009). Effects
J. Psychophysiol.27, 60–66. doi: 10.1027/0269-8803/a000087 of dopamine depletion on network entropy in the external globus pallidus.
Botcharova, M., Berthouze, L., Brookes, M. J., Barnes, G. R., and Farmer, J. Neurophysiol.102, 1092–1102. doi: 10.1152/jn.00344.2009
S. F. (2015). Resting state MEG oscillations show long-range temporal Culbreth, G., West, B. J., and Grigolini, P. (2019). Entropic approach
correlations of phase synchrony that break down during finger movement. to the detection of crucial events. Entropy21, 1–20. doi: 10.3390/e2102
Front. Physiol.6:183. doi: 10.3389/fphys.2015.00183 0178
Botcharova, M., Farmer, S. F., and Berthouze, L. (2012). Power-law distribution Daffertshofer, A., Ton, R., Kringelbach, M. L., Woolrich, M., and Deco, G. (2018).
of phase-locking intervals does not imply critical interaction. Phys. Rev. Distinct criticality of phase and amplitude dynamics in the resting brain.
E. Stat. Nonlin. Soft Matter Phys.86:51920. doi: 10.1103/PhysRevE.86. Neuroimage180, 442–447. doi: 10.1016/j.neuroimage.2018.03.002
051920 Deco, G., and Jirsa, V. K. (2012). Ongoing cortical activity at rest: criticality,
Botcharova, M., Farmer, S. F., and Berthouze, L. (2014). Markers of criticality multistability, and ghost attractors. J. Neurosci.32, 3366–3375. doi: 10.1523/
in phase synchronization. Front. Syst. Neurosci.8:176. doi: 10.3389/fnsys.2014. JNEUROSCI.2523-11.2012
00176 Deco, G., Kringelbach, M. L., Jirsa, V. K., and Ritter, P. (2017). The dynamics of
Bowman, D. D., and Sammis, C. G. (2004). Intermittent criticality and the resting fluctuations in the brain: Metastability and its dynamical cortical core.
Gutenberg-Richter distribution. Pure Appl. Geophys.161, 1945–1956. doi: 10. Sci. Rep.7:3095. doi: 10.1038/s41598-017-03073-5
1007/s00024-004-2541-z Deco, G., Senden, M., and Jirsa, V. (2012). How anatomy shapes dynamics: a semi-
Breakspear, M., and Terry, J. R. (2002). Detection and description of analytical study of the brain at rest by a simple spin model. Front. Comput.
non-linear interdependence in normal multichannel human EEG Neurosci.6:68. doi: 10.3389/fncom.2012.00068
data. Clin. Neurophysiol.113, 735–753. doi: 10.1016/S1388-2457(02)0 Dehghani, N., Hatsopoulos, N. G., Haga, Z. D., Parker, R. A., Greger, B., Halgren,
0051-2 E., et al. (2012). Avalanche analysis from multielectrode ensemble recordings in

cat, monkey, and human cerebral cortex during wakefulness and sleep. Front. Hartley, C., Berthouze, L., Mathieson, S. R., Boylan, G. B., Rennie, J. M., and
Physiol.3:302. doi: 10.3389/fphys.2012.00302 Farmer, S. F. (2012). Long-range temporal correlations in the EEG bursts
Del Papa, B., Priesemann, V., and Triesch, J. (2017). Criticality meets learning: of human preterm babies. PLoS One7:e0031543. doi: 10.1371/journal.pone.
criticality signatures in a self-organizing recurrent neural network. PLoS 0031543
One12:e0178683. doi: 10.1371/journal.pone.0178683 Hausdorff, J. M. (2009). Gait dynamics in Parkinson’s disease: common and
Dimitriadis, S. I., Laskaris, N. A., Simos, P. G., Micheloyannis, S., Fletcher, distinct behavior among stride length, gait variability, and fractal-like scaling.
J. M., Rezaie, R., et al. (2013). Altered temporal correlations in resting-state Chaos19:026113. doi: 10.1063/1.3147408
connectivity fluctuations in children with reading difficulties detected via MEG. He, B. J. (2011). Scale-free properties of the functional magnetic resonance
Neuroimage83, 307–317. doi: 10.1016/j.neuroimage.2013.06.036 imaging signal during rest and task. J. Neurosci.31, 13786–13795. doi: 10.1523/
Escolano, C., Navarro-Gil, M., Garcia-Campayo, J., Congedo, M., De Ridder, JNEUROSCI.2111-11.2011
D., and Minguez, J. (2014). A controlled study on the cognitive effect of He, B. J. (2014). Scale-free brain activity: past, present, and future. Trends Cogn.
alpha neurofeedback training in patients with major depressive disorder. Front. Sci.18, 480–487. doi: 10.1016/j.tics.2014.04.003
Behav. Neurosci.8:296. doi: 10.3389/fnbeh.2014.00296 Hesse, J., and Gross, T. (2014). Self-organized criticality as a fundamental property
Euler, M. J., Wiltshire, T. J., Niermeyer, M. A., and Butner, J. E. (2016). Working of neural systems. Front. Syst. Neurosci.8:166. doi: 10.3389/fnsys.2014.00166
memory performance inversely predicts spontaneous delta and theta-band Hobbs, J. P., Smith, J. L., and Beggs, J. M. (2010). Aberrant neuronal avalanches in
scaling relations. Brain Res.1637, 22–33. doi: 10.1016/j.brainres.2016.02.008 cortical tissue removed from juvenile epilepsy patients. J. Clin. Neurophysiol.27,
Ezaki, T., Fonseca Dos Reis, E., Watanabe, T., Sakaki, M., and Masuda, N. (2020). 380–386. doi: 10.1097/WNP.0b013e3181fdf8d3
Closer to critical resting-state neural dynamics in individuals with higher fluid Hoeft, F., Meyler, A., Hernandez, A., Juel, C., Taylor-Hill, H., Martindale, J. L., et al.
intelligence. Commun. Biol.3:52. doi: 10.1038/s42003-020-0774-y (2007). Functional and morphometric brain dissociation between dyslexia and
Fagerholm, E. D., Lorenz, R., Scott, G., Dinov, M., Hellyer, P. J., Mirzaei, N., reading ability. Proc. Natl. Acad. Sci. U.S.A.104, 4234–4239. doi: 10.1073/pnas.
et al. (2015). Cascades and cognitive state: focused attention incurs subcritical 0609399104
dynamics. J. Neurosci.35, 4626–4634. doi: 10.1523/JNEUROSCI.3694-14.2015 Hoffmann, H., and Payton, D. W. (2018). Optimization by self-organized
Fiest, K. M., Sauro, K. M., Wiebe, S., Patten, S. B., Kwon, C. S., Dykeman, J., criticality. Sci. Rep.8:2358. doi: 10.1038/s41598-018-20275-7
et al. (2017). Prevalence and incidence of epilepsy. Neurology88, 296–303. doi: Hohlefeld, F. U., Huebl, J., Huchzermeyer, C., Schneider, G.-H., Schoenecker, T.,
10.1212/WNL.0000000000003509 Kuehn, A. A., et al. (2012). Long-range temporal correlations in the subthalamic
Fraiman, D., Balenzuela, P., Foss, J., and Chialvo, D. R. (2009). Ising-like dynamics nucleus of patients with Parkinson’s disease. Eur. J. Neurosci.36, 2812–2821.
in large-scale functional brain networks. Phys. Rev. E Stat. Nonlinear, Soft Matt. doi: 10.1111/j.1460-9568.2012.08198.x
Phys.79(Pt 1):061922. doi: 10.1103/PhysRevE.79.061922 Hove, M. J., Suzuki, K., Uchitomi, H., Orimo, S., and Miyake, Y. (2012).
Fransson, P., Metsäranta, M., Blennow, M., Åden, U., Lagercrantz, H., and Interactive rhythmic auditory stimulation reinstates natural 1/f timing in gait
Vanhatalo, S. (2013). Early development of spatial patterns of power-law of parkinson’s patients. PLoS One7:e32600. doi: 10.1371/journal.pone.0032600
frequency scaling in fMRI resting-state and EEG data in the newborn brain. Huang, Y., Saleur, H., Sammis, C., and Sornette, D. (1998). Precursors, aftershocks,
Cereb. Cortex23, 638–646. doi: 10.1093/cercor/bhs047 criticality and self-organized criticality. Europhys. Lett.41, 43–48. doi: 10.1209/
Frette, V., Christensen, K., Malthe-Sørenssen, A., Feder, J., Jøssang, T., and Meakin, epl/i1998-00113-x
P. (1996). Avalanche dynamics in a pile of rice. Nature379, 49–52. doi: 10.1038/ Irrmischer, M., Houtman, S. J., Mansvelder, H. D., Tremmel, M., Ott, U., and
379049a0 Linkenkaer-Hansen, K. (2018a). Controlling the temporal structure of brain
Friedman, N., Ito, S., Brinkman, B. A. W., Shimono, M., Deville, R. E. L., oscillations by focused attention meditation. Hum. Brain Mapp.39, 1825–1838.
Dahmen, K. A., et al. (2012). Universal critical dynamics in high resolution doi: 10.1002/hbm.23971
neuronal avalanche data. Phys. Rev. Lett.108:208102. doi: 10.1103/PhysRevLett. Irrmischer, M., Poil, S.-S., Mansvelder, H. D., Intra, F. S., and Linkenkaer-
108.208102 Hansen, K. (2018b). Strong long-range temporal correlations of beta/gamma
Friston, K., Brown, H. R., Siemerkus, J., and Stephan, K. E. (2016). The oscillations are associated with poor sustained visual attention performance.
dysconnection hypothesis (2016). Schizophr. Res.176, 83–94. doi: 10.1016/j. Eur. J. Neurosci.48, 2674–2683. doi: 10.1111/ejn.13672
schres.2016.07.014 Iyer, K. K. (2018). Sleep, wake, and critical brain states: corollaries from brain
Frohlich, J., Irimia, A., and Jeste, S. S. (2015). Trajectory of frequency stability in dynamics. Front. Neurosci.12:948. doi: 10.3389/fnins.2018.00948
typical development. Brain Imaging Behav.9, 5–18. doi: 10.1007/s11682-014- Iyer, K. K., Roberts, J. A., Hellström-Westas, L., Wikström, S., Hansen Pupp, I.,
9339-3 Ley, D., et al. (2015). Cortical burst dynamics predict clinical outcome early
Gärtner, M., Irrmischer, M., Winnebeck, E., Fissler, M., Huntenburg, J. M., in extremely preterm infants. Brain138(Pt 8), 2206–2218. doi: 10.1093/brain/
Schroeter, T. A., et al. (2017). Aberrant long-range temporal correlations awv129
in depression are attenuated after psychological treatment. Front. Hum. Jannesari, M., Saeedi, A., Zare, M., Ortiz-Mantilla, S., Plenz, D., and Benasich, A. A.
Neurosci.11:340. doi: 10.3389/fnhum.2017.00340 (2020). Stability of neuronal avalanches and long-range temporal correlations
Gautam, S. H., Hoang, T. T., McClanahan, K., Grady, S. K., and Shew, W. L. (2015). during the first year of life in human infants. Brain Struct. Funct.225, 1169–
Maximizing sensory dynamic range by tuning the cortical state to criticality. 1183. doi: 10.1007/s00429-019-02014-4
PLoS Comput. Biol.11:e1004576. doi: 10.1371/journal.pcbi.1004576 Jia, H., Li, Y., and Yu, D. (2018). Attenuation of long-range temporal correlations
Gisiger, T., Linkenkaer-Hansen, K., Nikouline, V. V., Palva, J. M., Ilmoniemi, of neuronal oscillations in young children with autism spectrum disorder.
R. J., Worrell, G. A., et al. (2014). Spike avalanches in vivo suggest a driven, NeuroImage Clin.20, 424–432. doi: 10.1016/j.nicl.2018.08.012
slightly subcritical brain state. Front. Syst. Neurosci.9:108. doi: 10.3389/fnsys. Jiang, L., Qiao, K., Sui, D., Zhang, Z., and Dong, H. (2019). Functional criticality
2014.00108 in the human brain?: physiological, behavioral and neurodevelopmental
Haldeman, C., and Beggs, J. M. (2005). Critical branching captures activity in living correlates. PLoS One14:e0213690. doi: 10.1371/journal.pone.0213690
neural networks and maximizes the number of metastable states. Phys. Rev. Jiang, L., Sui, D., Qiao, K., Dong, H.-M., Chen, L., and Han, Y. (2018). Impaired
Lett.94:058101. doi: 10.1103/PhysRevLett.94.058101 functional criticality of human brain during Alzheimer’s disease progression.
Hanslmayr, S., Staudigl, T., and Fellner, M. C. (2012). Oscillatory power decreases Sci. Rep.8:1324. doi: 10.1038/s41598-018-19674-7
and long-term memory: the information via desynchronization hypothesis. Kinouchi, O., and Copelli, M. (2006). Optimal dynamical range of excitable
Front. Hum. Neurosci.6:74. doi: 10.3389/fnhum.2012.00074 networks at criticality. Nat. Phys.2, 348–351. doi: 10.1038/nphys289
Hardstone, R., Poil, S., Schiavone, G., Jansen, R., Nikulin, V. V., Mansvelder, H. D., Kitzbichler, M. G., Smith, M. L., Christensen, S. R., and Bullmore, E. (2009).
et al. (2012). Detrended fluctuation analysis?: a scale-free view on neuronal Broadband criticality of human brain network synchronization. PLoS Comput.
oscillations. Front. Physiol.3:450. doi: 10.3389/fphys.2012.00450 Biol.5:e1000314. doi: 10.1371/journal.pcbi.1000314
Harris, T. (1963). The Theory of Branching Processes. Berlin and Heidelberg: Kramer, M. A., Truccolo, W., Eden, U. T., Lepage, K. Q., Hochberg, L. R., Eskandar,
Springer-Verlag OHG. E. N., et al. (2012). Human seizures self-terminate across spatial scales via a

critical transition. Proc. Natl. Acad. Sci. U.S.A.109, 21116–21121. doi: 10.1073/ Meisel, C., and Kuehn, C. (2012). Scaling effects and spatio-temporal multilevel
pnas.1210047110 dynamics in epileptic seizures. PLoS One7:e30371. doi: 10.1371/journal.pone.
Krzemiński, D., Kamiński, M., Marchewka, A., and Bola, M. (2017). Breakdown 0030371
of long-range temporal correlations in brain oscillations during general Meisel, C., and Loddenkemper, T. (2019). Seizure prediction and intervention.
anesthesia. Neuroimage159, 146–158. doi: 10.1016/j.neuroimage.2017.07.047 Neuropharmacology172:107898. doi: 10.1016/j.neuropharm.2019.107898
Kwok, E. Y. L., Cardy, O., Allman, B., Allen, P., and Herrmann, B. (2019). Meisel, C., Olbrich, E., Shriki, O., and Achermann, P. (2013). Fading signatures of
Dynamics of spontaneous alpha activity correlate with language ability in young critical brain dynamics during sustained wakefulness in humans. J. Neurosci.33,
children. Behav. Brain Res.359, 56–65. doi: 10.1101/679670 17363–17372. doi: 10.1523/JNEUROSCI.1516-13.2013
Lai, M. C., Lombardo, M. V., Chakrabarti, B., Sadek, S. A., Pasco, G., Wheelwright, Meisel, C., Storch, A., Hallmeyer-Elgner, S., Bullmore, E., and Gross, T. (2012).
S. J., et al. (2010). A shift to randomness of brain oscillations in people with Failure of adaptive self-organized criticality during epileptic seizure attacks.
autism. Biol. Psychiatry68, 1092–1099. doi: 10.1016/j.biopsych.2010.06.027 PLoS Comput. Biol.8:e1002312. doi: 10.1371/journal.pcbi.1002312
Latham, P. E., and Nirenberg, S. (2004). Computing and stability in cortical Meresse, I. G., Zilbovicius, M., Boddaert, N., Robel, L., Philippe, A., Sfaello, I.,
networks. Neural Comput.16, 1385–1412. doi: 10.1162/089976604323057434 et al. (2005). Autism severity and temporal lobe functional abnormalities. Ann.
Lee, J. S., Yang, B. H., Lee, J. H., Choi, J. H., Choi, I. G., and Kim, S. B. (2007). Neurol.58, 466–469. doi: 10.1002/ana.20597
Detrended fluctuation analysis of resting EEG in depressed outpatients and Milton, J. G. (2012). Neuronal avalanches, epileptic quakes and other transient
healthy controls. Clin. Neurophysiol.118, 2489–2496. doi: 10.1016/j.clinph.2007. forms of neurodynamics. Eur. J. Neurosci.36, 2156–2163. doi: 10.1111/j.1460-
08.001 9568.2012.08102.x
Leistedt, S., Dumont, M., Coumans, N., Lanquart, J. P., Jurysta, F., and Linkowski, Minadakis, G., Ventouras, E., Gatzonis, S. D., Siatouni, A., Tsekou, H., Kalatzis, I.,
P. (2007a). The modifications of the long-range temporal correlations of the et al. (2014). Dynamics of regional brain activity in epilepsy: a cross-disciplinary
sleep EEG due to major depressive episode disappear with the status of study on both intracranial and scalp-recorded epileptic seizures. J. Neural
remission. Neuroscience148, 782–793. doi: 10.1016/j.neuroscience.2007.06.032 Eng.11:026012. doi: 10.1088/1741-2560/11/2/026012
Leistedt, S., Dumont, M., Lanquart, J. P., Jurysta, F., and Linkowski, P. (2007b). Mitzenmacher, M. (2004). A brief history of generative models for power law
Characterization of the sleep EEG in acutely depressed men using detrended and lognormal distributions. Internet Math.1, 226–251. doi: 10.1080/15427951.
fluctuation analysis. Clin. Neurophysiol.118, 940–950. doi: 10.1016/j.clinph. 2004.10129088
2007.01.003 Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., Altman, D., Antes, G., et al.
Li, X., Polygiannakis, J., Kapiris, P., Peratzakis, A., Eftaxias, K., and Yao, X. (2005). (2009). Preferred reporting items for systematic reviews and meta-analyses: the
Fractal spectral analysis of pre-epileptic seizures in terms of criticality. J. Neural PRISMA statement. PLoS Med.6:e1000097. doi: 10.1371/journal.pmed.1000097
Eng.2, 11–16. doi: 10.1088/1741-2560/2/2/002 Montez, T., Poil, S. S., Jones, B. F., Manshanden, I., Verbunt, J. P. A., Van
Linkenkaer-Hansen, K., Monto, S., Rytsälä, H., Suominen, K., Isometsä, E., and Dijk, B. W., et al. (2009). Altered temporal correlations in parietal alpha and
Kähkönen, S. (2005). Breakdown of long-range temporal correlations in theta prefrontal theta oscillations in early-stage Alzheimer disease. Proc. Natl. Acad.
oscillations in patients with major depressive disorder. J. Neurosci.25, 10131– Sci. U.S.A.106, 1614–1619. doi: 10.1073/pnas.0811699106
10137. doi: 10.1523/JNEUROSCI.3244-05.2005 Monto, S., Vanhatalo, S., Holmes, M. D., and Palva, J. M. (2007). Epileptogenic
Liu, X., Ward, B. D., Binder, J. R., Li, S.-J., and Hudetz, A. G. (2014). Scale- neocortical networks are revealed by abnormal temporal dynamics in seizure-
free functional connectivity of the brain is maintained in anesthetized healthy free subdural EEG. Cereb. Cortex17, 1386–1393. doi: 10.1093/cercor/bhl049
participants but not in patients with unresponsive wakefulness syndrome. PLoS Mora, T., and Bialek, W. (2011). Are biological systems poised at criticality?J. Stat.
One9:e92182. doi: 10.1371/journal.pone.0092182 Phys.144, 268–302. doi: 10.1007/s10955-011-0229-4
Lo, C. C., Bartsch, R. P., and Ivanov, P. C. (2013). Asymmetry and basic pathways in Moran, J. K., Michail, G., Heinz, A., Keil, J., and Senkowski, D. (2019). Long-
sleep-stage transitions. Europhys. Lett.102:10008. doi: 10.1209/0295-5075/102/ range temporal correlations in resting state beta oscillations are reduced in
10008 schizophrenia. Front. Psychiatry10:517. doi: 10.3389/fpsyt.2019.00517
Lombardi, F., Gómez-Extremera, M., Bernaola-Galván, P., Vetrivelan, R., Saper, Moretti, P., and Munoz, M. A. (2013). Griffiths phases and the stretching of
C. B., Scammell, T. E., et al. (2020). Critical dynamics and coupling in criticality in brain networks. Nat. Commun.4:2521. doi: 10.1038/ncomms3521
bursts of cortical rhythms indicate non-homeostatic mechanism for sleep- Mormann, F., Andrzejak, R. G., Elger, C. E., and Lehnertz, K. (2007). Seizure
stage transitions and dual role of VLPO neurons in both sleep and wake. prediction: the long and winding road. Brain.130(Pt 2), 314–333. doi: 10.1093/
J. Neurosci.40, 171–190. doi: 10.1523/JNEUROSCI.1278-19.2019 brain/awl241
Ma, Q., Ning, X., Wang, J., and Bian, C. (2006). A new measure to characterize Morrison, J. H., and Hof, P. R. (1997). Life and death of neurons in the aging brain.
multifractality of sleep electroencephalogram. Chinese Sci. Bull.51, 3259–3264. Science278, 412–419. doi: 10.1126/science.278.5337.412
doi: 10.1007/s11434-006-2213-y Mu, J., Rampp, S., Carrette, E., Roessler, K., Sommer, B., Schmitt, F. C., et al. (2014).
Ma, Z., Turrigiano, G. G., Wessel, R., and Hengen, K. B. (2019). Cortical circuit Clinical relevance of source location in frontal lobe epilepsy and prediction of
dynamics are homeostatically tuned to criticality in vivo. Neuron104, 655– postoperative long-term outcome. Seizure23, 553–559. doi: 10.1016/j.seizure.
664.e4. doi: 10.1016/j.neuron.2019.08.031 2014.04.006
Maass, W., Natschläger, T., and Markram, H. (2002). Real-time computing without Newman, M. E. J. (2005). Power laws, pareto distributions and Zipf ’s law. Contemp.
stable states: a new framework for neural computation based on perturbations. Phys.46, 323–351. doi: 10.1080/00107510500052444
Neural Comput.14, 2531–2560. doi: 10.1162/089976602760407955 Nikulin, V. V., and Brismar, T. (2004). Long-range temporal correlations in alpha
Mahmoodi, K., West, B. J., and Grigolini, P. (2017). Self-organizing complex and beta oscillations: effect of arousal level and test-retest reliability. Clin.
networks: individual versus global rules. Front. Physiol.8:478. doi: 10.3389/ Neurophysiol.115, 1896–1908. doi: 10.1016/j.clinph.2004.03.019
fphys.2017.00478 Nikulin, V. V., Jönsson, E. G., and Brismar, T. (2012). Attenuation of long-range
Mares, J., Vysata, O., Prochazka, A., and Valis, M. (2013). Age-dependent complex temporal correlations in the amplitude dynamics of alpha and beta neuronal
noise fluctuations in the brain. Physiol. Meas.34, 1269–1279. doi: 10.1088/0967- oscillations in patients with schizophrenia. Neuroimage61, 162–169. doi: 10.
3334/34/10/1269 1016/j.neuroimage.2012.03.008
Marinazzo, D., Pellicoro, M., Wu, G.-R., Angelini, L., Cortes, J., and Stramaglia, Nishimori, H., and Ortiz, G. (2011). Elements of Phase Transitions and
S. (2013). Information transfer of an Ising model on a brain network. BMC Critical Phenomena. Oxford: Oxford University Press, doi: 10.1093/acprof:oso/
Neurosci.14(Suppl. 1):376. doi: 10.1186/1471-2202-14-s1-p376 9780199577224.001.0001
Mehta, A. P., Dahmen, K. A., and Ben-Zion, Y. (2006). Universal mean moment Omori, F. (1895). On the aftershocks of earthquakes. J. Coll. Sci. Imp. Univ. Tokyo
rate profiles of earthquake ruptures. Phys. Rev. E Stat. Nonlinear, Soft Matter 7, 111–120.
Phys.73(Pt 2):056104. doi: 10.1103/PhysRevE.73.056104 Osorio, I., Frei, M. G., Sornette, D., and Milton, J. (2009). Pharmaco-
Meisel, C., Bailey, K., Achermann, P., and Plenz, D. (2017). Decline of long-range resistant seizures: self-triggering capacity, scale-free properties and
temporal correlations in the human brain during sustained wakefulness. Sci. predictability?Eur. J. Neurosci.30, 1554–1558. doi: 10.1111/j.1460-9568.2009.
Rep.7, 1–11. doi: 10.1038/s41598-017-12140-w 06923.x

Osorio, I., Frei, M. G., Sornette, D., Milton, J., and Lai, Y. C. (2010). Epileptic Roberts, J. A., Iyer, K. K., Vanhatalo, S., and Breakspear, M. (2014b). Critical
seizures: quakes of the brain?Phys. Rev. E Stat. Nonlinear, Soft Matter role for resource constraints in neural models. Front. Syst. Neurosci.8:154. doi:
Phys.82:021919. doi: 10.1103/PhysRevE.82.021919 10.3389/fnsys.2014.00154
Ouyang, G., Hildebrandt, A., Schmitz, F., and Herrmann, C. S. (2020). Rolls, E. T., Loh, M., Deco, G., and Winterer, G. (2008). Computational models
Decomposing alpha and 1/f brain activities reveals their differential associations of schizophrenia and dopamine modulation in the prefrontal cortex. Nat. Rev.
with cognitive processing speed. Neuroimage205:116304. doi: 10.1016/j. Neurosci.9, 696–709. doi: 10.1038/nrn2462
neuroimage.2019.116304 Ros, T. J., Baars, B., Lanius, R. A., and Vuilleumier, P. (2014). Tuning pathological
Padilla, N., Eklöf, E., Mårtensson, G. E., Bölte, S., Lagercrantz, H., and Ådén, brain oscillations with neurofeedback: a systems neuroscience framework.
U. (2017). Poor brain growth in extremely preterm neonates long before the Front. Hum. Neurosci.8:1008. doi: 10.3389/fnhum.2014.01008
onset of autism spectrum disorder symptoms. Cereb. Cortex27, 1245–1252. Ruiz, M. H., Bin Hong, S., Hennig, H., Altenmueller, E., and Kuehn, A. A. (2014).
doi: 10.1093/cercor/bhv300 Long-range correlation properties in timing of skilled piano performance:
Padilla, N., Saenger, V. M., Van Hartevelt, T. J., Fernandes, H. M., Lennartsson, the influence of auditory feedback and deep brain stimulation. Front.
F., Andersson, J. L. R., et al. (2020). Breakdown of whole-brain dynamics Psychol.5:1030. doi: 10.3389/fpsyg.2014.01030
in preterm-born children. Cereb. Cortex30, 1159–1170. doi: 10.1093/cercor/ Sachs, M. K., Yoder, M. R., Turcotte, D. L., Rundle, J. B., and Malamud, B. D. (2012).
bhz156 Black swans, power laws, and dragon-kings: earthquakes, volcanic eruptions,
Papanikolaou, S., Bohn, F., Sommer, R. L., Durin, G., Zapperi, S., and Sethna, J. P. landslides, wildfires, floods, and SOC models. Eur. Phys. J. Spec. Top.205,
(2011). Universality beyond power laws and the average avalanche shape. Nat. 167–182. doi: 10.1140/epjst/e2012-01569-3
Phys.7, 316–320. doi: 10.1038/nphys1884 Saleur, H., Sammis, C. G., and Sornette, D. (1996). Discrete scale invariance,
Parish, L. M., Worrell, G. A., Cranstoun, S. D., Stead, S. M., Pennell, P., and complex fractal dimensions, and log-periodic fluctuations in seismicity.
Litt, B. (2004). Long-range temporal correlations in epileptogenic and non- J. Geophys. Res. Solid Earth.101, 17661–17677. doi: 10.1029/96jb00876
epileptogenic human hippocampus. Neuroscience 125, 1069–1076. doi: 10.1016/ Sammis, C. G., and Smith, S. W. (1999). Seismic cycles and the evolution of stress
j.neuroscience.2004.03.002 correlation in cellular automaton models of finite fault networks. Pure Appl.
Parr, T., Rees, G., and Friston, K. J. (2018). Computational neuropsychology Geophys.155, 307–334. doi: 10.1007/s000240050267
and bayesian inference. Front. Hum. Neurosci.12:61. doi: 10.3389/fnhum.2018. Scheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V.,
00061 et al. (2009). Early-warning signals for critical transitions. Nature461, 53–59.
Pearlmutter, B. A., and Houghton, C. J. (2009). A new hypothesis for sleep: tuning doi: 10.1038/nature08227
for criticality. Neural Comput.21, 1622–1641. doi: 10.1162/neco.2009.05-08- Shew, W. L., and Plenz, D. (2013). The functional benefits of criticality in the cortex.
787 Neuroscientist19, 88–100. doi: 10.1177/1073858412445487
Pearlmutter, B. A., and Houghton, C. J. (2013). Dreams, mnemonics, and tuning Shew, W. L., Yang, H., Petermann, T., Roy, R., and Plenz, D. (2009). Neuronal
for criticality. Behav. Brain Sci.36, 625–659. doi: 10.1017/S0140525X1300 avalanches imply maximum dynamic range in cortical networks at criticality.
1404 J. Neurosci.29, 15595–15600. doi: 10.1523/JNEUROSCI.3864-09.2009
Peng, C. K., Buldyrev, S. V., Goldberger, A. L., Havlin, S., Mantegna, R. N., Simons, Shew, W. L., Yang, H., Yu, S., Roy, R., and Plenz, D. (2011). Information capacity
M., et al. (1995a). Statistical properties of DNA sequences. Phys. A Stat. Mech. and transmission are maximized in balanced cortical networks with neuronal
its Appl.221, 180–192. doi: 10.1016/0378-4371(95)00247-5 avalanches. J. Neurosci.31, 55–63. doi: 10.1523/JNEUROSCI.4637-10.2011
Peng, C. K., Havlin, S., Stanley, H. E., and Goldberger, A. L. (1995b). Quantification Shriki, O., Alstott, J., Carver, F., Holroyd, T., Henson, R. N. A., Smith, M. L.,
of scaling exponents and crossover phenomena in nonstationary heartbeat time et al. (2013). Neuronal avalanches in the resting MEG of the human brain.
series. Chaos5, 82–87. doi: 10.1063/1.166141 J. Neurosci.33, 7079–7090. doi: 10.1523/JNEUROSCI.4286-12.2013
Perković, O., Dahmen, K., and Sethna, J. P. (1995). Avalanches, Barkhausen Shriki, O., and Yellin, D. (2016). Optimal information representation and
noise, and plain old criticality. Phys. Rev. Lett.75, 4528–4531. doi: 10.1103/ criticality in an adaptive sensory recurrent neuronal network. PLoS Comput.
PhysRevLett.75.4528 Biol.12:e1004698. doi: 10.1371/journal.pcbi.1004698
Petermann, T., Thiagarajan, T. C., Lebedev, M. A., Nicolelis, M. A. L., Chialvo, Simola, J., Zhigalov, A., Morales-Munoz, I., Palva, J. M., and Palva, S. (2017).
D. R., and Plenz, D. (2009). Spontaneous cortical activity in awake monkeys Critical dynamics of endogenous fluctuations predict cognitive flexibility in the
composed of neuronal avalanches. Proc. Natl. Acad. Sci. U.S.A.106, 15921– Go/NoGo task. Sci. Rep.7:2909. doi: 10.1038/s41598-017-02750-9
15926. doi: 10.1073/pnas.0904089106 Simonsen, I., Hansen, A., and Nes, O. M. (1998). Determination of the Hurst
Pisarenko, V. F., and Sornette, D. (2012). Robust statistical tests of Dragon-Kings exponent by use of wavelet transforms. Phys. Rev. E Stat. Physics, Plasmas,
beyond power law distributions. Eur. Phys. J. Spec. Top.205, 95–115. doi: 10. Fluids, Relat. Interdiscip. Top.58:2779. doi: 10.1103/PhysRevE.58.2779
1140/epjst/e2012-01564-8 Slezin, V. B., Korsakova, E. A., Dytjatkovsky, M. A., Schultz, E. A., Arystova,
Plenz, D., and Niebur, E. (2014). Criticality in Neural Systems. Hoboken, NJ: Wiley, T. A., and Siivola, J. R. (2007). Multifractal analysis as an aid in the
doi: 10.1002/9783527651009 diagnostics of mental disorders. Nord. J. Psychiatry61, 339–342. doi: 10.1080/
Priesemann, V., Valderrama, M., Wibral, M., and Le Van Quyen, M. (2013). 08039480701643175
Neuronal avalanches differ from wakefulness to deep sleep–evidence from Smit, D. J. A., de Geus, E. J. C., van de Nieuwenhuijzen, M. E., van Beijsterveldt,
intracranial depth recordings in humans. PLoS Comput. Biol.9:e1002985. doi: C. E. M., van Baal, G. C. M., Mansvelder, H. D., et al. (2011). Scale-free
10.1371/journal.pcbi.1002985 modulation of resting-state neuronal oscillations reflects prolonged brain
Radulescu, A. R., Rubin, D., Strey, H. H., and Mujica-Parodi, L. R. (2012). maturation in humans. J. Neurosci.31, 13128–13136. doi: 10.1523/JNEUROSCI.
Power spectrum scale invariance identifies prefrontal dysregulation in 1678-11.2011
paranoid schizophrenia. Hum. Brain Mapp.33, 1582–1593. doi: 10.1002/hbm. Sornette, A., and Sornette, D. (1989). Self-organized criticality and earthquakes.
21309 Europhys. Lett.9:197. doi: 10.1209/0295-5075/9/3/002
Ribeiro, T. L., Copelli, M., Caixeta, F., Belchior, H., Chialvo, D. R., Nicolelis, Sornette, D. (2004). Critical Phenomena in Natural Sciences: Chaos. Cham:
M. A. L., et al. (2010). Spike avalanches exhibit universal dynamics across Springer.
the sleep-wake cycle. PLoS One5:e0014129. doi: 10.1371/journal.pone.00 Stam, C. J., and de Bruin, E. A. (2004). Scale-free dynamics of global functional
14129 connectivity in the human brain. Hum. Brain Mapp.22, 97–109. doi: 10.1002/
Roberts, J. A., Boonstra, T. W., and Breakspear, M. (2015). The heavy tail of hbm.20016
the human brain. Curr. Opin. Neurobiol.31, 164–172. doi: 10.1016/j.conb.2014. Stam, C. J., Montez, T., Jones, B. F., Rombouts, S. A. R. B., van der Made, Y.,
10.014 Pijnenburg, Y. A. L., et al. (2005). Disturbed fluctuations of resting state EEG
Roberts, J. A., Iyer, K. K., Finnigan, S., Vanhatalo, S., and Breakspear, synchronization in Alzheimer’s disease. Clin. Neurophysiol.116, 708–715. doi:
M. (2014a). Scale-free bursting in human cortex following hypoxia 10.1016/j.clinph.2004.09.022
at birth. J. Neurosci.34, 6557–6572. doi: 10.1523/jneurosci.4701-13. Stramaglia, S., Pellicoro, M., Angelini, L., Amico, E., Aerts, H., Cortés, J. M., et al.
2014 (2017). Ising model with conserved magnetization on the human connectome:

implications on the relation structure-function in wakefulness and anesthesia. Vanhatalo, S., and Kaila, K. (2006). Development of neonatal EEG activity: from
Chaos.27:047407. doi: 10.1063/1.4978999 phenomenology to physiology. Semin. Fetal Neonatal Med.11, 471–478. doi:
Stumpf, M. P. H., and Porter, M. A. (2012). Critical truths about power laws. 10.1016/j.siny.2006.07.008
Science335, 665–666. doi: 10.1126/science.1216142 Vyšata, O., Procházka, A., Mareš, J., Rusina, R., Pazdera, L., Vališ, M., et al. (2014).
Suckling, J., Wink, A. M., Bernard, F. A., Barnes, A., and Bullmore, E. (2008). Change in the characteristics of EEG color noise in alzheimer’s disease. Clin.
Endogenous multifractal brain dynamics are modulated by age, cholinergic EEG Neurosci.45, 147–151. doi: 10.1177/1550059413491558
blockade and cognitive performance. J. Neurosci. Methods174, 292–300. doi: Wang, J. W. J. L., Lombardi, F., Zhang, X., Anaclet, C., and Ivanov, P. C.
10.1016/j.jneumeth.2008.06.037 (2019). Non-equilibrium critical dynamics of bursts in 2 and δ rhythms as
Surmeli, T., Ertem, A., Eralp, E., and Kos, I. H. (2012). Schizophrenia and the fundamental characteristic of sleep and wake micro-architecture. PLoS Comput.
efficacy of qEEG-guided neurofeedback treatment: a clinical case series. Clin. Biol.15:e1007268. doi: 10.1371/journal.pcbi.1007268
EEG Neurosci.43, 133–144. doi: 10.1177/1550059411429531 Wang, R., Dearing, J. A., Langdon, P. G., Zhang, E., Yang, X., Dakos,
Tagliazucchi, E., Balenzuela, P., Fraiman, D., and Chialvo, D. R. (2012). Criticality V., et al. (2012). Flickering gives early warning signals of a critical
in large-scale brain FMRI dynamics unveiled by a novel point process analysis. transition to a eutrophic lake state. Nature492, 419–422. doi: 10.1038/nature
Front. Physiol.3:15. doi: 10.3389/fphys.2012.00015 11655
Tagliazucchi, E., Carhart-Harris, R., Leech, R., Nutt, D., and Chialvo, D. R. Weiss, B., Clemens, Z., Bodizs, R., and Halasz, P. (2011). Comparison of fractal and
(2014). Enhanced repertoire of brain dynamical states during the psychedelic power spectral EEG features: effects of topography and sleep stages. BRAIN Res.
experience. Hum. Brain Mapp.35, 5442–5456. doi: 10.1002/hbm.22562 Bull.84, 359–375. doi: 10.1016/j.brainresbull.2010.12.005
Tagliazucchi, E., von Wegner, F., Morzelewski, A., Brodbeck, V., Jahnke, K., West, T., Farmer, S., Berthouze, L., Jha, A., Beudel, M., Foltynie, T., et al. (2016).
and Laufs, H. (2013). Breakdown of long-range temporal dependence in The parkinsonian subthalamic network: measures of power, linear, and non-
default mode and attention networks during deep sleep. Proc. Natl. Acad. Sci. linear synchronization and their relationship to L-DOPA treatment and OFF
U.S.A.110, 15419–15424. doi: 10.1073/pnas.1312848110 state motor severity. Front. Hum. Neurosci.10:517. doi: 10.3389/fnhum.2016.
Tanaka, T., Kaneko, T., and Aoyagi, T. (2009). Recurrent infomax generates 00517
cell assemblies, neuronal avalanches, and simple cell-like selectivity. Neural Wikstro, S., Iyer, K. K., Roberts, J. A., Hellstro, L., Pupp, I. H., Ley, D., et al. (2015).
Comput.21, 1038–1067. doi: 10.1162/neco.2008.03-08-727 Cortical burst dynamics predict clinical outcome early in extremely preterm
Thatcher, R. W., North, D. M., and Biver, C. J. (2009). Self-organized criticality infants. Brain138(Pt 8), 2206–2218. doi: 10.1093/awv147
and the development of EEG phase reset. Hum. Brain Mapp.30, 553–574. doi: Wilting, J., and Priesemann, V. (2019). 25 years of criticality in neuroscience
10.1002/hbm.20524 - established results, open controversies, novel concepts. Curr. Opin.
Thatcher, R. W., North, D. M., and Biver, C. J. (2014). LORETA EEG phase reset Neurobiol.58, 105–111. doi: 10.1016/j.conb.2019.08.002
of the default mode network. Front. Hum. Neurosci.8:529. doi: 10.3389/fnhum. Wink, A. M., Bernard, F., Salvador, R., Bullmore, E., and Suckling, J. (2006). Age
2014.00529 and cholinergic effects on hemodynamics and functional coherence of human
Thiery, T., Lajnef, T., Combrisson, E., Dehgan, A., Rainville, P., Mashour, hippocampus. Neurobiol. Aging.27, 1395–1404. doi: 10.1016/j.neurobiolaging.
G. A., et al. (2018). Long-range temporal correlations in the brain distinguish 2005.08.011
conscious wakefulness from induced unconsciousness. Neuroimage179, 30–39. Wink, A. M., Bullmore, E., Barnes, A., Bernard, F., and Suckling, J. (2008).
doi: 10.1016/j.neuroimage.2018.05.069 Monofractal and multifractal dynamics of low frequency endogenous brain
Tinker, J., and Velazquez, J. L. P. (2014). Power law scaling in synchronization of oscillations in functional MRI. Hum. Brain Mapp.29, 791–801. doi: 10.1002/
brain signals depends on cognitive load. Front. Syst. Neurosci.8:73. doi: 10.3389/ hbm.20593
fnsys.2014.00073 Witton, C., Sergeyev, S. V., Turitsyna, E. G., Furlong, P. L., Seri, S., Brookes, M.,
Tkačik, G., Marre, O., Mora, T., Amodei, D., Berry, M. J., and Bialek, W. (2013). et al. (2019). Rogue bioelectrical waves in the brain: the Hurst exponent as a
The simplest maximum entropy model for collective behavior in a neural potential measure for presurgical mapping in epilepsy. J. Neural Eng.16:56019.
network. J. Stat. Mech. Theory Exp. 2013. doi: 10.1088/1742-5468/2013/03/ doi: 10.1088/1741-2552/ab225e
P03011 Worrell, G. A., Cranstoun, S. D., Echauz, J., and Litt, B. (2002). Evidence for
Tolkunov, D., Rubin, D., and Mujica-Parodi, L. R. (2010). Power spectrum scale self-organized criticality in human epileptic hippocampus. Neuroreport13,
invariance quantifies limbic dysregulation in trait anxious adults using fMRI: 2017–2021. doi: 10.1097/00001756-200211150-00005
adapting methods optimized for characterizing autonomic dysregulation to Yan, J., Wang, Y., Ouyang, G., Yu, T., Li, Y., Sik, A., et al. (2016). Analysis of
neural dynamic time series. Neuroimage 50, 72–80. doi: 10.1016/j.neuroimage. electrocorticogram in epilepsy patients in terms of criticality. Nonlinear Dyn.83,
2009.12.021 1909–1917. doi: 10.1007/s11071-015-2455-9
Tomen, N., Ernst, U. A., and Herrmann, J. M. (2019). The Functional Role of Yang, H., Shew, W. L., Roy, R., and Plenz, D. (2012). Maximal variability of
Critical Dynamics in Neural Systems. Berlin: Springer International Publishing. phase synchrony in cortical networks with neuronal avalanches. J. Neurosci.32,
Tomko, G. J., Crapper, D. R., Wilson, K. G., Madison, D. V., Nicoll, R. A., Sauer, 1061–1072. doi: 10.1523/JNEUROSCI.2771-11.2012
T., et al. (2018). The theory of critical phenomena: an introduction to the Zhigalov, A., Kaplan, A., and Palva, J. M. (2016). Modulation of critical brain
renormalization group. Neuron.46:671. doi: 10.1101/107185 dynamics using closed-loop neurofeedback stimulation. Clin. Neurophysiol.127,
Touboul, J., and Destexhe, A. (2010). Can power-law scaling and neuronal 2882–2889. doi: 10.1016/j.clinph.2016.04.028
avalanches arise from stochastic dynamics?PLoS One5:e8982. doi: 10.1371/
journal.pone.0008982 Conflict of Interest: The author declares that the research was conducted in the
Turrigiano, G. G., and Nelson, S. B. (2004). Homeostatic plasticity in the absence of any commercial or financial relationships that could be construed as a
developing nervous system. Nat. Rev. Neurosci.5, 97–107. doi: 10.1038/nrn1327 potential conflict of interest.
Uhlhaas, P. J., Roux, F., Rodriguez, E., Rotarska-Jagiela, A., and Singer, W. (2010).
Neural synchrony and the development of cortical networks. Trends Cogn. Copyright © 2020 Zimmern. This is an open-access article distributed under the
Sci.14, 72–80. doi: 10.1016/j.tics.2009.12.002 terms of the Creative Commons Attribution License (CC BY). The use, distribution
Van De Leemput, I. A., Wichers, M., Cramer, A. O. J., Borsboom, D., Tuerlinckx, or reproduction in other forums is permitted, provided the original author(s) and
F., Kuppens, P., et al. (2014). Critical slowing down as early warning for the the copyright owner(s) are credited and that the original publication in this journal
onset and termination of depression. Proc. Natl. Acad. Sci. U.S.A.111, 87–92. is cited, in accordance with accepted academic practice. No use, distribution or
doi: 10.1073/pnas.1312114110 reproduction is permitted which does not comply with these terms.

Critical Clinical

Uploaded by

Copyright:

Available Formats

Critical Clinical

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Critical Clinical

Uploaded by

Copyright:

Available Formats

REVIEW

published: 26 August 2020

Why Brain Criticality Is Clinically

Frontiers in Neural Circuits | www.frontiersin.org 1 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 2 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 3 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 4 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 5 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 6 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 7 August 2020 | Volume 14 | Article 54

FIGURE 5 | PRISMA-ScR flow chart.

Frontiers in Neural Circuits | www.frontiersin.org 8 August 2020 | Volume 14 | Article 54

neurodevelopment, cognition, psychiatry, and sleep medicine.

CLINICAL APPLICATIONS OF BRAIN

Frontiers in Neural Circuits | www.frontiersin.org 9 August 2020 | Volume 14 | Article 54

TABLE 1 | Summary of anesthesia-related criticality literature.

Study Study population Modality Analysis Main findings

ECoG, electrocorticogram; fMRI, functional magnetic resonance imaging; EEG, electroencephalogram.

Frontiers in Neural Circuits | www.frontiersin.org 10 August 2020 | Volume 14 | Article 54

TABLE 2 | Summary of epilepsy-related criticality literature.

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 11 August 2020 | Volume 14 | Article 54

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 12 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 13 August 2020 | Volume 14 | Article 54

TABLE 3 | Summary of neurodegeneration-related criticality literature.

Study Study Population Modality Analysis Main Findings

Frontiers in Neural Circuits | www.frontiersin.org 14 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 15 August 2020 | Volume 14 | Article 54

TABLE 4 | Summary of neurodevelopment-related criticality literature.

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 16 August 2020 | Volume 14 | Article 54

Study Study population Modality Analysis Main findings

disability. Low burst slope values, moderated by effect of gestational age,

Frontiers in Neural Circuits | www.frontiersin.org 17 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 18 August 2020 | Volume 14 | Article 54

TABLE 5 | Summary of cognition-related criticality literature.

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 19 August 2020 | Volume 14 | Article 54

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 20 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 21 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 22 August 2020 | Volume 14 | Article 54

TABLE 6 | Summary of psychiatry-related criticality literature.

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 23 August 2020 | Volume 14 | Article 54

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 24 August 2020 | Volume 14 | Article 54

TABLE 7 | Summary of sleep-related criticality literature.

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 25 August 2020 | Volume 14 | Article 54

Study Study population Modality Analysis Main findings

Frontiers in Neural Circuits | www.frontiersin.org 26 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 27 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 28 August 2020 | Volume 14 | Article 54

Despite these challenges and controversies, there is also room

Frontiers in Neural Circuits | www.frontiersin.org 29 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 30 August 2020 | Volume 14 | Article 54

Frontiers in Neural Circuits | www.frontiersin.org 31 August 2020 | Volume 14 | Article 54