Quantitative Electroencephalogram (qEEG) as a Natural and Non-Invasive Window into Living Brain and Mind in the Functional Continuum of Healthy and Pathological Conditions

Featured Application

The presented theoretical–conceptual framework has neurobiological/etiological relevance and focuses on dimensionally conceptualized physiological characteristics, mental functions, and neuropsychopathology and, as such, may provide the better clinical diagnostic and prognostic utility of qEEGs.

Abstract

Many practicing clinicians are time-poor and are unaware of the accumulated neuroscience developments. Additionally, given the conservative nature of their field, key insights and findings trickle through into the mainstream clinical zeitgeist rather slowly. Over many decades, clinical, systemic, and cognitive neuroscience have produced a large and diverse body of evidence for the potential utility of brain activity (measured by electroencephalogram—EEG) for neurology and psychiatry. Unfortunately, these data are enormous and essential information often gets buried, leaving many researchers stuck with outdated paradigms. Additionally, the lack of a conceptual and unifying theoretical framework, which can bind diverse facts and relate them in a meaningful way, makes the whole situation even more complex. To contribute to the systematization of essential data (from the authors’ point of view), we present an overview of important findings in the fields of electrophysiology and clinical, systemic, and cognitive neuroscience and provide a general theoretical–conceptual framework that is important for any application of EEG signal analysis in neuropsychopathology. In this context, we intentionally omit detailed descriptions of EEG characteristics associated with neuropsychopathology as irrelevant to this theoretical–conceptual review.

1. Introduction

“The history of EEG studies of mental activity shows that a weak theoretical basis at certain stages can result not only in methodological crises but can also affect empirical data collection and interpretation. An adequate theory can lend strong support to the methodology with “brain-oriented” structuring of psychological tasks and such a theory improves the neurophysiological informative value of the EEG parameters referring to the psychological characteristics of mental processes etc” ([1], p. 384). “It is time to begin the daunting task of relating clinical manifestations of mental disorders to neuroscientific brain dynamics in a comprehensive unifying manner” ([2], p. 942).

In the course of everyday life conditions and within the context of health and disease, people can be evaluated across three different dimensions: behavioral (performance), brain functioning (neuroimaging), and introspection (subjective aspects—psychology). Changes in these different dimensions do not always parallel one another; however, the common denominator for all of them is brain functioning, which affects and reflects the other two dimensions—behavior/performance and psychology/subjectivity [1]. Indeed, the physiologic functioning of the brain underlies emotions, cognition, and behavior; hence, in this context, an objective assessment of brain dysfunction is especially critical for neurology and psychiatry [1,3].

Currently, brain dysfunction is assessed according to manuals such as the Diagnostic Statistical Manual in the USA (DSM) or the International Classification of Disorders in Europe (ICD). However, none of the DSM-/ICD-defined syndromes correlate with any neurobiological phenotypic marker or gene that could have etiological relevance or predict the efficacy of medications [2]. It seems that to arrive at a biological basis for disease categories, brain disorders should be classified in association with impairment of brain systems and diagnosed according to deviations from normality in the corresponding brain activity [3].

Several factors need to be considered when choosing an appropriate measure of brain activity. First, we need a natural and non-invasive ‘window’ into the living brain. This window should give us an ‘online’ view that directly captures the dynamics of brain activity, which reflects multiple interacting operational modules hierarchically organized to allow for complex information processing that (a) characterizes the neurophysiological type (combination of traits; trait refers to the constitutional characteristics that are temporally stable over longer periods of an individual’s life-span; an individual with a certain trait characteristic responds similarly over many situations over the period of several months or years; here, trait corresponds to a temporally stable neuro-psycho-physiological system) and multidimensional structure of the functional state of the brain, (b) has high heritability and is, thus, individually specific, (c) reflects individual neurodevelopment (‘historicism’) and age-related changes, (d) is associated with higher mental and cognitive functions and subjective experience, and (e) reflects or guides neuro- and psychopathology.

Second, from a biophysical perspective, “disease may be regarded not only as a functional or molecular–structural abnormality, as in the classic view, but also (and not by way of contrast) as a disturbance of an entire network of electromagnetic communications. This network is based on long-range interactions between elements […] which oscillate at frequencies which are coherent and specific and thus capable of resonance. This would be a disturbance of internal oscillators and their communications” ([4], pp. 107–108; see also [5]).

Considering these two aspects, it seems that electroencephalogram (EEG) is the best candidate for measuring brain activity compared to other methods of brain signal acquisition, such as the magnetoencephalogram (MEG), functional magnetic resonance imaging (fMRI), and positron emission tomography (PET).

2. Why the Electroencephalogram (EEG)?

MEG, fMRI, and PET are expensive, non-portable, partially invasive, and are usually associated with high stress due to noise, space confinement, and the need to be motionless. Additionally, fMRI and PET scans provide indirect measures of brain activity, with poor temporal resolution. Further, many types of mental activities, brain disorders, and malfunctions of the brain cannot be registered using fMRI since its effect on the level of oxygenated blood is low [5]. In contrast to these neuroimaging techniques, EEG is, at the same time, the cheapest, fastest, and most portable technique that measures neuronal activity directly and non-invasively. EEG does not elicit feelings of claustrophobia, does not require overt cooperative behavior from the person, has a temporal resolution adequate to mental and cognitive processes, and may distinguish between different temporal scales of information processing inherent to mental and cognitive processes.

EEG is a summation of electric voltage fields produced by dendritic and postsynaptic currents of many cortical neurons firing in non-random partial synchrony [6,7,8]. The aggregate of these electric voltage fields can be detected by electrodes on the scalp. The brainstem and thalamus serve as subcortical generators to synchronize populations of neocortical neurons in both normal and abnormal conditions, thus influencing the EEG. It seems that the activity of subcortical structures can be ‘visible’ in the EEG either indirectly through their effects on cortical activity or—in contrast to popular belief—more directly via deep sources. See explanations given in [9] (p. 7): “While local field potentials indeed fall off rapidly within the brain, far less attenuation is observed when recording across skull and scalp. The reason is that the lower conductivity of the skull (compared to the brain and scalp) attenuates superficial sources more strongly than deep ones, thus acting like a spatial low-pass filter. This property causes strong blurring and attenuation of the focal superficial fields but has less of an effect on the more diffuse (“low spatial frequency”) fields from deeper sources […] Recent evidence suggests that a considerably larger range of brain structures, layers, and cell types than previously thought can contribute to spontaneous EEG phenomena”.

A quantitative electroencephalogram (qEEG) is a mathematically and algorithmically processed digitally recorded EEG that extracts information invisible to ‘naked’ eye inspections of the signal. For the rest of the paper, we will mostly refer to qEEGs as the majority of studies are performed using qEEGs.

After decades of studies, it is becoming clear that the qEEG is closely related to brain dynamics, with millisecond temporal resolution, functional properties, and global states of brain functioning, information processing, and cognitive activity [10,11,12,13,14,15,16,17]. The interaction of large populations of neurons gives rise to rhythmic electrical events in the brain, which can be observed at several temporal scales—qEEG oscillations. They are the basis of many different behavioral patterns and sensory mechanisms (for a review, see [18]). Indeed, a large body of evidence [19,20,21,22,23,24,25,26,27] has demonstrated that qEEG oscillations constitute a mechanism by which the brain can actively regulate changes in a state in selected neuronal networks to cause qualitative transitions between modes of information processing [20]. Thus, different qEEG oscillatory patterns are indicative of different information-processing states.

The qEEG has a number of important features that make it especially useful in clinical practice. In the following sections, we present a brief review of these features.

2.1. qEEG Historicism

An adult human qEEG is characterized by ‘historicism’—the information about primate phylogeny, pre- and post-natal maturation (individual development), and early life events (utero characteristics and early life stress).

Indeed, phylogenetically (phylogenesis—the evolutionary development and diversification of a species or group of organisms), the proportion of power of qEEG oscillations changes as a function of primate phylogeny [5,22,28]. Likewise, ontogenetically (ontogenesis—physical and psychological development of an individual organism from inception to maturity), qEEGs undergo significant transformation as a function of pre-natal (in utero) development (maternal stress exposure, anxiety, and depression during pregnancy are considered in utero adverse experiences and have been associated with future health problems [29,30,31,32]; this is so because intrauterine life events have a much greater impact on epigenetic profiles than stressful exposures during adult life due to heightened brain plasticity that is adversely affected by exposure to environmental insults [33]) [34] as well as a function of post-natal maturation (maturation refers to the timely appearance or unfolding of brain structures, events, and processes that are the result of the interaction between genes and the environment; brain maturation can be delayed, equal, or accelerated when compared to chronological age) [35,36,37,38]. It seems that ontogenetic differences mirror those of phylogenetic differences in the cause of brain development, where there is a gradual increase in qEEG complexity and change in the qEEG oscillations’ composition and proportions [17,39,40,41]. Why is this relevant? The qEEG has been found to have a high prognostic value for identifying the functional level of ‘brain maturity’ [42,43]. The knowledge of typical qEEG oscillatory patterns for a given phylogenesis/ontogenesis stage gives one the ability to assess the level of qEEG maturation or regression, which often accompanies the development of neuropsychopathology [44]. For example, a person with an immature qEEG is more easily swayed by external influences and has a lower threshold for aggressive and/or antisocial behaviors [45].

Additionally, early life stress (ELS) has been associated with abnormalities in the qEEG of adults and is also paralleled by a range of adverse outcomes in adults, such as personality dimensions, increased vulnerability to substance abuse, depression, anxiety, psychosis, and post-traumatic stress disorder (PTSD) [46,47,48]. Indeed, ELS such as protein energy malnutrition in the first year of life, extreme social and cognitive deprivation as a result of institutional care, physical or emotional neglect, and low socioeconomic status are all associated with abnormal qEEG characteristics on one hand and with developmental lag or deviation, persistent specific cognitive and behavioral deficits in adulthood, and accelerated cognitive decline [49,50,51,52,53,54] on another hand. Further, childhood traumas (including childhood sexual abuse) are also associated with adult qEEG deviations in parallel with cognitive dysfunction [55,56]. It seems that changes in catecholamine levels following a traumatic event can impede brain regional development, which, in turn, can compromise later cognitive functioning and emotional regulation and leave a person susceptible to stress later on in life.

Additionally, traumatic brain injury may also be reflected in qEEG deviations that correspond to complaints of cognitive symptoms that can persist anywhere from 11 [57], 22 [58], or even 27 [59] years post-injury, characterizing persistent post-concussive syndrome.

This brief overview suggests that the qEEG contains information that is a historical consequence of individual development, ELS, and significant life events. However, in order to adequately assess qEEG variability associated with pathology, within-subject stability over EEG recordings within an EEG session, test–retest reliability over time, and intra-subject specificity (i.e., the extent to which a qEEG pattern is uniquely associated with a given person) and specificity for different conditions need to be established.

2.2. qEEG Stability, Reliability, and Specificity

Studies have reportedly demonstrated that the majority of qEEG characteristics have high (up to 90%) within-subject stability (internal consistency measured by Cronbach’s alpha) within an EEG recording session, high (up to 90%) reproducibility (test–retest reliability) over a period of hours, weeks, months and years, and high (up to 99%) intra-subject specificity, meaning that qEEG can accurately identify subjects from a large group [60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78].

These results suggest that qEEG characteristics possess trait-like qualities (stability over time). In this context, intrinsic properties of brain activity measured by resting qEEG constitute a neural counterpart of personality traits (Section 4.2) and can be regarded as the statistical neuro-signature of a person. Such high stability, reliability, and specificity of qEEG characteristics suggest that genetic factors have a strong influence on qEEG variation.

2.3. qEEG Heritability

A large body of studies have suggested that qEEG characteristics and their variability are largely determined by genetics and, thus, are highly heritable (up to 90%) [79,80,81,82,83,84,85,86,87,88,89,90,91,92,93]. Additionally, it was demonstrated that the correlations for qEEG characteristics between family groups (each consisting of a biologically related father, mother, and two children) were greater than those obtained from the non-family groups (each consisting of biologically unrelated subjects) [94] (see also [89]).

Smit et al. [92] proposed several common genetic sources for EEG: (a) skull and scalp thickness may affect the conductive properties of the tissues surrounding the cortex, (b) genetic influence on cerebral rhythm generators such as the central ‘pacemaker’ in the septum for hippocampal activity or the thalamocortical and corticocortical generators of cortical rhythmicity, (c) genes directly involved in the bioelectric basis of the EEG signal itself: for example, genes influencing the number of pyramidal cells, the number of dendritic connections, or their orientation with respect to the scalp may directly influence the mass dendritic tree depolarization of pyramidal cells in the cortex that underlies the EEG. Begleiter and Porjesz [95] added another factor: regulatory genes that control the neurochemical processes of the brain and, therefore, influence neural function.

Besides high qEEG heritability, genetic loci underlying the functional organization of human neuroelectric activity and their associated conditions/behavior have also been identified. Below is a short overview of qEEG oscillations, the related genes, and the associated pathological conditions:

(a)

qEEG beta oscillations (beta rhythm is electromagnetic oscillations in the frequency range of brain activity above 13 Hz)

Winterer et al. [96] reported that three exonic variants of the gene encoding the human gamma-amino butyric acid (GABA)_B receptor on chromosome 6 modify the cortical synchronization measured as scalp-recorded qEEG coherence. Another genetic study indicated the importance of GABA_A receptor genes in the modulation of qEEG beta oscillations in the human brain: Porjesz et al. [97] found a significant genetic linkage between the beta frequency of the human qEEG and a cluster of GABA_A receptor genes on chromosome 4p. Additionally, this same GABA_A receptor gene was found to be associated with a DSM-IV diagnosis of alcohol dependence [98].

(b)

qEEG alpha oscillations (alpha rhythm is electromagnetic oscillations in the frequency range of 8–13 Hz, arising from the synchronous and coherent electrical activity of neurons in the human brain)

Low voltage qEEG alpha oscillations have also been reported to be linked to (a) the GABAergic system, as an association has been found between the exon 7 variant of the GABA_B receptor gene and alpha voltage [99], (b) a serotonin receptor gene (HTR3B), associated with alcoholism and antisocial behavior [100], and (c) a corticotrophin-releasing binding hormone (CRH-BP) [101,102], associated with depression, anxiety, and alcoholism. Low voltage alpha in females has also been reported to be associated with a genetic variant that leads to low activity of the enzyme that metabolizes dopamine and norepinephrine, catechol-o-methyltransferase (COMT) [103]. Additionally, low voltage alpha has been associated with a subtype of alcohol dependence with anxiety disorders [104,105] and with the brain-derived neurotrophic factor (BDNF) Val66Met polymorphism in depression [106]. High voltage qEEG alpha oscillations are heritable in a simple autosomal dominance manner [79]. The alpha peak frequency (APF) has been associated with the COMT gene, with the Val/Val genotype being marked by a 1.4 Hz slower APF compared to the Met/Met group [107].

(c)

qEEG theta oscillations (theta rhythm is electromagnetic oscillations in the frequency range of brain activity between 4 and 7.5 Hz)

There is evidence [108] that single nucleotide polymorphisms located in brain-expressed long intergenic non-coding RNAs (lincRNAs) on chromosome 18q23 are associated with posterior interhemispheric theta EEG coherence. These same variants are also associated with alcohol use behavior and posterior corpus callosum volume. Further, the Val¹⁵⁸Met polymorphism of the COMT gene is associated with low-frequency oscillation abnormalities in schizophrenia patients [109].

This short overview suggests that there are common genetic links between qEEG oscillation characteristics and specific health conditions. It seems that genetically influenced features of the intrinsic oscillatory activity are related to the structures and functions of the corresponding neural generators and that different features of qEEGs may predict individual differences in brain function and structures.

2.4. qEEG and Structural Integrity of the Brain

Indeed, numerous studies have demonstrated that qEEGs reflect the brain’s structural characteristics (or ‘hardware’), such as the number of connections between neurons, white matter density, axonal diameter, degree of myelination and white matter integrity, as well as the integrity of corticocortical and thalamocortical circuits, hippocampal volume, the number of active synapses in thalamic nuclei, and the number of potential neural pathways [7,8,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126]. For example, reduced EEG amplitude is believed to be partially due to a reduced number of synaptic generators and/or reduced integrity of the protein/lipid membranes of neurons [127,128].

2.5. qEEG and Functional Integrity of the Brain

Decades of studies have demonstrated that brain functional characteristics (or ‘software’), such as memory performance, attention and processing speed, emotional regulation, individual capacity for information processing, cognitive preparedness, and others, including functional states of the brain, are readily reflected in qEEGs at all ages in both healthy individuals and individuals with neurological or psychiatric conditions [14,21,22,24,115,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147] (for a review, see [18]). This is so because qEEG oscillatory activity is generated by synchronous neural populations that mirror the firing rate of their constituent neurons [148]: for example, during arousal, task execution, and/or a behavioral act, the underlying neuronal populations will increase spiking with respect to baseline. These increased firing rates will engage non-linear feedback loops, effectively changing the system’s response function and the specifics of its emergent oscillations. In contrast, during rest and states of quietness, the spiking activity decreases, which is also reflected by a decrease in oscillatory activity [148]. Further, qEEG oscillations are able to temporally coordinate and control neuronal firing and are proposed to be a basic principle of information processing in the human brain [149,150].

Considering that different qEEG oscillations reflect functionally different components of information processing acting on various temporal scales [140,141], it is possible to map qEEG oscillations onto specific mental and/or behavioral states [151]. qEEG oscillations from the same frequency band may express different functions depending on the conditions they are involved in [152]. This seems biologically plausible: qEEG oscillatory functional diversity creates a rich repertoire of brain activity that can meet the complex computational and communicational demands of the brain during healthy and pathological conditions.

In this context, qEEG measures can provide independent evidence of variations in alertness, attentiveness, memory, emotional regulation, or mental effort. Incorporating them into tests of cognitive function might lead to more sensitive and less ambiguous clinical assessment tools [153,154].

Since information-processing modes depend on the functional integrity of the brain, which, in turn, depends on the orchestrated oscillatory activity of neuronal pools (reflected in the characteristic qEEG rhythms); functional coupling between qEEG oscillations, cognitive functions, and vegetative processes is important.

2.6. qEEG and Vegetative Status/Autonomic Nervous Systems (ANS)

Several studies have demonstrated the association between qEEGs and ANS [155,156,157,158]. It seems that the brainstem mediates a functional coupling between the ANS and the central nervous system (CNS) assessed by qEEG [159,160,161]. A theoretical concept of the integration between the ANS and CNS was presented by Jennings and Coles [162]. The coordination and communication in and between the autonomic vegetative systems and the brain occur with tuned frequencies in the range of qEEG oscillations, suggesting the existence of resonant links in the brain with all organs of the body (for a review and discussion, see [5]; see also [163]). Basar [5] suggested that such mutual resonances form a coordinated dynamic system that maintains survival functions such as blood pressure, respiratory rhythms, cardiac pacemakers, and body temperature (see also [155,156,159,160,161,162]).

Since the dynamics of the physiologic variables (autonomic system) and the dynamics of brain activity depend on each other, it is reasonable to hypothesize that reduced variability in the activity of the neural networks should cause a concurrent decrease in the variability of autonomic physiologic functions. Indeed, it was demonstrated that a widespread brain injury that causes a derangement in neural networks leads to a reduced complexity of qEEG (measured by entropy, the dynamic repertoire of the probable qEEG states, and operational architectonics) [164,165,166] and reduced heart rate variability [167] in unresponsive patients compared to healthy subjects.

It seems that decreased qEEG variability is coupled with a decrease in the variability of other physiologic variables (autonomic system), which results in reduced physiological adaptability. In turn, reduced physiological adaptability can contribute to stress and weakened immunity, which may further impact the qEEG pattern, creating a downward spiral.

2.7. qEEG, Stress, and Immunity

There is a strong link between qEEG oscillatory patterns and stress regulatory systems: the hypothalamic–pituitary–adrenocortical (HPA) axis and the sympathetic–adrenomedullary axis [168]. For example, qEEGs recorded from stressed students before an exam revealed a correlation between greater right hemisphere (RH) activation and higher cortisol levels [169]. This is supported by the following facts: (a) the administration of cortisol to healthy participants has been shown to increase RH frontal activation [170], and (b) greater right-sided activation (measured by resting qEEG) is associated with higher levels of basal cortisol compared to their left-activated counterparts [171]. Cortisol also seems to reduce neural interactions between different areas of the brain. Indeed, an inverse relationship between basal cortisol levels and neural interaction between the frontal and parietal cortex has been demonstrated using qEEG connectivity analysis [172].

Considering the link between qEEG oscillatory patterns and stress regulatory systems, it is not surprising that the association between several factors of the immune system and qEEG activity has also been reported [173,174,175,176,177] (for a recent meta-analysis, see [178]). For example, higher levels of right-prefrontal qEEG activation (a) reliably predicted poorer immune response [176] and (b) are characterized by lower levels of natural killer cell activity [179]. These data support the hypothesis that individuals characterized by a more negative affective style have a weaker immune response and, therefore, may be at greater risk for illness than those with a more positive affective style. Additionally, RH activation is associated with hyprecortisolemia, which contributes to the deterioration of immune system functioning and puts depressed patients at a greater risk of developing other illnesses, accounting for depression’s high comorbidity with other diseases [180].

2.8. qEEG and Cerebral Haemodynamics and Metabolism

Studies have suggested that different qEEG characteristics are related to cerebral hemodynamics and metabolism [181,182,183,184,185,186,187,188,189,190]. Cerebral cortex metabolism disturbance is associated with and may be responsible for cortical neural synchronization anomalies that may manifest as abnormal qEEG oscillations [191]. Additionally, changes in the characteristics of qEEG oscillations (amplitude, power, frequency) are proportional to cerebrovascular damage (CVD) [119]. The qEEG has been shown to be a reliable marker of the decline in neuronal integrity associated with a decline in blood flow [192,193,194,195,196,197,198]. Additionally, studies show a sensitivity greater than 80%, false-positive rates below 5–10%, and correlations of 70% between qEEG and blood flow in ischemic and non-ischemic regions, thus suggesting that the qEEG can reliably detect focal features that can be quite abnormal even if the computer tomography (CT) or MRI scans are still normal (dysfunction without infarction) [199]. Similarly, in patients with subarachnoid hemorrhage, only qEEG could differentiate patients with and without cerebral infarction and not doppler/color-coded duplex sonography [200]. Further, recent meta-analyses have shown that qEEG has prognostic potential in predicting patient independence and stroke severity beyond that afforded by standard clinical assessments [201] (see also [202,203]). Indeed, qEEG changes precede that of multimodal monitoring or confirmation of infarction on CT [204].

Cerebral hemodynamics and metabolism are regulated by a complex interaction between different homeostatic mechanisms where neurotransmitters play a significant role.

2.9. qEEG and Neurotransmitters

Several studies have suggested a relation between different qEEG oscillations and neuromodulator balance [5,205]. This is because peculiarities of qEEG features result from the interaction of numerous resonance loops within the cortex and between the cortex and subcortical structures, and these interactions are significantly influenced by neurotransmitter concentrations in the brain [206]. Indeed, the levels of activity of different neurotransmitter systems (acetylcholinergic (ACh-ergic), noradrenalinergic (NA-ergic), dopaminergic (DA-ergic), serotonergic (ST-ergic), and GABA-ergic), as well as the patterns of their interaction, are important drivers of qEEG oscillations. For example, the activation of the NA-ergic system is associated with the desynchronization of qEEGs during behavioral excitation [207] and an increase in high-frequency qEEG oscillations [208]. It is also believed that the increased activity of the DA-ergic cerebral systems results in shifts of the frequencies of qEEG oscillations toward higher ranges and facilitates the reaction of desynchronization [209]. Additionally, posterior vs. anterior distribution of qEEG theta oscillations is informative on DA levels [210]. Low ST levels result in a higher power of low-frequency qEEG components [206]; conversely, high ST levels result in the decreased power of low-frequency qEEG components and the higher power of high-frequency qEEG components [211]. Higher relative levels of ACh promote qEEG alpha oscillations, whereas an increased tone of inhibitory monoamine receptors is associated with qEEG delta oscillations (delta rhythm is electromagnetic oscillations in the frequency range of brain activity between 1.5 and 3.5 Hz) [205]. It seems that for each qEEG oscillatory pattern, there is a correlated neurotransmitter mix [212].

Deficiencies or excesses of any of the neurotransmitters will produce a marked departure from homeostatically regulated normative qEEG oscillatory patterns and may contribute to neuro–psycho pathophysiology [199,205]. Indeed, a large body of data suggests that it is possible to unravel distinctive abnormal qEEG oscillatory profiles in terms of specific neurochemical imbalances in particular brain regions [213].

2.10. qEEG and Neuropsychopathology

The literature indicates that there is a greater proportion of abnormal EEGs in individuals with psychopathology: (a) up to 68% of qEEGs in psychiatric patients display evidence of pathophysiology, and these results have additional utility beyond simply ruling out ‘organic brain lesions’ [214,215]; (b) up to 73% of nonepileptic adults have qEEG epileptiform discharges (EDs) [216] that are attributable to underlying brain abnormalities (traumatic, vascular, tumor, metabolic), medications, and psychiatric disorders (see, for example, [217]); (c) the mean prevalence of interictal qEEG abnormalities in psychogenic nonepileptic seizures is estimated to be 26% [218,219,220,221,222,223,224,225,226]; (d) up to 30% of panic attack patients have demonstrable qEEG abnormalities, especially in atypical presentations of panic attacks, and the incidence of abnormal qEEG findings in mood disorders reaches 40% [227]; (e) up to 78% of antisocial and criminal populations have underlying qEEG abnormalities [228] that are more prevalent in subjects with violent crimes, repeated violence, and motiveless crimes; (f) up to 76% of children with reading disabilities but without severe disorders of behavior have EEG abnormalities [229], and (g) 69% of youngsters with behavior disorders with a predominance of aggressiveness have EEG deviations [230]. Additionally, there is evidence that abnormal EEGs are associated with the following clinical conditions: negative histories (13%), severe head injury or neuropsychiatric disorder (46%), psychopathic personality (88%), and family history of seizures (62%) [231].

Basic mechanisms of cerebral rhythmic activities in norm and pathology are described in detail in Steriade et al. [212]. This emphasizes that the presence of qEEG abnormalities should be inferred as ‘electrographic markers’ of underlying brain dysfunction and is suggestive of the potential usefulness of qEEGs in clinical practice.

Indeed, more recent research shows that certain neuropsychopathologies, such as attention deficit hyperactivity disorder (ADHD), specific learning disabilities, schizophrenia, obsessive–compulsive disorder (OCD), borderline personality disorder (BPD), depression, suicidal ideation, anxiety disorders, traumatic brain injury (TBI), mild cognitive impairment (MCI), Alzheimer’s disease (AD), and other disorders are associated with specific qEEG patterns and that these spontaneous electric potentials provide reliable markers of brain function and dysfunction [56,152,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246] (for reviews, see [199,213,247]).

Given that patients with different disorders display abnormal and distinct qEEG-profiles, it is not surprising that they can be differentially classified utilizing qEEG-variables [248]. For example, qEEG utility in discrimination/differentiation between affective disorders and schizophrenia [249], between Alzheimer and non-Alzheimer dementias [199,250], between sub-types of dementia [251], between depression and dementia [199], between schizophrenia and unipolar and bipolar depression [199], and between panic disorder and depression [199] have been demonstrated. For sensitivity and specificity values of qEEG-based detection/discrimination of patients with specific disorders, see

Table 1. qEEG-based detection/discrimination of patients with specific dysfunction/disorder.

Dysfunction/Disorder	Sensitivity	Specificity	References
Cerebrovascular disease	>80%		[199]
ADHD/ADD vs. normal children	90%	94%	[237]
ADD vs. DLD	97%	84.2%	[252]
LD vs. normal children	72%	80%	[253]
Depression	72–93%	75–88%	[254]
Panic disorder	71%	84%	[255]
Dementia	91.9%	92.2%	[256]
AD	71–81%		[257,258]
Declining to MCI	95%	94.1%	[259]
Converting to AD	96.3%	94.1%	[259]
Alcohol and drug abuse predicting relapse	61%	85%	[260]
mTBI vs. sTBI	95.5%	97.4%	[238,261]

ADHD/ADD = attention deficit disorders with or without hyperactivity; DLD = developmental learning disorders; LD = learning disorders; AD = Alzheimer’s disease; MCI = minimal cognitive impairment; m(s) TBI = mild(severe) traumatic brain injury.

.

It is argued that the levels of specificity found in qEEG studies are often higher than those found in routinely used clinical tests, such as mammograms, cervical screenings, and brain scans such as CT or single photon emission computed tomography (SPECT) [199,262,263].

Even within the same disorder, qEEGs may be beneficial in identifying the cause of the abnormal behavior. For example, Kropotov distinguishes five reasons for the neurophysiology of ADHD, stating that “[…] mentioned dysfunctions are associated with specific patterns in spontaneous and evoked electrical potentials, recorded from the head by multiple surface electrodes” ([264] (p. 74; see also [3,265]).

Additionally, qEEGs can play a unique role when it comes to dealing with ambiguous or edge cases in clinical practice. It may help to identify/differentiate:

• Electrical changes that precede the clinical onset of a seizure by tens of seconds to minutes—the early detection of a seizure. It has been shown that patients go through a preictal transition for approximately 0.5 to 1 h before a seizure occurs [266]. On average, the prediction rate is ~81% and has an average warning time of 63 min [267];
• Whether a given seizure is epileptic or nonepileptic in origin: For example, there are groups of disorders that produce symptoms similar to an epileptic seizure: (a) cardiac arrhythmias causing syncope, episodes caused by cerebrovascular disease, movement disorders, and unusual manifestations of sleep disorders; (b) events of psychiatric origin (often referred to as psychogenic nonepileptic seizures (PNES)) [268];
• Subclinical seizures: Some seizures recorded during prolonged EEG monitoring may be asymptomatic or ‘subclinical’;
• Whether the cognitive impairments and behavioral problems in question are due to emotional, psychological, or social factors or because of brain dysfunctions or sensory deficits with quantitatively demonstrable abnormalities in brain electrical activity;
• Whether the hyperactive sensation-seeking behavior (typical for ADHD and mania) is due to hypervigilance or vigilance autostabilization behavior, which is a compensatory behavioral pattern to counter regulate a hypovigilance state, and whether withdrawal behavior (typical for depression) is due to hypovigilance or the result of a compensatory behavioral pattern that counter-regulates hypervigilance [269,270];
• Between a degenerative disorder such as AD and pseudodementia due to psychiatric illness [271];
• Between normal and abnormal maturational patterns, such as brain maturation lag (characterized by a pattern of qEEG that is typical for younger age) and brain maturational deviation (characterized by a pattern of qEEG that is not normal at any age) [253];
• Between presence or absence of consciousness in minimally and unresponsive patients [166,272,273,274,275].

In this context, different spatial–temporal qEEG patterns may reflect different underlying mechanisms/functions/symptoms; this hints at the existence of several clinical sub-types within a given diagnostic group that are not recognized by the current diagnostic systems [276].

Distinct aspects of pathophysiologic mechanisms may be elucidated depending on which qEEG oscillations or their combinations are altered in the qEEG oscillatory pattern of any given neuropsychopathology. It seems that neuropsychopathology manifests through the considerable reorganization of the composition of qEEG oscillations and their ratios over a broad frequency range of 0.5–30 Hz, which constitutes the dynamic repertoires of qEEG states. These qEEG oscillations are ‘mixed’ or superimposed in proportions that depend on the specific neuromediators and neural circuit disturbance and also depend on the presence of various symptoms and affects. Spatial analysis has revealed that different cortical areas are characterized by varying numbers of qEEG oscillations, with a statistically significant difference in their relative presence and communication within the qEEG oscillatory pattern [152,244].

One aspect that often goes unnoticed by clinicians but is nearly always affected by neuropsychopathology is the experiential Selfhood.

2.11. qEEG and Experiential Selfhood

Indeed, in various neuropsychopathological conditions, self-consciousness alterations dominate the patient’s phenomenological experiences and have either a long-term or permanent presence [277,278,279]. Even though experiential Selfhood (also referred to as self-consciousness or self-awareness) is a multi-layered concept that is often conceptualized in different ways by various disciplines [280], the currently emerging consensus is that self-referential processing constitutes the core of Selfhood [281,282]. Empirical evidence from neuroscience [281,283,284,285,286] indicates that such self-referential processing is instantiated by a specific self-referential network (SRN) within the brain, sometimes also referred to as the default mode network (DMN) [283,284,285,286,287]. Further, it has been documented that specific qEEG oscillations have a significant positive correlation with the SRN [288,289,290,291,292].

Recently, a three-dimensional neurophysiological model of the complex experiential Selfhood (which is based on the qEEG analysis) was proposed [286,293,294] (for a detailed description, see [295]). This triad model of Selfhood considers the neurophysiological evidence that three major spatially separate yet functionally interacting brain subnets constitute the SRN and account for the phenomenological distinctions between three major aspects of Selfhood, namely, (i) first-person agency (conceptualized as the ‘witnessing observer’ or simply the ‘Self’), (ii) embodiment (conceptualized as ‘representational–emotional agency’ or simply ‘Me’), and (iii) reflection/narration (conceptualized as ‘reflective agency’ or simply ‘I’ ), all of which commensurate with one another [296] and, together, form a unified sense of Selfhood [286] (see also [297]). Each aspect of the triad can be enhanced or weakened depending on the current physiologic and mental state [286,298], voluntary training [293,294], and neuropsychopathology [299,300,301]. Since aspects of Selfhood rarely fall under the purview of clinical practice, we present below a few examples of the potential application of qEEGs in the assessment of experiential Selfhood for different neuropsychopathologies.

For example, in its ‘pure’ unmedicated form, major depressive disorder is associated with functional enhancement (measured by qEEG) of all three aspects (‘Self’, ‘Me’, and ‘I’) [300], thus reflecting the well-documented excessive self-focus, increased rumination, and increased embodiment in patients with depression [302,303,304,305,306,307]. One could speculate that “these three components of complex Selfhood (indexed by (qEEG)) synergize with one another in a maladaptive loop and, over time, become habitual, leading to a vicious circle that maintains a disordered affective state that clinically manifests as depression” [300] (p. 34). It has been further proposed that the ‘Self’ plays a chief role here as it organizes, represents, and appraises the salience of interoceptive/emotional/bodily information presented by ‘Me’ and the narrative and semantic-conceptual information presented by ‘I’ [300].

PTSD is characterized by rather different Self–Me–I dynamics [301]. Increased activity (measured by qEEG) of the ‘Self’ aspect was found to be significantly associated with the increased vigilance of PTSD sufferers to their surroundings, with a concurrent shift of their first-person perspective from the current moment in time to the moment of the traumatic event (criterion E, according to DSM-5 [308]). We have speculated [301] that such constant hypervigilance coupled with profound emotional arousal leads to sensory overload and further exacerbates alienation of the Self in such patients [309]. Indeed, the increased activity (measured by qEEG) of ‘Me’ was found to be significantly linked to enhanced emotional, sensory, and bodily states in PTSD sufferers (criterion D, according to DSM-5), such as fear, stress, frozenness, shivering, shaking, trembling, palpitations, and sweating [310,311,312]. These feelings and memories are usually reported as intrusive and unwanted (criterion B, according to DSM-5). Additionally, it was observed that the activity (measured by qEEG) of ‘I’ decreased and that this decrease was associated with a distinct lack of linguistic/contextual information and narrative to accompany the traumatic event (criterion C, according to DSM-5), which is a well-documented phenomenon in PTSD patients [312,313].

A six-year longitudinal analysis of a single patient’s recovery of self-consciousness (from a minimally conscious state until full self-consciousness) after a severe traumatic brain injury has revealed that the recovery of first-person agency (or ‘Self’), representational–emotional agency (or ‘Me’), and reflective agency (or ‘I’) was paralleled by restoration of functional integrity (measured by qEEG) in the three subnets of the SRN [299]. Of note, the recovery dynamic in the Self–Me–I aspects (and corresponding qEEG metrics) was not linear but followed a unique trajectory for every aspect (some recovered more quickly, while others lagged) and was tightly paralleled by (and significantly correlated with) findings from clinical exams and tests [299,314].

Further, converging evidence for a breakdown of qEEG integrity within the SRN in non- and minimally communicative patients with severe brain injuries was found, and this breakdown was proportional to the degree of expression of clinical self-consciousness [287]. More specifically, it was demonstrated that the strength of qEEG integrity within the SRN was smallest or even absent in patients in a vegetative state (VS), intermediate in patients in a minimally conscious state (MCS), and highest in healthy, fully self-conscious subjects. Curiously the strongest decrease in strength of qEEG integrity as a function of loss of self-consciousness was found in the ‘Self’ aspect compared to the ‘Me’ and ‘I’ SRN modules. The central role of ‘Self’ was also found for the prediction of self-consciousness recovery: those VS patients who later recovered stable minimal or full self-consciousness in the course of the disease (up to six years post-injury) showed stronger ‘Self’ functional integrity (measured by qEEG) in the early stage (three months post-injury) compared to those patients who continued to stay in the persistent VS [315].

This brings us to another reason for the clinical and ethical importance of qEEG utility in the assessment of the neurophysiological and neurophenomenological status of unresponsive patients.

2.12. qEEG and Disorders of Consciousness

A vegetative state (VS), recently re-termed as ‘unresponsive wakefulness syndrome’ (UWS) [316], and MCS belong to the so-called disorders of consciousness or DoCs [317]. While, by convention, VS/UWS patients are unresponsive to their external and internal environments and are thus unconscious [318], patients in MCS show some level of overt awareness and fluctuating ability to follow commands non-reflexively [319] (see also [320]).

The factual simplicity of the qEEG assessment, its portability and adaptability for longitudinal protocols, and its relatively low cost have opened up a wide area of qEEG investigations in the recent decade—these assessments aim to study the pathophysiology of DoCs as well as look for prognostication markers for the recovery of consciousness in DoC patients [321,322]. Already, the simple description of standard EEGs (guided by accurate qualitative scales) has shown a robust correlation of such patterns with both the level of consciousness impairment (VS/UWS or MCS) and the degree of short-term consciousness recovery [323]. These studies reveal that the overall electrical activity of the brain is differentially impaired in patients that fall under different DoCs and that it may be related to the degree of recovery, as follows from the group-analyses [321].

The implementation of more complex numerical computations of the EEG signal—qEEG analyses—has contributed in a much more nuanced way to the evaluation of DoC patients [322], leading to a better understanding of the neural constituents of consciousness’ impairment [166]. For example, studies on qEEG oscillations have demonstrated that patients in VS/UWS have a considerably reduced repertoire of local qEEG oscillations compared to those in MCS or a fully conscious state [272]. Additionally, unawareness in patients with VS/UWS was associated with an altered composition of qEEG oscillations and their proportions compared with a full consciousness state [272,275]. These results confirmed previous observations that loss of consciousness is associated with altered oscillatory contents of the qEEG [324,325,326].

In agreement with these findings, it has been proposed that the degree of reduction in the dynamic correlates of neuronal networks’ complexity measured by the qEEG may be useful for distinguishing patients with different levels of consciousness impairment (VS/UWS vs. MCS) or even as a prognostic measure [165,275,321,326,327,328]. Indeed, evaluation of qEEG spatial–temporal patterns (which reflect functionally connected neuronal assemblies and their dynamics over time) [166,327,329,330] in DoC patients demonstrated that neuronal assemblies become considerably smaller, with shortened life-spans, and they became highly unstable and functionally disconnected (desynchronized) in patients in VS/UWS [166]. In contrast, fluctuating (minimal) awareness in patients who are in MCS is paralleled by partial restoration of qEEG functional integrity, whose parameters approach those of the levels found in healthy, fully conscious participants [166]. These studies lead to the conclusion that consciousness is likely to vanish in the presence of many very small, extremely short-lived, and highly unstable neuronal assemblies that perform their operations completely independently of one another (functional disconnection) and, thus, are not capable of supporting any coherent content to be experienced subjectively. Importantly, it has been documented that the observed impairment in the brain’s functional integrity in DoC patients is independent of brain damage etiology and, thus, reflects functional (and potentially reversible) damage, as opposed to irreversible structural neuronal loss [273]. As a whole, these findings are in keeping with a recent study [331], where it was shown that, in contrast to MCS, the VS/UWS brain is characterized by small, disconnected networks that do not contribute to higher integrative processes [332].

Another factor that may complicate diagnosis and affect both healthy and diseased individuals is aging.

2.13. qEEG and Aging

Since age-related processes affect both the structural and functional integrity of the brain, it is reasonable to suggest that qEEGs possess age-dependent changes that are both pathology-independent (healthy aging) and pathology-dependent (pathological aging). Indeed, many studies have demonstrated that the aging process is reflected in qEEG changes [63,68,144,333,334,335,336,337,338,339,340,341,342,343,344,345] and is associated with age-related conditions such as cognitive decline, Alzheimer’s disease, mild cognitive impairment, vascular dementia or other dementias, multiple sclerosis, and cerebral tumors [112,116,117,119,124,185,187,346,347,348].

Aging, as is well known, eventually results in death; and death is no longer understood to be an all-or-nothing state but rather a process, the aspects of which may be captured by qEEG.

2.14. qEEG and Death

Death is often a tragic and somewhat baffling finale of a person’s life. Since the person is unresponsive near or during death, we know little (if anything) about it, especially from the neurophenomenological point of view (neurophenomenology is scientific research aimed at combining neuroscience with phenomenology in order to study the human experience [349]). However, recent studies suggest that qEEG may shed some light on this mysterious phenomenon. The data suggest that the mammalian brain has the potential for high levels of internal information processing (consistent with conscious processing) during clinical death [350,351], suggesting that patients near death may generate a replay of memories [352]. This is supported by electrophysiological studies that have demonstrated (a) that the post-mortem human brain may retain latent capacities to respond with potential life-like properties [353], (b) that auditory systems (measured by event-related potentials) respond similarly to those of healthy controls just hours before death [354], and (c) the resting-state default mode—task-positive network anticorrelations were present among unresponsive hospice patients [355], thus suggesting that unresponsive patients may possess functional architecture in the brain that can support internally oriented thought (mind-wandering) at the end of life. Moreover, analysis of qEEG-unresponsive patients just hours before death demonstrated that they might be able to listen to music, despite being unable to overtly indicate their awareness [356].

Furthermore, studies have shown that the prevalence of qEEGs with electrocerebral activity despite a clinical diagnosis of brain death (BD) was 3.5% [357] to 19.6% [358], thus posing a challenge for the diagnostic criteria of BD and stressing the importance of qEEG utility for the confirmation of BD. Further, the association between qEEG patterns and eventual death has been demonstrated [165,328,359], thus suggesting that the qEEG may have potential prognostic value for evaluating near-term patients’ survival or death.

2.15. Causality of qEEG Oscillatory Patterns in Neuropsychopathology

The above brief review of qEEG features and properties and their association with neuropsychopathology suggests the existence of circular causality, where, on the one hand, different pathological processes affect the qEEG pattern and, on the other hand, changes in the qEEG pattern affect pathological processes. This supposition is supported by converging empirical evidence: (a) central nervous system (CNS)-active drugs that affect known neuromediators change different features of the qEEG oscillatory pattern in a consistent and predictable manner, with a parallel reduction in symptoms [360,361,362,363]; (b) specific features of the qEEG oscillatory pattern have better predictive power for medication response compared to a syndrome-based diagnosis [364,365,366,367,368,369,370,371]; for example, the overall predictive accuracy in differentiating treatment responders from non-responders is 84%, with a sensitivity of 77% and a specificity of 92% [372]; (c) different features of the qEEG oscillatory pattern predict future (i) decline within the next 7 years in normal elderly people with subjective cognitive complains (no objective evidence of cognitive deficit) [259], (ii) clinical outcomes in patients in the vegetative state 6 years after brain injury [315,327], and (iii) developments of delinquent (antisocial) behavior [373]; (d) normalization of the distorted structure of the qEEG oscillatory pattern by an exogenous magnetic field stimulation changes the subjective experience of neuropsychopathology, accompanied by a clinical decrease (>50% reduction) of symptom severity [374] (see also [375,376]); (e) normalization of atypical qEEG oscillatory patterns through operant conditioning with neurofeedback results in symptom reduction in neuropsychopathologies such as epilepsy [377,378], depression and anxiety [379], schizophrenia [380], addiction [381], ADHD [382,383], sleep disorders [384], autism [385], chronic pain [386], learning difficulties [387], and dyslexia [388]; last but not least, (f) cognitive enhancement in the elderly by qEEG neurofeedback [389,390].

A substantial corpus of evidence supports the proposition that the successful treatment of psychiatric patients results in the normalization of the previously demonstrated qEEG abnormalities [213].

Such circular causality is possible because the qEEG oscillatory pattern is not just a correlate of information processing, communication, integrated phenomenal experience, and the associated neuropsychopathology but is, indeed, a constitute (substrate) of these very things [330,391].

From the above review, it is clear that the qEEG is a natural and non-invasive ‘window’ into the living brain and mind since only the qEEG permits direct observation of the ongoing dynamics and coordinated processes organized in the patterns of brain activity that reflect the overall architecture of information processing, behavior, and subjective experience during both healthy and pathological conditions [392].

To make sense of the ‘view’ from this ‘window’, many different methods have been suggested. However, when processing the qEEG signal, it is essential to remember that it is a neurophysiological phenomenon that has its own peculiarities, regularities, and complex rules of organization which are functionally relevant [13,393,394,395,396] (for reviews, see [18,329,397,398,399]). Only when one knows these characteristics is it possible to make proper use of the qEEG as a tool and to give a more neurophysiologically adequate interpretation of the data. In connection to this, a much deeper understanding of the brain dynamics reflected in the qEEG is essential for progress in psychophysiological, cognitive, and clinical sciences.

3. qEEG Functional Structure and Signal Processing

With advances in qEEG signal processing methods, a wide range of statistical and mathematical techniques and analyses has been implemented to analyze complex oscillatory activity in spatial and multi-temporal dimensions. All of these have revealed new insights into the functional neural networks during normal functioning and neuropsychopathology.

Since there are many excellent reviews dedicated to qEEG signal processing methods [6], in this section, we will overview only the most important aspects of the functional structure of the qEEG signal that should be considered during processing. “Understanding the [q]EEG “grammar”, its internal structural organization would place a “Rozetta stone” in researchers’ hands, allowing them to more adequately describe the information processes of the brain in terms of [q]EEG phenomenology” ([400], p. 111), which is functionally relevant for healthy and pathological conditions.

Studies focused on the structural organization of qEEG signals have demonstrated that the qEEG is an extremely non-stationary, highly composite, and very complex signal [18,329,397,398,399]. This qEEG multivariability, in contrast to the popular view, is not noise but a reflection of the underlying integral neurodynamics, thus being functionally significant, information-rich [395,396,401], and individually specific [73]. The qEEG multivariability is characterized by a piecewise stationary structure where stationary processes with different probability characteristics are ‘glued’ to one another [329,398,399,401,402,403]. It is proposed that each piecewise stationary qEEG segment reflects the oscillatory state of the underlying transient neuronal assembly [13,404,405,406,407,408,409,410] that signifies a functional cortical state [330,411,412,413], which can be local (part of the cortex), global (all cortex), micro (ranging from milliseconds to seconds), or macro (ranging from minutes to hours). Here, the qEEG oscillatory state is a steady, transient, and self-organized operational unit [414] that has been proposed to present the basic building blocks of cortical activity accompanied by mentation, thinking, and information processing [415]. Activity within each state is stable (or quasi-stable) and is likely to represent a fingerprint of a functionally distinct neuronal network mode. Each qEEG oscillatory state (either local or global) is characterized by multiple qEEG oscillations, where different oscillations are mixed in different proportions depending on the level of vigilance, perceptual, cognitive and mental operations, health, or pathology (for more details, see [152]).

Analysis of the non-stationary behavior of the main qEEG signal characteristics (amplitude, frequency, and phase) has demonstrated that all three change abruptly with the progression of time: qEEG amplitude, frequency, and phase persist for some time around a stable average and then abruptly ‘jump’ to a new stable average, which, after some time, is replaced by yet another average level (for qEEG amplitude, see: [329,398,399]; for qEEG frequency, see: [395,396,416]; for qEEG phase, see: [417]). These ‘jumps’ in qEEG characteristics (or rapid transitional periods (RTPs), as we have named them [397,399]) mark the boundaries of segments of relatively stable brain functioning. The abrupt transition from one quasi-stationary qEEG segment to another, in this sense, reflects a ‘switching’ between brain states (micro, macro, or both) in specific neuronal networks or the whole cortex by the transient formation and disassembling of interconnecting cortical neuronal assemblies (neuronal assembly is defined as a set of neurons that cooperate (synchronize their activity) to perform a specific computation (operation) required for a specific function or task [418,419,420,421]) [394,398,413,422]. During such a transition, there is an abrupt change in the entropy, information, and dimensionality of the neuronal assembly (for details, see [391]). A multitude of different microstates may exist within any one particular macrostate. Consecutive macrostates, in their turn, comprise a new sequence on yet another timescale. Such functional qEEG structures comprise a nested hierarchical multivariability that reflects the poly-operational structure of brain activity [329,401,403].

The co-existence of the high multivariability of qEEG characteristics, along with the transient stabilization of these characteristics in time (metastability), has been demonstrated (the parameters of temporary stabilization of oscillatory states differ from ‘random’ EEGs, thus providing evidence for the non-occasional character of stabilization of the main parameters of neuronal activity [395]) [395]. Perhaps the high multivariability of qEEG characteristics is a reflection of the range of the brain states’ repertoire and their possible variations. On the other hand, the temporal stabilization of qEEG characteristics reflects the maintenance of some persistent pattern in neurodynamics within a particular time interval on both micro and macro levels. This suggests that the overall brain dynamics is a balancing act between multivariability and metastability [14,401].

Considering that all activities (influences) from multiple primary sources are not just mixed, summed, or averaged in a given cortex area but are integrated within the current state (activity) of this area, the local qEEG is considered to represent a functional source, which is defined as the part or parts of the brain that contribute to the activity recorded at a single sensor [423,424]. A functional source is an operational concept that does not have to coincide with a well-defined anatomical part of the brain and is neutral with respect to the problems of localization of primary source and volume conduction [423,424]) In this context, the local EEGs can be described by (a) the size of the oscillatory state repertoires (the number of the qEEG quasi-stationary segments types—the neurodynamics diversity); (b) the life-span (illustrating the functional life-span of a neuronal assembly or the duration of operation produced by this assembly; because a transient neuronal assembly functions during a particular time interval, this period is reflected in the qEEG as a stabilized interval of quasi-stationary activity [425]) of oscillatory states of each type (duration of the qEEG quasi-stationary segments of each type or period of the temporal stabilization, which shows the time during which the brain ‘maintains’ the underlying neurodynamics); (c) the probability of occurrence of a particular type of oscillatory state (the number of the most probable types of qEEG quasi-stationary segments, which indicates the most ‘preferred oscillations’ of the brain); (d) the number of functionally active oscillatory states (the types of qEEG quasi-stationary segments that change along the changes in the condition, task, or function); (e) the relative incidence of change in the type of oscillatory states (gives an estimation of the rate of relative alteration in the type of qEEG quasi-stationary segment); (f) the sequence of types of oscillatory states (consistent groupings or bundling of the types of qEEG quasi-stationary segments, representing more integral blocks of qEEG structural organization); (g) the size of the neuronal ensemble (indeed, the more neurons recruited into an assembly through local synchronization of their activity, the higher the resulting amplitude of oscillations in the corresponding qEEG channel [13]) that generates the oscillatory state, measured by the average amplitude within qEEG quasi-stationary segments; (h) stability of local neuronal synchronization within a neuronal assembly, estimated by the coefficient of amplitude variability within qEEG quasi-stationary segments; (i) neuronal assembly growth (recruitment of new neurons) or disassembly (functional elimination of neurons), measured by the average amplitude relation among adjacent qEEG quasi-stationary segments); and (j) the speed of neuronal assembling or disassembling, estimated by the average steepness among adjacent qEEG quasi-stationary segments, measured in areas near RTP [329,395,399,401,426].

This is the first level of multivariability and metastability [427] (Figure 1), where:

Multivariability is characterized by ‘switching’ from one local neurodynamic to another, with new oscillatory patterns being continually created, destroyed, and, subsequently, recreated. Therefore, there is an increase in:

• The size of oscillatory states repertoire;
• The number of functionally active oscillatory states;
• The relative incidence of change in oscillatory state types;
• Neuronal assembly disassembling;
• The speed of neuronal assemblies disassembling;

and a decrease in:

• The life-span of oscillatory states;
• The size of the neuronal ensemble;
• The probability of occurrence of a particular type of oscillatory state;
• The stability of local neuronal synchronization within a neuronal assembly;
• The sequence of oscillatory state types.

Metastability is characterized by the temporal stabilization of oscillatory states in sequential combinations. Therefore, there is a decrease in:

• The size of the repertoire of oscillatory states;
• The number of functionally active oscillatory states;
• The relative incidence of change in the type of oscillatory states;

and an increase in:

• The life-span of oscillatory states;
• The size of the neuronal ensemble;
• The probability of occurrence of a particular type of oscillatory state;
• Neuronal assembly growth;
• The speed of neuronal assemblies growing;
• The sequence of oscillatory state types.

Figure 1. The first level of multivariability and metastability measured by qEEG.

Figure 1. The first level of multivariability and metastability measured by qEEG.

Different cortex regions have different dominant qEEG oscillations [150,428] that act as resonant communication networks through large populations of neurons [140,429]. Usually, cortical oscillators communicate only with oscillators that have specific resonance frequencies [430]. They do not communicate with oscillators that have nonresonant frequencies, even though there may be synaptic connections between them. In such a way, various assemblies of oscillators can process information without any cross-interference. By changing the frequency content of bursts and subthreshold oscillations, the brain determines communication at any particular moment [431]. These oscillatory systems may provide a general communication framework that is parallel to and faster than the morphology of sensory networks [140].

It seems that RTPs (see above) also contribute to this communication framework. Studies on the spatial–temporal distribution of RTPs in qEEG amplitude [329,397,399,403], qEEG phase [427,432,433], and qEEG frequency [416] have demonstrated that (a) RTPs observed in different local qEEG signals systematically coincide in time and that (b) this RTP temporal synchronicity is not random—it occurs at significantly higher or lower levels than is expected by chance alone. This non-random RTP synchrony reflects periods of mutual temporal stabilization of quasi-stationary segments in the multichannel qEEG. At the neurophysiological level, this implies that various neuronal assemblies located in different cortical regions synchronize (temporally coordinate) their operations on a particular timescale [403,427]. Such synchronization is the brain’s true functional connectivity (as it is defined by Friston et al. [434,435]) and reflects the synchronization of operations; therefore, it is operational synchrony [397]. Since operational synchrony has been demonstrated for qEEG amplitude, phase, and frequency, it is reasonable to suggest that operational synchrony is a universal phenomenon for different characteristics of the electromagnetic brain field in which complex brain functioning is reflected.

It has been demonstrated that the pattern of the functional stabilization of cortical inter-area relations can be expressed as a mosaic of dynamic constellations of different operations executed by remote brain regions—‘operational modules’ (OMs) [397,399,401,403,436]. The lifetime of such spatial OMs is determined by the duration of the period of joint stabilization of the main dynamic parameters of the activity of neuronal assemblies that are involved in these modules. At the level of the qEEG, this process is reflected in the stabilization of the quasi-stationary qEEG segments in corresponding EEG channels that comprises a metastable state [329,399,401,437]. Here, metastability relates to the phenomenon of a constant interplay/competition between the complementary tendencies of cooperative integration and autonomous fragmentation in the activity of multiple distributed nested neuronal assemblies that defines brain activity dynamics [438] (see also [14,401]). During this metastability, the restriction in the brain’s degrees of freedom is what permits the neuronal systems to have the possibility for the interactive information exchange of the essential variables, which are important for reaching a ‘consensual decision’ that is appropriate for the functional requirements engendered by each successive stage of behavioral performance. It seems that brain areas are able to mutually influence each other in order to reach a common functional state, stabilizing the main parameters of their activities. It is likely that optimal informational processing is possible only under a proper level of the functional stabilization of intercortex relations [244,329].

In this context, each OM is a metastable spatial–temporal pattern of brain activity because the neuronal assemblies that constitute it have different operations/functions and do their own inherent tasks (thus expressing autonomous tendency) while, at the same time, being temporally entangled with each other (and thus expressing coordinated activity) in order to execute a common complex operation or complex cognitive act of a higher hierarchy [401,403,439]. Here, it is important to stress that discrete parts of the cortical networks may gain another functional meaning when they are recruited by other OM and, therefore, take part in the realization of another functional act [397]. This confirms the dominant principle of the nervous constellation centers suggested by Ukhtomsky, who discussed the variable functional role of different brain cortical areas depending on their participation in various working constellations [440].

In this context, much like the local oscillatory states described earlier, the multichannel EEGs can be described by (a) the size of the OMs’ repertoire (the number of the types (OM type is characterized by the number and topographic locations of the cortex areas that mutually stabilize (temporally synchronize) the main parameters of the neuronal networks involved (temporal synchronization of RTPs)) of spatial configurations of qEEG quasi-stationary segments (temporal synchronization of RTPs)—coordinated neurodynamic diversity); (b) the life-span of OMs of every type (duration of spatial configurations of the qEEG quasi-stationary segments of every type—period of RTPs’ temporal stabilization—shows the time window during which the brain ‘maintains’ underlying coordinated neurodynamics); (c) the probability of occurrence of a particular type of OM (the number of the most probable types of spatial configurations of qEEG quasi-stationary segments—synchronized RTPs); (d) the number of functionally active OMs (the types of spatial configurations of the qEEG quasi-stationary segments that change along with the changes in condition, task, or function); (e) the relative incidence of change in the type of OMs (presents an estimation of the rate of relative alteration in the type of spatial configuration of the qEEG quasi-stationary segments); and (f) the sequence of types of OMs (consistent grouping or bundling of the types of spatial configurations of the qEEG quasi-stationary segments, reflecting the more integral blocks of qEEG structural coordinated organization) [329,401,403,439].

This is the second level of multivariability and metastability (Figure 2) where:

Multivariability is characterized by ‘switching’ from one coordinated neurodynamic to another, with new OMs being continually created, destroyed, and, subsequently, recreated. Therefore, there is

an increase in:

• The size of the OMs’ repertoire;
• The number of functionally active OMs;
• The relative incidence of change in the type of OM;

and a decrease in:

• The life-span of OMs;
• The probability of occurrence of a particular type of OM;
• The stability of the sequence of types of OMs.

Metastability is characterized by the temporary stabilization of RTPs (formation of OMs) in sequential combinations. Therefore, there is

a decrease in:

• The size of the OMs’ repertoire;
• The number of functionally active OMs;
• The relative incidence of change in the type of OM;

and an increase in:

• The life-span of OMs;
• The probability of occurrence of a particular type of OM;
• The sequence of types of OMs.

Figure 2. The second level of multivariability and metastability measured by qEEG.

Figure 2. The second level of multivariability and metastability measured by qEEG.

In this context, the participation of cortex areas in the organization of a common functional act is reflected not so much by the presence of a shared qEEG rhythm in different EEG channels (distant neuronal ensembles) but by the systematic synchronization of the moments of switching (RTPs) between qEEG oscillatory modes in the different cortex areas. The fact that operational synchrony is sensitive to the morpho-functional organization of the cortex rather than the volume conduction and/or reference electrode differs from surrogate data (random combination of RTPs) and is functionally sensitive to different cognitive tasks as well as healthy and pathological conditions; this suggests that operational synchrony reflects the remote temporal coordination of brain operations performed by local neuronal assembles (for relevant details, see [244,329,398,399,403]).

It has been suggested that disturbed synchrony in distributed qEEG oscillations may reflect dysfunction within resting-state networks in neuropsychopathology. It seems that a loss of optimal metastable balance between independence and the integrative processes in large-scale cortical activity, which may be due to a dysfunction in large-scale cortical integrative processes and a poverty of regional physiological variation [441], is associated with the ‘cognitive disintegration’ [442], ‘cognitive dysmetria’ [443], and ‘thalamocortical dysrhythmia’ [444] that are typical for a number of neuropsychopathologies [244,276,445]. This view is consistent with the modern concept of brain and mind disorders, where the disease is considered to be a process, with a change in the balance of autonomy (low functional connectivity) and connectedness (high functional connectivity) of different brain systems that sustains health [244,276,438]. Indeed, on one hand, deficits in the ability of cortical areas to coordinate could produce a lack of mutual constraint, leading to excessive expression of local processing unrestrained by large-scale context. On the other hand, an excess of coordination could stifle independent area expression and cause a stereotyped processing rigidity. Thus, the alteration in brain functional connectivity might serve as a contributing factor to the disorganization syndrome [446].

From this overview, it is clear that any qEEG signal processing method should take into account the functional structure of the signal in order to be able to extract neurophysiologically relevant information.

In order to provide accurate and practical methods for separating individuals with cognitive impairments or behavioral disturbances into those with and without quantitatively demonstrable abnormalities in brain electrical activity (qEEG) and, thus, altered neurophysiology, statistical evaluation of qEEG characteristics relative to population normative values is needed. Such possibility was demonstrated by the introduction of so-called ‘neurometrics’ [153]. Machine learning algorithms and artificial intelligence techniques are used for this purpose as well. However, the limitation of these methods is their inherent non-explainability: no insight can be obtained from the inferred output. Without the explainability of the learned inference mechanism, not much insight can be gained in terms of the underlying brain activity patterns and mechanisms, which are important to better understand neuropsychopathology. Within this neurometrics, a given qEEG characteristic of the individual is transformed from its original units (voltage, time, latency, coherence, and symmetry) to a common metric reflecting the relative probability of that value within a healthy population normative reference. This allows researchers to compare or combine measures that have not initially been dimensionally comparable. From this, ‘abnormality’ is derived and defined as statistically improbable values exceeding those expected by chance alone. Thus, a clustering of qEEG deviations provides evidence of underlying functional neurophysiology abnormality that can be associated with the patient’s clinical condition or neurological/psychiatric problems. This neurometric analysis of qEEG characteristics was successfully applied to a variety of neuropsychopathologies, with high replicability, specificity, and sensitivity to a wide variety of cognitive, developmental, neurologic, and psychiatric disorders [69,248,413,447,448,449].

Although much has been learned through systemic, cognitive, and clinical neuroscience about the underlying neural mechanisms and functional correlates of qEEG oscillatory activity, surprisingly little conceptual integration of this knowledge is present in clinical applications of qEEG. Even though the qEEG (a) reflects developmental maturation and aging, anatomical and functional integrities of the brain, including cognitive processes and the functional status of the brain as well as diverse neuropsychopathologies; (b) has high predictive capacity, sensitivity, and specificity for identifying responders and non-responders; and (c) provides qualitative predictions for a patient’s state after treatment courses (see Section 2.1, Section 2.2, Section 2.3, Section 2.4, Section 2.5, Section 2.6, Section 2.7, Section 2.8, Section 2.9, Section 2.10, Section 2.11, Section 2.12, Section 2.13 and Section 2.14), all of this knowledge should be put in a wider conceptual context of the health–disease continuum in order to refine the diagnostic capability of qEEGs. In the following section, we will attempt to integrate conceptually recent neuropsychophysiological and electrophysiological findings within a common theoretical framework in order to reduce the divide between state-of-the-art research and current clinical practice.

4. Common/Unifying Theoretical–Conceptual Framework

A general framework needs to be developed in order to allow researchers and clinicians to organize systematically and understand the enormous diversity of observations related to qEEG characteristics in neuropsychopathology. As a first step, a general concept of dynamic qEEG oscillatory patterns has been proposed [450]. Within this concept, a dynamic qEEG oscillatory pattern is considered as a spatial–temporally organized superimposition of ongoing multiple qEEG oscillations in many frequency bands [26], where different oscillations are mixed in varying proportions based on the vigilance level; perceptual, cognitive, and mental operations; behavior; and the extent of pathologic processes. It seems that different qEEG descriptors can be combined within the dynamic qEEG oscillatory pattern, which is characterized by (a) frequency content, including the composition of delta (0.5–3 Hz), theta (3–7 Hz), alpha (7–13 Hz), beta (13–30 Hz), and gamma (>30 Hz) qEEG oscillations, along with their proportions, dominant frequencies, amplitudes and powers, and (b) spatial heterogeneity (expressed in spatially structured extracellular electrical fields), including spatial complexity (amount of brain connectivity), interhemispheric symmetry, and hubs (cortex areas with the highest neuropsychopathology effect or highest functional connectivity). This concept was successfully applied to major depressive disorder [450].

The concept of dynamic qEEG oscillatory patterns is useful since it enables researchers to combine all known and newfound qEEG characteristics within one entity. Studies (see above) have suggested that dynamic qEEG oscillatory patterns reflect the structural and functional integrity of the brain, including an information-processing mode that is dynamically regulated by interactions within a homeostatic system that is mediated by many different neurotransmitters on one side and functional activity and various perceptual and cognitive operations associated with a mental or behavioral condition in health and pathology on the other.

The categorization of different types of dynamic qEEG oscillatory patterns resulted in the second concept—the qEEG phenotype.

4.1. qEEG Phenotype and Neuropsychophysiological Type

qEEG phenotypes are the clusters of commonly occurring qEEG characteristics within qEEG oscillatory patterns found in the general population that are believed to be the net result of genetics, pre- and post-natal individual development, and significant life events such as brain traumas or neuro-diseases (see above). In contrast to the classical view, which sees qEEG phenotypes as classes of qEEG abnormalities, where, usually, one qEEG characteristic defines a qEEG phenotype and each phenotype is static [451], we believe that qEEG phenotypes characterize all variabilities of the population, covering the whole spectrum, from health to pathology, and, further, that such phenotypes are dynamic. In this view, some of the qEEG phenotypes are characteristic of healthy conditions, while others are typical for different degrees of deviation from the healthy state up to the situation where the qEEG phenotype is pathological. Additionally, every phenotype is defined by the combination of qEEG features (qEEG oscillatory pattern—a coherent functional whole) and is dynamic, meaning that every qEEG phenotype may exhibit changes, to some extent, within its own qEEG oscillatory pattern due to dysfunction or development of pathology (Figure 3). This is so because the type of qEEG oscillatory pattern is a phenotypic expression (qEEG phenotype) of (a) cellular and biochemical (dys)function; (b) maturational processes (or delaying factors), partially genetically and epigenetically determined; (c) neurotransmitter (im)balance; (d) regulatory systems (and their disturbances); (e) early subclinical organic brain damage; or (f) morpho-functional disturbances that may be present in neuropsychiatric disorders.

Such conceptualization underwent theoretical development and empirical verification in the work of Zhirmunskaya and colleagues [452,453,454,455,456] and was successfully applied to several clinical neuropsychopathologies as well as various modes of pharmacological influence [156,457,458,459,460,461,462].

In this context, a given qEEG phenotype reflects a neuropsychophysiological type (neurodynamic constitution) of the individual. Indeed, numerous studies have demonstrated that qEEG characteristics can encompass functionally different psychophysiological determinants [1,463]. On one hand, a qEEG phenotype reflects the inherent brain functional organization and dynamic structure of brain activity, which are intra-individually stable traits, as evidenced by test–retest reliability (Section 2.2) and genetic studies (Section 2.3). On the other hand, since intrinsic brain activity supports and conditions individual cognitive and information processing, self-regulatory functions, decision-making, behavior, and consciousness, a qEEG phenotype reflects the neurophysiological predispositions of the underlying cognition and personality, temperament, and character factors thus reflecting the individual’s psychological and behavioral traits. Indeed, it was found that individual differences in qEEG variability relate strongly to stable indicators of subject identity [464]. Hence, both an individual’s neurophysiological and psychological differences can be neuropsychophysiologically interpreted under the same unified notion of the qEEG phenotype [1,465,466,467,468,469,470].

This brings us to the idea that a qEEG phenotype is not just a concept but a real phenomenon. This is supported by the observations that qEEG phenotypes reliably predict the effectiveness of drug interventions, while nosological or behavioral groupings do not [451]: for example, effective treatment of ADHD [366,471], refractory depression [472], and major depressive disorder [369] was achieved when it was based on prospectively identified qEEG phenotypes related to different sub-types within the diagnostic group, thus suggesting that nosological heterogeneity is well-reflected in the multiplicity of spatial–temporal parameters of qEEG oscillatory patterns. Additionally, qEEG phenotypes are a better indicator of pathology vulnerability when compared to classical evaluation: for example, it was demonstrated that qEEG has a better indication of alcoholism susceptibility than the customary dichotomous affection status [473,474].

In summary, a constellation of different qEEG characteristics that are united within the qEEG oscillatory pattern and expressed as the qEEG phenotype should be considered more appropriate for diagnostic and medication/therapy-response purposes. It seems that every qEEG phenotype represents a set of quantitative neuropsychophysiological, cognitive, and behavioral traits that determine an individual’s liability or vulnerability to develop or manifest a particular neurodysfunction or disease [3,475,476]. Thus, qEEG phenotypes may aid in revealing disease-specific causal pathways and may aid people in finding a work/lifestyle balance that is more in keeping with their natural predispositions.

4.2. qEEG Phenotype and Neurophysiological, Cognitive, and Behavioral Traits

Studies on intra- and interindividual differences have revealed that qEEG oscillatory patterns reflect individually specific peculiarities of homeostatic and adaptive regulation [468,477,478,479,480]. Additionally, it has been demonstrated that the qEEG phenotype reflects the specificity of intracortical and cortico-subcortical interrelations and is, to a significant extent, a neurodynamic substrate of the psychological properties of the personality [481,482,483,484,485]. Together, these suggest that neurophysiological, cognitive, and behavioral traits are reflected in the characteristics of the qEEG phenotype. Indeed, studies have demonstrated the association between personality traits and qEEG characteristics [486,487,488,489,490,491,492,493,494,495,496,497,498,499,500,501]. Moreover, normal EEG patterns have been correlated with a well-integrated personality (general personality fitness) [502] (for a review, see [503]).

Personality traits have been proposed to constitute vulnerability factors for psychopathology, including mental diseases and affective disorders [504,505,506,507,508,509,510,511], thus suggesting that personality traits and psychopathology are not distinct entities [512]. They both (a) reflect increased vulnerability for and (b) underly any given psychopathology.

Thus, it seems that, on one hand, the qEEG phenotypes are related to personality types (see above and also [513]) and psychopathology. On the other hand, personality features are related to psychopathology too. This similarity of qEEG relations between psychopathology and personality traits suggests that they are mediated by the same neurophysiological mechanisms and behavioral patterns. Therefore, it is reasonable to suggest that behaviors intrinsic to personality traits are the same as those that are exaggerated in psychopathological conditions [514].

There are many different individual traits with varying degrees of significance for neurophysiologic regulation, information and emotions processing, and behavior. For the purpose of this article, we will focus on those traits which are (a) expressed along the entire continuum of functioning, from health to pathology; (b) transdiagnostic; and (c) reflected in the characteristics of qEEG phenotypes, thus being fundamental and primary functions that condition (modify, modulate, or mediate) other dimensions.

The Research Domain Criteria (RDoC) project, launched by the National Institute of Mental Health in 2009 [515], is the first comprehensive attempt to identify fundamental neurobehavioral dimensions that cut across current heterogeneous disorder categories. According to the RDoC, there are five major domains of functioning:

• Negative valence domain;
• Positive valence systems;
• Cognitive systems;
• Systems for social processes;
• Arousal/regulatory systems.

Since some qEEG characteristics can be related to several domains, constructs, or subconstructs of the RDoC matrix and phenomenally different clinical characteristics are, at a more fundamental level, implementations of one and the same process, we suggest the following fundamental and primary domains of functioning that are (a) expressed along the entire continuum of functioning, from health to pathology; (b) transdiagnostic; and (c) reflected in the characteristics of qEEG phenotypes:

• Tonic level of vigilance (corresponds to arousal/regulatory systems in RDoC);
• Speed of information processing (corresponds to cognitive systems in RDoC);
• Directedness of the attention (internal vs. external focus) (corresponds to cognitive systems in RDoC);
• Emotional–motivational tendency (corresponds to positive and negative valence systems in RDoC);
• Sociability (sensory stimulation and excitement tolerability) (corresponds to systems for social processes in RDoC);
• Anxiety tendency (anxious arousal vs. anxious apprehension) (corresponds to negative valence and arousal/regulatory systems in RDoC);
• Stress regulation (resistance and recovery);
• Overall brain resources (resilience).

The fact that individual differences in dispositional moods, stress resilience, behavioral orientation to physical or social objects, temporal processing (primordial sense of flow between events), and brain resources are shaped/present in very early childhood, long before cultural standards and knowledge are internalized by an individual [516,517,518], supports a fundamental and primary nature of these domains of functioning.

One can see that the majority of the suggested domains can be mapped onto the RDoC matrix. In the following sections, we will give a brief overview of these domains.

4.2.1. Tonic Vigilance

Vigilance (or arousal) here refers to the individual’s predominant level of supply energy available to the brain’s regulatory systems. It is responsible for generating the activation of neural systems appropriate for various contexts and providing appropriate homeostatic regulation of systems such as energy balance and sleep.

There are more or less stable individual differences in the baseline vigilance: some individuals are constantly in a highly activated state (hypervigilance), while others have a chronically low level of activity (hypovigilance) [519], thus reflecting an individual’s predominant vigilance state—tonic vigilance. The underlying cause is believed to be different levels of activity in the loop connecting the brainstem reticular formation with the cortex. As such, tonic vigilance reflects the predominant baseline cortical and mental arousal that sets the overall level of activity of the entire brain and, as a consequence, the body. In this way, tonic vigilance, as the brain’s trait energetic capacity, sets the stage for cognitive and behavioral performance.

It seems that tonic vigilance is one of the most important traits of an individual since it determines the optimal range of activating factors (a sympathovagal balance) on the cerebral cortex that is necessary for the formation of an active, energetic state [520]. Outside this optimal range, work capacity drops and the subjective sense of well-being decreases, while psychopathology vulnerability increases [521]. Indeed, studies have demonstrated that hypervigilance is a common feature of various anxiety and affective disorders (depression), including PTSD [522,523,524,525], hyperaroused fatigue (overloading) with reduced sleep propensity, inhibition of drive, and eventual exhaustion [526,527,528]. Conversely, hypovigilance is associated with hypoaroused fatigue with increased sleepiness, a lack of drive, sickness behavior [526,529,530,531], burnout [532], impulsivity [533], ADHD [241,534,535], and AD [135]. This is supported by the Arousal Regulation Model of Affective Disorders [269], where hypostable and hyperstable levels of arousal contribute to manic or ADHD and depressive-like behavior, respectively [270,536]. Importantly, tonic vigilance has been suggested to causally contribute to mental illness [269]. Additionally, the relationship between tonic vigilance, personality traits [501,513,537], and psychopathology have been proposed [538,539] (see also Section 4.1).

Besides (a) being expressed along a health–pathology continuum and (b) being transdiagnostic, tonic vigilance is reflected in the characteristics of the qEEG phenotype [157,523,540,541,542,543,544,545,546,547]. Additionally, vigilance-relevant qEEG characteristics are ~80% heritable [548].

To summarize, tonic vigilance affects cognition, emotions, and behavior and is implicated in the etiology of psychiatric disorders and reflected in the characteristics of qEEG phenotypes.

4.2.2. Speed of Information Processing

Even though this domain is dependent on tonic vigilance, it has added significance in relation to normal functioning and neuropsychopathology.

The speed of information processing through periodic cycles of sampling and sensory integration determines the pace of perceptual performance and behavior [148,549,550]. In other words, the speed of information processing determines the resolution at which information is sampled and/or processed by cortical neurons, and, therefore, it limits the capacity for storage, transfer, and retrieval of information in the brain [551,552,553]. In such a way, the speed of information processing regulates the overall amount of information that reaches the cortex and is related to reaction time, speed, and capacity of cognitive and motor task performance, cognitive preparedness, temporal integration of information across the senses, as well as reasoning ability.

The speed of information processing is an individually stable trait (varies within an individually specific range) and is predictive of the individual’s reaction and decision times, attention, and working memory performance across the age span [21,69,78,144,553,554,555,556,557,558,559,560,561].

The speed of information processing is expressed along the health–pathology continuum, where slowed information flow is observed in sorrow and fear emotions, physiologic and pathologic aging, mild cognitive impairment and AD, vascular dementia, psychosis and schizophrenia, ADHD, chronic fatigue, burnout syndrome, and melancholic depression [117,119,124,530,532,544,562,563,564,565,566,567], whereas an increased speed of information processing is typical for anger, anxious arousal and panic disorder, hyperaroused fatigue (overloading), anxious depression, and PTSD [523,562]. Additionally, temporality (inner and outer time speed perception) is recognized as an important factor in psychopathology [568,569,570], where an imbalance between inner and outer time speed perception was demonstrated. For the transdiagnostic character of the speed of information processing, see [518].

Importantly, qEEG characteristics are causally related to the speed of information processing [21,144,148,549,550,559,561] and the temporal integration of information across the senses [552,553,571,572]. Additionally, qEEG characteristics associated with the speed of information processing have a heritability of 81% [89].

To summarize, the speed of information processing is controlled by the characteristics of qEEG phenotypes that are thought to function as a timing or gating mechanism in operation processing [573]: shorter-duration qEEG oscillations provide more gating signals per unit of time and, thus, result in faster information-processing rates and shorter RTs, whereas longer-duration qEEG oscillations provide fewer gating signals and lead to slower processing rates and longer RTs [574]. This is in line with ‘physiological lability’ (introduced by Wedensky [575]), defined as the capacity of the system to perform a certain amount of complete work cycles per unit of time [576]. Additionally, alterations in the speed of information processing are characteristic of various neuropsychopathologies.

4.2.3. Directedness of the Attention (Internal vs. External Focus)

Attention is one of the basic human cognitive abilities that allow the discrimination of relevant parts of information and the ignoring of others; it usually refers to a more focused activation of the cerebral cortex that enhances information processing [577]. Attention depends on the vigilance level and, therefore, can be predicted through it; however, directedness of attention is independent of vigilance and, thus, should be considered separately.

It seems that directedness of attention can be conceptualized as internal attention (self-focus) and external attention (environment-focus). Some healthy individuals are generally well tuned into both internal and external events, which helps them to flexibly reallocate attention ‘in’ and ‘out’ in order to adapt their behavior to the needs of the current situation. However, others have the propensity to be self-focused or environment-focused. Exaggerated unbalanced attentional focus is typically associated with neuropsychopathology and is known as attentional bias. Attentional biases are characterized by an inability to flexibly reallocate attention to relevant internal or external stimuli, for example, an increase in the orienting of attention towards threat-related stimuli and attentional avoidance/difficulty disengaging attention from irrelevant but negatively valanced stimuli [578].

When enhanced focus on the external environment prevails (e.g., family, friends, social and work duties, including the distracting effects of environment), the individual is more alert, anxious, or irritable due to the inability to inhibit irrelevant/inappropriate external information. Therefore, there is a predominance of perceptually guided contents over self- and somatically-guided contents; this is usually associated with anxiety, irritability, stress, and mania [579,580,581,582].

When an individual has excessive self-focus, it elevates volitional control (including both execution monitoring and internal focus) but also interferes with automatic actions, which stems from an inability to inhibit irrelevant/inappropriate internal information [583]. Excessive self-focus occurs at the expense of external environmentally oriented contents and their respective social and psychomotor functions. This shift may be manifested in symptoms such as ruminations, various somatic and vegetative symptoms, social withdrawal, lack of motivation, and psychomotor retardation, usually observed in depression and fatigue [300,581,582,584,585]. Moreover, because excessive resources are allocated to processing internal mental contents, other outward-oriented aspects such as attention, working memory, and episodic memory are compromised as well, which can lead to various neuropsychological deficits.

Notice that in both scenarios (excess or deficit of internal attention), perceptually guided contents (environment) are not properly integrated with the somatic- and self-generated contents (self), which results in their disbalance. This disbalance of internal and external contents seems to be abnormally tilted towards either the internal or external content, which, importantly, also goes along with different degrees in the expression of self [300,305].

Converging evidence suggests that directedness of attention is reflected in the characteristics of qEEG phenotypes [21,485,579,586,587,588,589,590,591,592,593].

4.2.4. Emotional–Motivational Tendency

Emotional–motivational tendency reflects the predisposition of an individual to engage in certain types of emotional (positive vs. negative) and motivational (approach vs. withdrawal) responses [594]. Hence, the emotional–motivational tendency may be viewed as a ‘diathesis’ (trait) that predisposes an individual toward a particular affective and motivational style and establishes risk factors for developing certain psychopathologies [595,596,597,598,599,600].

Indeed, excessive negative and withdrawal tendencies can result in behavioral inhibition expressed as fatigue, lack of energy, apathy, and slow psychomotor functioning, with a stronger hormonal response to stress (higher cortisol level), where more situations/stimuli are perceived as stressors [169,171,599,601,602]. Likewise, it is accompanied by reduced baseline cellular immune function, negativity bias (alongside reduced reward sensitivity), and sadness, fear, or depression [596,599,603,604,605]. The tendency toward negativity and withdrawal is associated with a personal pessimism bias—the person believes that negative events are more likely and positive events are less likely to happen to him/her than they are to other people. The tendency toward negativity and withdrawal has also been observed in psychiatrically healthy offspring of individuals with depression [606].

Conversely, excessive positive and approach tendencies can result in the individual having excessive behavioral activation, expressed as increased muscle tension, agitation, and somatic symptoms of arousal, with smaller hormonal responses to stress (fewer situations/stimuli are perceived as stressors), positive emotions, mania, jealousy, anxiety, or anger [599,602,607,608,609,610]. The tendency toward positivity and approaching tasks/challenges is associated with a personal optimism bias—the person believes that positive events are more likely and unpleasant events are less likely to happen to him/her than they are to other people [611,612].

Similar to other domains of functioning, converging evidence has demonstrated that emotional–motivational tendency is reflected causally in the characteristics of qEEG phenotypes with high reliability and internal consistency [76,170,613,614,615,616,617,618,619,620,621,622] and is associated with a genetic risk for depression [623].

4.2.5. Sociability (Sensory Stimulation and Excitement Tolerability)

Sociability is a trait related to sensation/experience seeking, disinhibition propensity, and boredom susceptibility. It reveals a tendency to interact well with others and the degree to which a person can tolerate/enjoy sensory stimulation from people and situations.

Sociability is a complex facet of introversion–extraversion [499,624], which is related to behavioral inhibition and activation systems (BIS/BAS) [499,625,626] and is associated with cortical arousal via the ascending reticular activating system (ARAS) [627]. In this context, introverts are believed to have a lower threshold for arousal; therefore, they are assumed to be chronically ‘over-aroused’ and, thus, tend to seek a state of lower arousal. Conversely, extraverts are believed to have a higher threshold for arousal, as they are assumed to be chronically ‘under-aroused’ and tend to seek a state of higher arousal [537,628].

Both under- and over-sociability are associated with a raft of psychopathologies [629,630]. Thus, lack of sociability can result in the individual preferring low sensory stimulation due to heightened baseline cortical arousal, having a lower threshold for arousal, and, therefore, high behavioral inhibition, increased emotional tension, and depression. Such a person is excited by low-intensity signals and inhibited by high-intensity signals.

In contrast, excess sociability can result in the individual preferring strong sensory stimulation and a propensity for sensations/excitement/novelty-seeking due to baseline cortical under-arousal, having a higher threshold for arousal and, therefore, low behavioral inhibition [631]. Such a person is excited by high-intensity input signals and inhibited by low-intensity ones. In the extreme, over-sociability is associated with psychopathology [632,633,634], including mania, narcissism, psychopathy, substance abuse, excessive venturesomeness (excitement-seeking), and various forms of externalizing (risk-taking, grandiosity, exhibitionism, manipulativeness) [635], suggesting that excessive sociability potentially represents a vulnerability factor for other conditions.

Converging evidence suggests that sociability is also reflected in the characteristics of qEEG phenotypes [494,495,496,499,501,636,637,638,639,640,641,642].

4.2.6. Anxiety Tendency (Anxious Arousal vs. Anxious Apprehension)

Anxiety is one of the best-known and oldest evolutionary systems evolved in humans. It results from a set of information-gathering reactions that allow the individual to face uncertainty and danger and survive. Despite being adaptive, since it helps us avoid dangerous situations and to achieve our goals, anxiety also causes significant suffering and, in its extreme forms, can be debilitating.

Biologically, there are two subcomponents of anxiety: somatic and cognitive [643]. Somatic anxiety (or anxious arousal) is a physiological component of anxiety, characterized by autonomic arousal and somatic tension (high blood pressure, pounding heart, sweating, dryness of mouth, difficulty breathing) [643,644,645]. Cognitive anxiety (or anxious apprehension) refers to the mental component of anxiety and consists of expectations about and anticipations of a difficult task or threat, success or self-evaluation, worrying, negative self-talk, and disrupted attentional processes [643,645,646]. It is important to stress that each of the two components of anxiety is associated with different psychopathologies.

A person with an anxious arousal tendency is usually alert during the distraction-free resting-state period and experiences rest periods as more aversive and anxiety-inducing. Extreme anxious arousal is typical for neurosis, panic or phobic disorders, and PTSD [647,648].

A person with an anxious apprehension tendency usually has (a) higher expectation for the perceptual events, which reflects higher nonselective readiness for perception and action, even in the absence of any goal-directed task and (b) excessive worry for the future and verbal rumination about those expectations. Highly expressed anxious apprehension is typical for generalized anxiety disorder, obsessive–compulsive disorder, and separation anxiety disorder [647,648].

Again, converging evidence suggests that both components of anxiety are reflected in the characteristics of qEEG phenotypes [497,649,650,651,652,653].

4.2.7. Stress Regulation (Resistance and Recovery)

Considering the link between qEEG oscillatory patterns and stress regulatory systems such as the hypothalamic–pituitary–adrenocortical (HPA) axis and the sympathetic–adrenomedullary axis [168] (see Section 2.7), as well as the association between early life adversity and the subsequent stress regulatory profile in the adult [654,655], it is reasonable to consider stress regulation as a trait.

Stress regulation determines the ability of an individual to withstand, adapt to, and recover from stress [656]. The brain plays a central role here since it perceives and determines what is threatening and executes behavioral and physiological responses to the stressor [657]. When stress regulation is altered (due to chronic stress or individual vulnerability), there is (a) a decreased capacity to resist, adapt, and recover from stress and (b) a tendency to perceive the stressor as a threat (ether to one’s physical safety or to one’s ego/social sense of self). For example, early life adversity may lead to persistent changes in the neural network balance that increase sensitivity to emotional stimuli [658] and is often associated with blunted HPA-axis reactivity.

Stress dysregulation is associated with a transition from adaptive to maladaptive stress responsivity and stress-related disorders [654,655] and neuropsychopathologies such as inattention, depression, anxiety, and insomnia.

It seems that stress regulation is also reflected in the characteristics of qEEG phenotypes [177,659,660,661,662,663].

4.2.8. Overall Brain Resources (Resilience)

The overall brain resources reflect the brain’s morpho-functional integrity, capacity for self-reorganization, self-regulation, and adaptation, and information processing efficiency. The brain resources domain relates to the general capacity of the brain to withstand neuropathological changes before overt behavioral, functional, or cognitive impairment manifests. This domain unites brain and cognitive reserves [664,665,666]. Brain reserve refers to quantitative aspects of the brain (structure or ‘hardware’), including but not limited to the number and size of neurons, the number of connections between neurons, fiber density, axonal diameter, the degree of myelination, the integrity of corticocortical and thalamocortical circuits, hippocampal volume, the number of active synapses in the thalamic nuclei, and the number of potential neural pathways [664,667,668,669]. Cognitive reserve refers to the ‘neuropsychological competence’ aspects of the brain, that is, how well the underlying ‘hardware’ is used (functions or ‘software’). It reflects a process where the brain actively attempts to cope with brain challenges or damage by using pre-existing neural networks and/or by recruiting additional or alternative brain regions to support the task network. Cognitive reserve also includes the processes of network efficiency and neural compensation [664,666,669,670].

In this view, the dualism of brain and cognitive reserves is considered within a single framework—brain resources—where the term ‘resource’ refers to the joint structural and functional characteristics of brain networks that offer cognitive protection in disease.

A person with high overall brain resources has increased compensatory and neuroprotective mechanisms that give the person increased capacity for effective cognitive functioning in spite of neuropathophysiological challenges or aging. A person with high brain resources has a younger brain phenotype and is more likely to remain within normal limits for a longer period of time despite the possible parallel progression of underlying disease [671]. Therefore, a person with high brain resources and a high disease burden may remain asymptomatic due to compensatory and resistant adaptations of the functional brain network [667,672,673] (Figure 4).

A person with low overall brain resources has decreased compensatory and neuroprotective mechanisms that decrease the capacity for effective cognitive functioning even without neuropathophysiological challenges. A person with fewer brain resources has an older brain phenotype, so he or she may express symptoms after a trivial brain dysfunction due to redundant neural pathways and an inflexible or intolerant functional network. Likewise, the brain has a lower threshold for the expression of functional impairments following the onset of brain pathology [667,672,673] (Figure 4).

Since the qEEG oscillatory pattern reflects structural and functional integrity of the brain (see Section 2.4 and Section 2.5) and causally relates to neuropsychopathology (see Section 2.15), it is not surprising that overall brain resources are reflected in the characteristics of the qEEG phenotype [8,112,113,116,119,121,122,136,143,144,185,187,671,675,676,677,678].

From this brief overview of the fundamental and primary domains of functioning, it is clear that each (a) exhibits trait properties; (b) is expressed along the entire continuum of functioning, from health to pathology; (c) is transdiagnostic; and (d) is reflected in the characteristics of qEEG phenotypes.

It has been proposed that trait characteristics and qEEG phenotypes should only be evaluated during a resting state [453,519].

4.3. Why the Resting State?

Studies of the closed-eyes resting condition provide an important opportunity to examine baseline qEEG patterns unbiased by any task. The resting-state condition avoids the confounding effects of visual scenes, instructions, and task execution (i.e., expectation matching, strategies employed, motivation or lack of it, fatigue, and anxiety associated with task performance). Additionally, the resting state seems more self-relevant than standard cognitive tasks, which typically drive subjects to direct their attention away from their personal concerns [50]. The resting-state condition permits the assessment of the ‘pure’ self-relevant baseline brain and mind activity [51]. This activity reflects the individual type of spontaneous processing of an internal mental context (top-down processing) [52], such as random episodic memory [53] and related imagery [54], conceptual processing [55], stimulus-independent thought [56], self-reflection, internal narrative, and autobiographical self [57,58,59]. The frequently expressed concern that unconstrained brain activity varies unpredictably does not apply to the passive resting-state condition of the human brain. Studies have shown that “it is rather intrinsically constrained by the default functionality of the resting-state condition [60]” (references in this citation can be found in [450], p. 1051) and that this constrained default functionality comprises the individual neurophysiological type.

Indeed, numerous studies have demonstrated that the resting-state qEEG represents the default functional infrastructure of the brain that is involved in information processing related to inherent and relatively stable capacities of the individual, such as emotional regulation, cognition, and behavior as well as individually specific neurocognitive mechanisms underlying adaptation to motivationally and intellectually challenging tasks or conditions and the systemic self-regulation of brain functions [1,11,679,680,681]. Additionally, relations between personality traits and resting-state activity have been found for each of the Big-5 personality traits (for a review, see [682]). The resting qEEG has a non-random complex spatial–temporal structure (see Section 3), is highly predictable (Section 2.2), and is regulated by a complex neuroanatomical and neurochemical homeostatic system (see Section 2.9). This system is genetically based (Section 2.3) but also demonstrates some flexibility, e.g., epigenetic changes as a result of lifestyle changes (Section 2.1). The resting-state qEEG may thus serve as an intrinsic functional ‘fingerprint’ that captures trait-like features of brain organization that are relevant for neurocognitive functioning in healthy and pathological conditions. This intrinsic functional ‘fingerprint’ reflects an individual’s neurophysiological type, which can be captured by the qEEG phenotype [1,392,421].

The importance of the resting-state qEEG is supported by evidence that functional brain organization, as measured in the resting condition, is predictive of task execution, performance, and behavioral reactions during the actual activity or task [329,683,684,685,686]. Additionally, most of the energy used by the brain goes into supporting resting and ongoing neuronal activity [687,688,689]. A task-related increase in neuronal metabolism is usually small when compared with this large resting energy consumption [689]. These facts also support the importance of the resting-state neuronal activity that consumes most of the brain’s energy.

In this context, the resting-state qEEG manifests baseline trait mechanics of self-organization that regulate multiple brain systems, thus adapting the brain and body to an ever-changing environment [690,691]. Thus, the resting-state qEEG reflects intrinsic baseline/default activity that instantiates the maintenance of information for interpreting, responding to, and even predicting environmental demands. Here, the resting-state constitutes a reference baseline, relative to which all cognitive and physiological states in healthy and pathological conditions can be considered [288,289,680].

Thus, following Fox and Raichle [692], it can be suggested that the pattern of the resting-state qEEG (qEEG phenotype) serves as a functional localizer (‘content’), providing a priori information about the way in which the brain will respond across a wide variety of tasks during healthy and pathological conditions (‘context’). Indeed, in all of the neuropsychopathologic conditions we studied, none of the participants could reach a proper resting state that is typical for a healthy brain [152,244]. The corollary is that such a system is less able to cope with the demands of a constantly changing environment.

Based on this logic, we consider abnormality in a closed-eyes resting qEEG to be a core feature of any neuropsychodysfunction or -pathology [152,244]. In this context, alteration in the closed-eyes resting qEEG oscillatory pattern (qEEG phenotype) may constitute a tonic component of qEEG microstructural organization that can serve as the field of action for abnormalities governed by multiple causes [450].

4.4. qEEG Phenotype, Personality Traits, and Norm-Pathology Continuum

The above review (Section 4.2) demonstrates that variability in personality traits and their associated characteristics of qEEG phenotypes are expressed in both healthy and pathological conditions. This suggests that personality traits and neuropsychopathology are not distinct entities, but rather, they manifest along a unified continuum of functioning, where mental diseases can be considered to fall along multiple continuous trait dimensions, with traits (and corresponding characteristics of qEEG phenotype) ranging from normal to extreme (pathological). Here, neuropsychopathology is considered in terms of dysfunctions of various kinds and degrees. In other words, neuropsychopathology is the over-expression or under-expression of personality traits and the associated characteristics of qEEG phenotypes that are otherwise moderately expressed in healthy conditions.

Such a view assumes a dimensional approach to neuropsychopathology, which considers the full range of variation, from normal to abnormal functioning, where both extremes of a dimension may be considered ‘abnormal’ [515]. For example, when anxiety increases beyond an optimal level, (a) one’s perceptual field narrows and attention to task-relevant cues fails [693], (b) the ability to store and retrieve task-relevant information deteriorates [694,695], and (c) distracting and task-irrelevant thoughts increase [696]. In its extreme form, it is associated with anxiety disorders. On the opposite end of the dimension, a complete lack of anxiety may be associated with aggressive or psychopathic behavior. Consider another example from [697] (p. 116): “[…], persons that fall in the high-end of sociability dimension, have positive emotional and high approach-motivation tendencies are also high in the novelty-seeking trait and low in the harm-avoidance trait (Davies, 2012; Eysenck, 1990). Such people are likely to find extravagance, novelty, and excitement motivating but will be relatively insensitive to the feelings of others, punishment for breaching rules, or the possibility of failure (Cavanagh, 2005)” (references in this citation can be found in [697]). In the extreme, it may be associated with mania, narcissism, psychopathy, substance abuse, excessive venturesomeness (excitement-seeking), risk-taking, grandiosity, exhibitionism, and manipulativeness [635]. On the opposite end of the dimension, lack of sociability may be associated with emotional tension or depression.

Considering a dimensional approach to neuropsychopathology, the following functional continuum, through normality (health) and mental disorders (psychopathology), can be proposed (Figure 5).

Here, the focus is on the expression of the characteristics of qEEG phenotypes (neural mechanism) over a continuous range. A corresponding mental characteristic that is linked to a particular qEEG phenotype (or its features) would also show a range of expression. Importantly, ‘functionality’ is the degree to which the individual is able to function, given various degrees of expression of the mental and associated qEEG phenotype’s characteristics [519,582]. Hence, the degree of functionality associated with the expression of the mental characteristic corresponds to the degree of the expression of a given characteristic(s) of the qEEG phenotype. The expression of the characteristics of qEEG phenotypes in the middle range (green, Figure 5) is expected to correspond to optimal neurophysiological and mental functioning. Indeed, it has been shown that optimal characteristics of the qEEG phenotype are associated with optimal cognition, good personality structure, and overall well-being [506,582,612,677,698,699]. Similarly, it has been shown that (a) moderate values of self-esteem are optimal for psychological functioning, where both high and low values lead to dysfunction [700], (b) moderate anxiety is associated with the best performance [701,702], and (c) moderate vigilance/arousal is likewise associated with optimal performance [703,704,705]. For more examples, see [582].

Expression of qEEG phenotype characteristics outside the optimal range is expected to be associated with changes in the functionality of mental characteristics that are ‘normal’ but fall in a sub-optimal zone—ordinary functioning (in blue, Figure 5), representing a state of increased risk for mental dysfunction. The inefficient functioning range (orange area, Figure 5) on either end of the spectrum of the ‘normal’ range can be seen as a transition zone towards probable psychopathology. It can be defined as a range where personal features and behaviors cause difficulties but are nonetheless adaptable enough to avoid being classified as disordered [706]. Further extremes in the expression of the qEEG phenotype characteristics (purple area, Figure 5) are associated with states that are dysfunctional and may occur in a pathological condition as it leads to impairments in behavior. Extremes are always maladaptive, with no exception. Indeed, studies on healthy people showed that extreme neural values, even in otherwise normal individuals, impair their functionality and that the amelioration of extreme readings leads to improved functioning, whereas the middle values favor optimal functioning [582].

From an information-processing point of view, neurocognitive capacities follow a normal distribution in the human population, varying from being extremely efficient to extremely inefficient, depending on the underlying genetic makeup, trait predispositions, learning history, life events and style, and state variables. Since neurocognitive processes form and shape individual behaviors, sub-optimal neurocognitive capacity can translate into behavioral symptoms. The frequency and intensity of these symptoms will be continuously distributed in the general population—with most below (or above) the threshold for clinical significance. Here, normal function becomes impaired when symptoms escalate, making it difficult to maintain normal relationships and occupational productivity.

In the proposed continuum, the optimal range has special importance and, thus, demands more explanation. The optimal range of the qEEG phenotype characteristics represents certain ‘idealized’ characteristics displayed by the majority of healthy subjects within the same age group without current or past neurologic or mental health complaints or other illnesses or traumas that might be associated with brain dysfunction and without a family history of neurologic and psychiatric diseases [455,707,708]. It is assumed that the optimal range of the qEEG phenotype characteristics is a zone in which an individual has a higher probability of achieving an optimal performance compared to a qEEG phenotype with characteristics that fall outside the optimal zone (see above). Optimal functioning ensures the efficient recruitment of resources, the adequate mobilization of energy, and the utilization of skills in proportion to the task at hand. In this context, optimal functioning includes:

• Successful performance;
• Maximum efficiency;
• Minimal cost;
• Temporal adequacy.

Here, optimality is understood in terms of a trade-off that balances the accuracy or benefit (result) of performance against an appropriate cost (the time, energy, memory, and computational resources). It is consistent with (a) an evolutionary trade-off approach, according to which a biological system (here, the brain) maximizes a specific fitness function that results in an optimal phenotype [709] that reflects trade-offs among traits to optimize fitness and (b) the principle of optimality in biology (formulated by Liberman et al. [710]) that relates to the establishment of spatiotemporal patterns that are maximally predictable and can hold the living state for a prolonged time [711]. In this context, optimal functioning can be considered a complex trait-like construct—an optimal phenotype.

Since there are neurophysiological limits as to how high or low qEEG characteristics are able to go, only the optimal range offers enough ‘operating room’ to increase or decrease a given qEEG characteristic (deviation to both sides away from the optimal range) when faced with a task and in dependence on the internal or external environment; thus, the optimal (mid) range is by definition maximally adaptive.

In between optimality and dysfunction, there lies a whole raft of conditions associated with varying kinds and degrees of functionality. Therefore, health and diseases can be conceptualized non-categorically as the heterogeneity of phenotypes that exist along a continuum between optimality and dysfunction. In this context, health can be defined as the optimal, flexible, and successful interaction of an organism with a complex and changing environment. Thus, brain health can be defined as the development and preservation of optimal brain integrity and neural network functioning for a given age [712] and not merely the absence of disease. Disease can be defined as the rigid persistence of context-insensitive operational set-points. Here, healthiness is not a difference in kind from the clinical population but merely a matter of dimensional divergence distance.

Deviation of qEEG characteristics from the optimal range does not necessarily reflect gross abnormality or a pathologic process. Deviation means that the brain functions less efficiently, thus spending more energy and resources to achieve a needed function, operation, or task (Figure 5) [153,449,707] or having increasing difficulties in finding resources available for compensation in order to preserve core cognitive functions [713]. The latter places more stress on available resources. Inefficiency in handling the normal challenges of daily life, as well as the adverse physiological consequences, can be genetically or developmentally programmed or mediated by lifestyle choices. Additionally, depending on the degree, deviations may limit the range of the cognitive, emotional, and behavioral repertoire accessible to the individual.

If the compensatory mechanisms of the brain are intact, then a small deviation is unlikely to be pathologically significant and lies within the bounds of normal individual variability (despite being sub-optimal). Indeed, the brain may have many built-in alternative solutions, which are revealed and manifested via neurological cases of resilience. Consider, for example, the mechanisms of dynamic rerouting of information streams in the brain that are necessary to maintain functional integrity in the face of structural network failure. However, the pathogenic significance of qEEG deviations from the optimal range increases when compensatory mechanisms of the brain are decreased or exhausted. This is usually associated with strong and very strong deviations from the optimal range (Figure 5). Indeed, since neuro-cognitive processes form and shape individual behaviors, sub-optimal neurocognitive capacity can translate into behavioral symptoms and particular dysfunctions or abnormalities that may already be associated with neurological, developmental, and psychiatric disorders when the clinical significance threshold is crossed [455,707,714].

An important consideration regarding dimensionality is that (a) the relationship between increasing disruptions in functional mechanisms reflected in the characteristics of qEEG phenotypes and the severity of symptoms may be non-linear, with set-points that mark a transition to more severe pathology [515], and (b) a particular characteristic of the qEEG phenotype may have a different meaning and reflect a different underlying pathogenetic process as a function of the overall context within which it emerges (‘contextual functionality’) [715].

Since there is abundant scientific evidence (see above) that normative qEEG values are the result of brain electrical rhythm autoregulation by a complex homeostatic system [248,449] that displays a characteristic metastability around certain homeostatic levels, then deviations in characteristics of qEEG phenotypes may reflect a departure from homeostatic regulation that is a new stable state of altered brain activity [716] and can be a pathological ‘set point’ if this deviation is chronic and outside the normal range. This adaptation is referred to as ‘allostasis’ and is defined as ‘homeostasis through change’ [716,717], where normal mechanisms for homeostatic regulation have spun out of the physiological range, which can lead to a chronic condition of heightened vulnerability to pathology. The price the brain pays to adapt to adverse psychosocial or physical situations (environments), which is related to how inefficient the response is and how many challenges the brain experiences for a sustained period of time, is defined as the allostatic load [718,719]. Allostatic load is mediated by factors such as genes and early development, as well as learned behaviors, reflecting lifestyle choices of diet, exercise, drug use, and so on [716], all of which are reflected in the qEEG phenotype (see above), which is a long lasting ‘brain signature’.

Considering (a) the dimensional approach to neuropsychopathology (see above), (b) the functional continuum of normality–pathology (see above), and (c) the fact that the same symptoms (though with different expressions) are present in the majority of psychiatric disorders [720] due to the transdiagnostic nature of primary domains of functioning (traits) and their associated qEEG phenotypes’ characteristics (see above), neuropsychopathology can be conceptualized as the degree of deviations across every domain (assessed by qEEG phenotypes’ characteristics) and their disposition (stabilized relative to each other’s positions on the normality–pathology continuum), where the domain with the largest deviation (or combination of several deviant domains) may represent the leading pathogenic marker or risk factor that manifests as either in the permanent mental constitution or a heightened vulnerability to mental disorder. This is in line with animal models of neuropsychiatric disorders, where domain interplay was demonstrated [721].

In this context, developing a new diagnostic model should involve building a brain-based conceptualization of psychiatric illnesses [722], where neuropsychopathology is quantified through qEEG phenotypes that are defined by the degree of deviation from the normative (age- and sex-matched) data and the disposition of qEEG characteristics and the associated transdiagnostic primary domains of functioning (traits) that are placed in the functional continuum of normality–pathology. Here, the deviation from the normative values provides a probabilistic measure of the likelihood that the individual’s electrical activity reflects abnormal brain functioning. Thus, concepts of ‘normal’ or ‘abnormal’ can be redefined based on the probability of qEEG characteristics associated with transdiagnostic primary domains of functioning relative to normative data, where ‘abnormality’ is defined statistically as improbable values exceeding those expected randomly [153]. It is assumed that the more statistically unusual the observation, the more likely it is that the underlying brain system is clinically abnormal. This provides a quantitative and objective criterion for the severity of brain dysfunction in an individual. From this viewpoint, the distinction between ‘normality’ and ‘abnormality’ depends upon the threshold value established in any particular set of transdiagnostic primary domains of functioning [153] along the normality–pathology continuum. The proposed framework of brain profiling diagnosis provides valuable insights into the etiology of psychiatric diseases and allows researchers to understand how usually adaptive processes may become part of vicious circles that result in pathology; it also has important implications for identifying at-risk individuals, initiating early prevention, and tailoring treatments, thereby providing more cost-effective and efficient diagnostic tools.

5. Summary and Concluding Remarks

There is mounting evidence that the problems experienced by the current paradigm of psychiatric diagnoses are due to a lack of brain-related etiological knowledge about neuropsychopathology [2,276,722,723,724,725,726,727,728,729]. An important step towards solving these problems is through formulating a theoretical–conceptual framework that relates clinical manifestations of mental disorders to individual history, lifestyles, and brain dynamics in a comprehensive unifying manner.

Here, such a framework is proposed, and it is based on knowledge accumulated over many decades (including recent developments) from systemic, cognitive, and clinical neuroscience, where mathematically and statistically derived analyses (neurometrics) of qEEGs provide quantitative information about brain activity that is related to anatomical and functional integrity, developmental maturation, and the mediation of sensory, perceptual, and cognitive processes as well as clinical manifestations of mental disorders. The resulting patterns of qEEG characteristics—qEEG phenotypes—are placed on the functional continuum of health and pathology of primary domains of functioning (traits), which have a dimensional nature and are transdiagnostic. In this context, the typicality or atypicality of each qEEG phenotype is quantified by the disposition of qEEG characteristics on the distribution of phenotypic parameters along the health–pathology continuum, where more atypical phenotypes have more extreme positions. This theoretical–conceptual framework has neurobiological/etiological relevance and uses dimensionally parameterized physiological characteristics, mental functions, and neuropsychopathology; we believe it provides better clinical diagnostic and prognostic utility of qEEGs.

However, more work is needed to arrive at (a) the optimal number and types of primary domains of functioning (traits), (b) the optimal repertoire of qEEG phenotypes, (c) the ‘library’ of qEEG characteristics within the qEEG phenotypes associated with the primary domains of functioning, and (d) adequate thresholds in the values of qEEG characteristics (age- and sex-stratified) that signal the transition from one functioning condition to another along the health–pathology continuum.