Systems and methods for improving sleep

Abstract

A method for transplanting the sleep state of a first subject (donor) into a second subject (recipient), the method comprising: capturing the sleep state of the first subject as represented by a brain activity pattern; and transplanting the sleep state of the first subject into the second subject by inducing the brain activity pattern in the second subject.

Description

Technical Field

[0001] This invention relates generally to the field of neural modulation and neural enhancement (NE), and more specifically to systems and methods for improving sleep states in humans or animals. Background Technology

[0002] Each reference and document cited in this paper is explicitly incorporated into this paper in its entirety by reference for all purposes.

[0003] Biological Temporal Changes: Almost everything in biology changes over time. These changes occur on many different timescales, which vary greatly. For example, there are evolutionary changes that affect entire populations over time, rather than individual organisms. Evolutionary changes are generally slower than the human timescale, which spans many years (typically a human lifetime). Faster changes in the timing and duration of biological activities in living organisms occur, for example, in many important biological processes of daily life: in humans and animals, these changes occur in, for example, eating, sleeping, mating, hibernation, migration, and cell regeneration. Other rapid changes can involve the transmission of neural signals, for example, through synapses such as the calyx ofheld, particularly large synapses in the auditory central nervous system of mammals, which can reach transmission frequencies up to 50 Hz. With recruitment modulation, the effective frequencies can be even higher. Individual nerve impulses can reach speeds up to 100 meters per second (0.06 miles per second) (Kraus, David, *Concepts in Modern Biology*, New York: Globe Book Company, 1969: 170). Myelination of axons can increase transmission speed through the segmentation membrane depolarization process.

[0004] Many of these changes over time are repetitive or rhythmic and are described as having a certain frequency or oscillation. The field of chronobiology studies this periodicity (cycle) in living organisms, for example, and their adaptation to rhythms related to the sun and moon, for example [DeCoursey et al., (2003)]. These cycles are also known as biological rhythms. In some cases, the related terms time-varying omics and chronomics have been used to describe the molecular mechanisms involved in chronobiological phenomena or more quantitative aspects of chronobiology, especially when comparing cycles between organisms is required. Chronobiological research includes, but is not limited to, comparative anatomy, physiology, genetics, molecular biology, and the behavior of organisms within the mechanics of biological rhythms [DeCoursey et al., (2003)]. Other aspects include epigenetics, development, reproduction, ecology, and evolution.

[0005] The most important rhythm in chronobiology is the circadian rhythm, a roughly 24-hour cycle manifested through physiological processes in all these organisms. It is regulated by the diurnal clock. The circadian rhythm can be further broken down into regular cycles throughout the 24-hour day [Nelson RJ., 2005, An Introduction to Behavioral Endocrinology, Sinauer Associates, Inc., Massachusetts, p. 587]. All animals can be classified according to their activity cycles: diurnal, which describes organisms active during the day; nocturnal, which describes organisms active at night; and twilight, which describes animals active during dusk and twilight (e.g., white-tailed deer, some bats).

[0006] Although circadian rhythms are defined as being regulated by endogenous processes, other biological cycles can be regulated by exogenous signals. In some cases, multitrophic systems can exhibit rhythms driven by the circadian clock of one of their members (which may also be influenced by or reset by external factors).

[0007] Many other important cycles have also been studied, including: infrared rhythms with periods longer than one day. Examples include: annual or yearly cycles that govern the migration or reproductive cycles of many plants and animals, or the human menstrual cycle; sub-diurnal rhythms, which are cycles shorter than 24 hours, such as the 90-minute REM cycle, the 4-hour nasal cycle, or the 3-hour growth hormone production cycle; tidal rhythms commonly observed in marine life, which follow a roughly 12.4-hour transition from high tide to low tide and back; lunar rhythms, which follow the lunar month (29.5 days), and are associated with, for example, marine life because tidal levels are modulated across the lunar cycle; and gene oscillations – some genes are expressed more during certain hours of the day than at other times.

[0008] Within each cycle, the more active period of the process is called the peak phase [Refinetti, Roberto (2006), Circadian Physiology, CRC Press / Taylor & Francis Group, ISBN 0-8493-2233-2, Concise Abstract]. When the process is less active, the cycle is in its trough or valley phase. The specific moment of highest activity is the peak or maximum; the lowest point is the nadir. How high (or low) the process reaches is measured by amplitude.

[0009] Sleep cycles and sub-circadian rhythms: The normal cycle of sleep and wakefulness indicates that at specific times, one nervous system is activated while others are deactivated. Therefore, understanding the various stages of sleep is crucial to the neurobiology of sleep. In 1953, Nathaniel Kleitman and Eugene Aserinksy, using electroencephalogram (EEG) recordings from normal human subjects, demonstrated that sleep comprises different stages that occur in a characteristic sequence.

[0010] Humans fall asleep in successive phases, one after another, within the first hour after going to bed. These characteristic phases are primarily defined by electroencephalogram (EEG) standards. Initially, during "drowsy" periods, the EEG spectrum shifts towards lower values, and the amplitude of cortical waves increases slightly. This drowsy phase is called stage I sleep, eventually succumbing to light sleep or stage II sleep, characterized by a further decrease in the frequency of EEG waves and an increase in their amplitude, accompanied by intermittent clusters of high-frequency spikes known as sleep spindles. Sleep spindles are periodic bursts of activity at approximately 10–12 Hz, typically lasting 1 or 2 seconds, and are generated by interactions between thalamic and cortical neurons. In stage III sleep, which represents moderate to deep sleep, the number of spindles decreases, while the amplitude of low-frequency waves still increases more. In the deepest sleep (stage IV sleep), the main EEG activity consists of low-frequency (1–4 Hz), high-amplitude fluctuations known as delta waves (the characteristic slow waves of this phase of sleep). The entire sequence from drowsiness to deep stage IV sleep typically takes about an hour.

[0011] These four SS (stages of sleep) are called non-rapid eye movement (non-REM) sleep, and their most prominent feature is slow-wave (stage IV) sleep. Awakening people from slow-wave sleep is the most difficult, and therefore, it is considered the deepest sleep stage. However, after a period of slow-wave sleep, EEG recordings show that the sleep stages reverse to reach a rather different state, known as rapid eye movement or REM sleep. During REM sleep, EEG recordings are very similar to those of the waking state. This pattern is peculiar: the dreamer's brain becomes highly active, while the body's muscles are weak, and the breathing and heart rates become erratic. After about 10 minutes of REM sleep, the brain typically cycles back to non-REM SS. Slow-wave sleep usually reappears in the second cycle of this continuous cycle rather than during the rest of the night. On average, four additional stages of REM sleep occur, each cycle being longer than the previous one. The typical 8 hours of sleep experienced each night actually includes several cycles alternating between non-REM and REM sleep, with the brain being highly active during most of this assumed period of restful sleep. For reasons unknown, the amount of REM sleep per day decreased from about 8 hours at birth to 2 hours at age 20, and only about 45 minutes at age 70.

[0012] Falling asleep: When falling asleep, a series of highly coordinated events puts the brain into sleep during the aforementioned stages. Technically, sleep begins in brain regions that produce slow-wave sleep (SWS). Two cell groups—the ventral preoptic nucleus in the hypothalamus and the parafacial region in the brainstem—have been shown to be involved in promoting SWS. When these cells are activated, they trigger loss of consciousness. After SWS, REM sleep begins. Despite increasing understanding of its biochemistry and neurobiology, the purpose of REM sleep remains a biological mystery. A small group of cells in the brainstem called cyanobacterial nuclei has been shown to control REM sleep. When these cells become injured or diseased, people do not experience the muscle weakness associated with REM sleep, which can lead to REM sleep behavior disorder, a severe condition in which patients experience intense dreaming.

[0013] Neural relevance: The neural relevance of a sleep state is a state conceived by electroneurobiology or some biophysical subsystem of the brain, the existence of which is necessarily and regularly associated with a particular sleep state. All properties of thought, including consciousness, emotion, and desire, are considered to have direct neural relevance. For purposes of study, the neural relevance of a sleep state can be defined as the minimal set of neuronal oscillations corresponding to a given sleep state (SS). Neuroscientists use empirical methods to discover the neural relevance of SS.

[0014] Mental state: A mental state is the state of mind experienced by a subject. Some mental states are pure and defined, while humans are capable of complex states consisting of combinations of mental representations, which may possess contradictory characteristics in their pure form. Several paradigmatic thought states exist that subjects experience: love, hate, pleasure, fear, and pain. Mental states can also include wakefulness, sleep, flow (or being in a “zone”), and emotion (mental state). A mental state is a hypothetical state corresponding to thought and feeling and is composed of an aggregation of mental representations. Mental states are related to emotion, although they are also related to cognitive processes. Because mental states are complex and potentially have inconsistent properties, it is difficult or impossible to clearly interpret them through external analysis (other than self-report). However, some studies have reported that certain properties of mental states or thought processes can actually be determined by passive monitoring, such as EEG or fMRI, which have some degree of statistical reliability. In most studies, characterizing the mental state is the endpoint, and the original signal is replaced after statistical classification or semantic labeling. The remaining signal energy is considered noise. Current technology does not allow for precise abstract coding or representation of the entire range of mental states based on neural correlations of mental states.

[0015] Brain: The brain is a key part of the central nervous system, enclosed within the skull. In humans and more commonly, mammals, the brain controls both voluntary and cognitive processes. The brain (and to a lesser extent, the spinal cord) controls all volitional functions of the body and interprets information from the outside world. The brain controls intelligence, memory, emotion, speech, thinking, movement, and creativity. The central nervous system also controls voluntary functions and many homeostatic and reflexive actions, such as breathing and heart rhythm. The human brain consists of the cerebrum, cerebellum, and brainstem. The brainstem contains the midbrain, pons, and medulla oblongata. Sometimes it also contains the diencephalon, the caudal part of the forebrain.

[0016] The brain is composed of neurons, glial cells (also known as glial cells), and other cell types in a network of connections that integrate sensory input, control movement, facilitate learning and memory, activate and express emotions, and control all other behavioral and cognitive functions. Neurons primarily communicate via electrochemical impulses, which transmit signals between connective cells within brain regions and between brain regions. Therefore, the desire to non-invasively capture and replicate neural activity associated with cognitive states has become a subject of interest for behavioral and cognitive neuroscientists.

[0017] Technological advancements now allow for the non-invasive recording of vast amounts of information from the brain at multiple spatial and temporal scales. Examples include electroencephalography (“EEG”) data using multichannel electrode arrays placed on the scalp or inside the brain, magnetoencephalography (“MEG”), magnetic resonance imaging (“MRI”), functional data using functional magnetic resonance imaging (“fMRI”), positron emission tomography (“PET”), near-infrared spectroscopy (“NIRS”), single-photon emission computed tomography (“SPECT”), and more.

[0018] Non-invasive neuromodulation techniques have also been developed that can modulate patterns of neural activity and thereby induce changes in behavior, cognitive states, perception, and motor output. Integrating non-invasive measurements and neuromodulation techniques to identify and transfer brain states from neural activity is highly valuable for clinical therapies, such as brain stimulation and related techniques commonly used to treat cognitive impairments.

[0019] The brainstem provides major motor and sensory innervation to the face and neck via cranial nerves. Of the twelve pairs of cranial nerves, ten originate in the brainstem. This is an extremely important part of the brain because the neural connections from the main part of the brain to the motor and sensory systems of the rest of the body pass through the brainstem. This includes the corticospinal tract (motor), the posterior column-medial lemniscus pathway (fine touch, vibration sensation, and proprioception), and the spinothalamic tract (pain, temperature, itch, and coarse touch). The brainstem also plays a crucial role in the regulation of cardiac and respiratory functions. It also regulates the central nervous system and is key in maintaining consciousness and modulating the sleep cycle. The brainstem has many fundamental functions, including controlling heart rate, breathing, sleep, and eating.

[0020] The skull functions to protect the delicate brain tissue from damage. It consists of eight fused bones: the frontal bone, two parietal bones, two temporal bones, the sphenoid bone, the occipital bone, and the ethmoid bone. The face is formed by 14 pairs of bones, including the maxilla, zygomatic bone, nasal bone, palatine bone, lacrimal bone, inferior nasal bone, mandible, and vomer. The bony skull is separated from the brain by the dura mater (a membranous organ) which contains cerebrospinal fluid. Typically, the cortical surface of the brain is not subjected to localized pressure from the skull. Therefore, the skull creates an obstacle to electrical access to brain function, and in healthy humans, breaching the dura mater to access the brain is highly discouraged. The result is a filtering of electrical readings of brain activity through the dura mater, cerebrospinal fluid, skull, scalp, and skin appendages (e.g., hair), leading to a loss of potential spatial resolution and amplitude of signals emanating from the brain. Although the magnetic fields generated by brain electrical activity are accessible, the spatial resolution using feasible sensors is also limited.

[0021] The cerebrum is the largest part of the brain and consists of a left hemisphere and a right hemisphere. It performs higher functions, such as interpreting input from the senses, as well as speech, reasoning, emotion, learning, and fine motor control. The surface of the cerebrum has a folded appearance and is called the cortex. The human cortex contains about 70% nerve cells (neurons) and has a gray appearance (gray matter). Beneath the cortex are long connecting fibers between neurons that make up white matter, called axons.

[0022] The cerebellum is located behind the cerebrum and brainstem. It coordinates muscle movement and helps maintain balance and posture. The cerebellum may also be involved in some cognitive functions, such as attention and language, as well as regulating fear and pleasure responses. There is substantial evidence that the cerebellum plays a crucial role in some types of motor learning. The tasks in which the cerebellum most clearly functions are those requiring fine-tuning of how movements are performed. There is controversy regarding whether learning occurs within the cerebellum itself or whether it merely serves to provide signals that facilitate learning in other brain structures. The cerebellum also plays an important role in sleep and long-term memory formation.

[0023] The brain communicates with the body via the spinal cord and twelve pairs of cranial nerves. Ten of these pairs, which control hearing, eye movement, facial sensation, taste, swallowing, and movement of the muscles in the face, neck, shoulders, and tongue, originate in the brainstem. The cranial nerves for olfaction and vision originate in the cerebrum. Neurons are the basic units of the nervous system, which includes the autonomic nervous system and the central nervous system.

[0024] The right and left hemispheres of the brain are connected by a structure of fibers called the corpus callosum. Each hemisphere controls the opposite side of the body. The right eye sends visual signals to the left hemisphere, and vice versa. However, the right ear sends signals to the right hemisphere, and the left ear sends signals to the left hemisphere. Not all functions of the hemispheres are shared. For example, speech is processed only in the left hemisphere. The cerebral hemispheres have different structures that divide the brain into lobes. Each hemisphere has four lobes: the frontal lobe, temporal lobe, parietal lobe, and occipital lobe. There are very complex relationships between the lobes of the brain and between the left and right hemispheres: the frontal lobe controls judgment, planning, problem-solving, behavior, emotion, personality, speech, self-awareness, concentration, intelligence, and body movement; the temporal lobe controls language comprehension, memory, organization, and hearing; the parietal lobe controls: the interpretation of language; input from vision, hearing, sensation, and movement; temperature, pain, tactile signals, memory, spatial and visual perception; and the occipital lobe interprets visual input (movement, light, color).

[0025] Brain structures and specific areas within brain structures include, but are not limited to, rhomboid structures (e.g., terminal brain structures such as the medulla oblongata, medullary pyramids, olivary body, inferior olivary nucleus, respiratory center, cuneate nucleus, gracile nucleus, intercalation nucleus, medullary and cranial nerve nuclei, inferior salivary nucleus, nucleus ambiguus, dorsal nucleus of the vagus nerve, hypoglossal nucleus, solitary tract nucleus, etc.), posterior brain structures (e.g., pons, pontine cranial nerve nuclei, main nucleus of the trigeminal sensory nucleus (V) or pontine nucleus, motor nucleus of the trigeminal nerve (V), external rotator nucleus (VI), facial nerve nucleus (VII), vestibulocochlear nucleus (vestibular nucleus and cochlear nucleus) (VIII), superior salivary nucleus, pontine tegmentum, respiratory center, respiratory regulation center, long inhalation center, pontine micturition center (Barrington's nucleus)). The structures of the midbrain include: locus coeruleus, peduncular nucleus, dorsolateral tegmental nucleus, tegmental pontine reticular nucleus, superior olivary complex, midline parapontine reticular formation, cerebellar peduncle, superior cerebellar peduncle, middle cerebellar peduncle, inferior cerebellar peduncle, fourth ventricle, cerebellum, cerebellar vermis, cerebellar hemispheres, anterior lobe, posterior lobe, flocculonodular lobe, cerebellar nucleus, parietal nucleus, interpositional nucleus, globular nucleus, tethered nucleus, dentate nucleus, etc.; midbrain structures (e.g., tectum, tetradymature, inferior colliculus, superior colliculus, anterior tectum, tegmentum, periaqueductal gray matter, parabrachial region, medial parabrachial nucleus, lateral parabrachial nucleus, inferior parabrachial nucleus (Kolliker-Fuse nucleus), rostral space nucleus of medial longitudinal fasciculus, midbrain reticular formation, dorsal suture nucleus, red nucleus, ventral tegmental region, substantia nigra, pars compacta, reticular region, interpeduncular nucleus, cerebral peduncle, cerebral peduncle. (cerebri), midbrain cranial nerve nuclei, oculomotor nucleus (III), trochlear nucleus (IV), cerebral aqueduct (cerebral aqueduct, cerebral aqueduct, etc.), forebrain structures (e.g., diencephalon), epithalamic structures (e.g., pineal gland, habenula, stria medullaris, thalamus, etc.), third ventricle, thalamic structures (e.g., anterior nuclei, anterior ventral nuclei, anterior dorsal nuclei, anterior medial nuclei, medial nuclei, medial dorsal nuclei, midline nuclei, parazonal nuclei, connective nuclei, rhomboid nuclei, intralaminar nuclei, central medial nucleus, parafascicular nuclei, paracentral nuclei, central lateral nucleus, central medial nucleus). (nucleus), lateral nuclei, lateral dorsal nucleus, lateral posterior nucleus, occipital, ventral nuclei, ventral anterior nucleus, ventral lateral nucleus, ventral posterior nucleus, ventral posterolateral nucleus, posterior thalamus, medial geniculate body, lateral geniculate body, thalamic reticular nucleus, etc.), hypothalamic structures (e.g.,Anterior, middle and lateral regions, parts of the preoptic region, middle preoptic nucleus, suprachiasmatic nucleus, paraventricular nucleus, supraoptic nucleus (main), anterior hypothalamic nucleus, lateral region, parts of the preoptic region, lateral preoptic nucleus, anterior part of the lateral nucleus, parts of the supraoptic nucleus, other nuclei of the preoptic region, medial preoptic nucleus, periventricular preoptic nucleus, tubercle, middle and lateral region, dorsomedial hypothalamic nucleus, ventromedial nucleus, arcuate nucleus, lateral region, tuberous part of the lateral nucleus, lateral eminence nucleus, posteromedial region, mammary nucleus (part of the mammary body), optic chiasm, subfornular organs, periventricular nucleus, pituitary stalk, tuber cinereum, tuberous nucleus, tuberous-mammary nucleus, tuberous region, mammary nucleus, etc.), subthalamic structures (e.g., thalamic nuclei, zona indeterminate zone, etc.), pituitary structures (e.g., neurohypophysis, middle part (middle lobe of pituitary), adenohypophysis, etc.), telencephalon structures, white matter structures (e.g., corona radiata, internal capsule, external capsule, outermost capsule, arcuate fasciculus, uncinate fasciculus, perforator fasciculus, etc.). Subcortical structures (e.g., hippocampal structures (medial temporal lobe), dentate gyrus, hippocampal angle (CA field), hippocampal angle 1, hippocampal angle 2, hippocampal angle 3, hippocampal angle 4, amygdala (limbic system) (limbic lobe), central nucleus (autonomic nervous system), medial nucleus (auxiliary olfactory system), cortical nuclei and basomedial nuclei (primary olfactory system), lateral [disambiguation] and basal nuclei (frontotemporal cortical system), claustrum, basal ganglia, striatum, dorsal striatum (also neostriatum), putamen, caudate nucleus, ventral striatum, nucleus accumbens, olfactory tubercle, globus pallidus (forming the lentiform nucleus with the putamen), hypothalamic nucleus, basal forebrain, anterior perforated substance, innocuous substance, basal nucleus, Broca's oblique band, medial septal nucleus, etc.), olfactory brain structures (e.g., olfactory bulb, piriform cortex, anterior olfactory nucleus, olfactory tract, anterior commissure, uncus, etc.), cerebral cortical structures (e.g.,Anterior lobe, cortex, primary motor cortex (precentral gyrus M1), supplementary motor cortex, premotor cortex, prefrontal cortex, gyri, superior frontal gyrus, middle frontal gyrus, inferior frontal gyrus, Brodmann areas: 4, 6, 8, 9, 10, 11, 12, 24, 25, 32, 33, 44, 45, 46, 47, parietal lobe, cortex, primary somatosensory cortex (S1), secondary somatosensory cortex (S2), posterior parietal cortex, gyri, posterior central gyrus (primary somatosensory area), others, precuneus, Brodmann areas 1, 2, 3 (primary somatosensory areas); 5, 7, 23, 26, 29, 31, 39, 40, occipital lobe, cortex, primary visual cortex (V1), V2, V3, V4, V5 / MT, gyri, lateral occipital gyrus, cuneus, Brodmann areas 17 (V1, primary visual cortex, secondary somatosensory area ... Auditory cortex); 18, 19, temporal lobe, cortex, primary auditory cortex (A1), secondary auditory cortex (A2), inferior temporal cortex, posterior inferior temporal cortex, superior temporal gyrus, middle temporal gyrus, inferior temporal gyrus, entorhinal cortex, perinasal cortex, parahippocampal gyrus, fusiform gyrus, Broadman area: 9, 20, 21, 22, 27, 34, 35, 36, 37, 38, 41, 42, medial superior temporal area (MST), insular cortex, cingulate cortex, anterior cingulate cortex, posterior cingulate cortex, posterior cortical cortex, gray area, infragenual area 25, Broadman area 23, 24, 26, 29, 30 (posterior cortical area), 31, 32, etc.

[0026] Neuron: A neuron is an electrically actuated cell that receives, processes, and transmits information, and based on that information, sends signals to other neurons, muscles, or glands via electrical and chemical signals. These signals between neurons occur through specialized connections called synapses. Neurons can connect to each other to form neural networks. The fundamental purpose of a neuron is to receive incoming information and, based on that information, send signals to other neurons, muscles, or glands. Neurons are designed to send signals rapidly over physiologically long distances. They do this using electrical signals called nerve impulses or action potentials. When a nerve impulse reaches the neuronal terminus, it triggers the release of chemicals or neurotransmitters. Neurotransmitters travel rapidly across short gaps (synapses) between cells and act as signals to neighboring cells. See www.biologyreference.com / Mo-Nu / Neuron.html#ixzz5AVxCuM5a.

[0027] Neurons can receive thousands of inputs from other neurons via synapses. Synaptic integration is the mechanism by which neurons integrate these inputs before generating nerve impulses or action potentials. The ability of synaptic inputs to influence neuronal outputs is determined by several factors: the magnitude, shape, and relative timing of the potentials generated by the synaptic inputs; the geometry of the target neuron; the physical location of the synaptic inputs within that structure; and the expression of voltage-gated channels in different regions of the neuronal membrane.

[0028] Neurons receive and send information from many other cells at specialized junctions called synapses. Synaptic integration is the computational process by which individual neurons process their synaptic inputs and convert those inputs into output signals. Synaptic potentials occur when neurotransmitters bind to the dendritic membrane and open ligand-operated channels within it, allowing ions to move to and from the cell according to their electrochemical gradients. Synaptic potentials can be excitatory or inhibitory, depending on the direction and charge of ion movement. Action potentials occur when the total synaptic input to a neuron reaches a depolarization threshold level and triggers the re-opening of voltage-gated ion channels. Synaptic potentials are typically brief and small in amplitude, thus usually requiring the sum of temporal inputs (temporal summation) or the sum of inputs from multiple synapses (spatial summation) to reach the action potential activation threshold.

[0029] There are two types of synapses: electrical synapses and chemical synapses. An electrical synapse is a direct electrical coupling between two cells mediated by a gap junction, a pore constructed of connective proteins. Essentially, an electrical synapse results in the transmission of a gradient potential (which can be depolarized or hyperpolarized) between the two cells. Electrical synapses are very rapid (without synaptic delay). They are passive processes, where the signal may degrade with distance and may not produce a depolarization large enough to initiate an action potential in the postsynaptic cell. Electrical synapses are bidirectional; that is, the postsynaptic cell can actually send messages to the presynaptic cell.

[0030] A chemical synapse is a coupling between two cells via neurotransmitters, ligands, or voltage-gated channels and receptors. It is influenced by the concentration and type of ions on either side of the membrane. Among neurotransmitters, glutamate, sodium, potassium, and calcium are positively charged. GABA and chloride are negatively charged. Neurotransmitter junctions provide opportunities for pharmacological intervention, and many different drugs, including illicit drugs, act at synapses.

[0031] Excitatory postsynaptic potentials (EPSPs) are postsynaptic potentials that make postsynaptic neurons more likely to generate action potentials. Hyperpolarization of the postsynaptic neuron membrane is caused by the binding of inhibitory neurotransmitters from the presynaptic cell to the postsynaptic receptor. This makes it more difficult for postsynaptic neurons to generate action potentials. Depolarization of the postsynaptic neuron membrane is caused by the binding of excitatory neurotransmitters from the presynaptic cell to the postsynaptic receptor. This makes postsynaptic neurons more likely to generate action potentials. For example, in the synapse of a neuron using glutamate as a receptor, the receptor opens an ion channel that allows cations to flow nonselectively. When these glutamate receptors are activated, both Na+ and K+ flow across the postsynaptic membrane. The reverse potential (Erev) of the postsynaptic current is approximately 0 mV. The resting potential of the neuron is approximately -60 mV. The generated EPSP will depolarize the postsynaptic membrane potential, causing it to move towards 0 mV.

[0032] Inhibitory postsynaptic potentials (IPSPs) are a type of synaptic potential that makes postsynaptic neurons less likely to generate action potentials. An example of an inhibitory postsynaptic action potential is a neuronal synapse that uses γ-aminobutyric acid (GABA) as its neurotransmitter. At such synapses, GABA receptors typically open channels that selectively allow the permeation of Cl- ions. When these channels are open, negatively charged chloride ions can flow across the membrane. The resting potential of a postsynaptic neuron is -60 mV, and the action potential threshold is -40 mV. Neurotransmitter release at this synapse will inhibit the postsynaptic cell. Because the ECI is more negative than the action potential threshold, for example, -70 mV, it reduces the likelihood that the postsynaptic cell will generate an action potential.

[0033] Some types of neurotransmitters, such as glutamate, continuously produce EPSP. Other neurotransmitters, such as GABA, continuously produce IPSP. Action potentials last about one millisecond (1 msec). In contrast, EPSP and IPSP can last for 5 to 10 msec. This allows the action of one postsynaptic potential to build upon the next, and so on.

[0034] Membrane leakage, and to a lesser extent, the potential itself, can be influenced by external electric and magnetic fields. These fields can be generated focally, for example, through implanted electrodes, or less specifically, through transcranial stimulation. Transcranial stimulation can be subthreshold or suprathreshold. In the former case, the external stimulation modulates the resting membrane potential, making the nerve more or less excitable. This stimulation can be direct current or alternating current. In the latter case, this will tend to depolarize the neuron in sync with the signal. Suprathreshold stimulation can be painful (at least because the stimulation directly excites painful neurons) and must be pulsed. Because this corresponds to electroconvulsive therapy, suprathreshold transcranial stimulation is rarely used.

[0035] Many neurotransmitters are known, as are pharmacological interventions and therapies that affect these compounds. Typically, major neurotransmitters are small monoamine molecules such as dopamine, adrenaline, noradrenaline, serotonin, GABA, histamine, and acetylcholine. In addition, neurotransmitters also include amino acids, gaseous molecules such as nitric oxide, carbon monoxide, carbon dioxide, and hydrogen sulfide, and peptides. The presence, metabolism, and modulation of these molecules can influence learning and memory. The supply, oxidation, and control of neurotransmitter precursors, as well as other effects on brain chemistry related to learning and memory, can be used to promote memory, learning, and adaptive transfer of learning.

[0036] Neuropeptides and their corresponding receptors are widely distributed throughout the mammalian central nervous system. During learning and memory processes, in addition to structural synaptic remodeling, changes at the molecular and metabolic levels, along with alterations in neurotransmitter and neuropeptide synthesis and release, have been observed. Although cholinergic neurotransmission in the brain is generally considered to play a crucial role in processes related to learning and memory, it is well known that these functions are influenced by a vast number of neuropeptides and non-peptide molecules. Arginine vasopressin (AVP), oxytocin, angiotensin II, insulin, growth factor, serotonin (5-HT), melanin-concentrating hormone, histamine, serotonin, gastrin-releasing peptide (GRP), glucagon-like peptide-1 (GLP-1), cholecystokinin (CCK), dopamine, and corticotropin-releasing factor (CRF) have modulatory effects on learning and memory. Among these peptides, CCK, 5-HT, and CRF play key roles in the modulation of memory processes under stress. CRF is considered a major neuropeptide involved in both physical and emotional stress, and its protective effect during stress may be achieved by activating the hypothalamus-pituitary (HPA). The peptide CCK has been proposed to promote memory processing, and CCK-like immunoreactivity in the hypothalamus has been observed under stress exposure, suggesting that CCK may be involved in the central control of stress response and stress-induced memory dysfunction. On the other hand, 5-HT appears to play a role in behaviors involving high cognitive demands, and stress exposure activates the serotonergic system in various brain regions.

[0037] Mental state: Some studies have reported that certain attributes of mental state or thought processes can actually be determined by passive monitoring, such as EEG, which has a certain degree of statistical reliability. In most studies, characterizing mental state is the endpoint, and after statistical classification or semantic labeling, the original signal is replaced, and the remaining signal energy is treated as noise.

[0038] Neural relevance: The neural relevance of a mental state is a state conceived by electroneurobiology or some biophysical subsystem of the brain, the existence of which is necessarily and regularly associated with a particular mental state. All properties attributed to en.wikipedia.org / wiki / Mind, including consciousness, emotion, and desire, are considered to have direct neural relevance. The neural relevance of a mental state can be viewed as a minimal set of neuronal oscillations corresponding to a given mental state. Neuroscientists use empirical methods to discover the neural relevance of objective mental states.

[0039] Brainwaves: The root of all thought, emotion, and behavior lies in the communication between neurons in the brain—the rhythmic or repetitive neural activity of the central nervous system. Oscillations can be generated by a single neuron or by synchronized electrical impulses from groups of neurons communicating with each other. Interactions between neurons can cause oscillations at frequencies different from the firing frequencies of individual neurons. Synchronized activity of a large number of neurons produces macroscopic oscillations, which can be observed in electroencephalography (EEG). These oscillations are divided into bandwidths used to describe their claimed function or functional relationship. Oscillatory activity in the brain has been observed at various levels of tissue and is considered to play a key role in processing neural information. Numerous experimental studies support the functional role of neural oscillations. However, a unified explanation remains elusive. Neural oscillations and synchronization are associated with many cognitive functions, such as information transfer, perception, motor control, and memory. Electroencephalogram (EEG) signals are relatively easy and safe to acquire, have a long history of analysis, and can be high-dimensional, for example, with up to 128 or 256 individual recording electrodes. Although the information represented in each electrode is not independent of each other, and the signals are highly noisy, much information remains to be obtained from such signals, which have not yet been fully characterized.

[0040] Brain waves have been extensively studied through neural activity generated by large neuronal groups (primarily via EEG). Typically, EEG signals exhibit oscillatory activity in specific frequency bands (regularly synchronized activation of neuronal groups): α (7.5–12.5 Hz), detectable in the occipital lobe during relaxed wakefulness and increasing when the eyes are closed; δ (1–4 Hz), θ (4–8 Hz), β (13–30 Hz), low γ (30–70 Hz), and high γ (70–150 Hz) bands, where faster rhythms, such as γ activity, are associated with cognitive processing. Higher frequencies imply coordinated firing of multiple neuronal groups in parallel, series, or both, as individual neurons do not fire at a rate of 100 Hz. Specific characteristics of neural oscillations are associated with cognitive states, such as consciousness and alertness, and different SS (senses of consciousness).

[0041] The Nyquist Theorem states that the highest frequency that can be accurately represented is half the sampling rate. In practice, the sampling rate should be ten times higher than the highest frequency of the signal (see www.slideshare.net / ertvk / eeg-examples). While EEG signals are largely band-limited, superimposed noise may not be. Furthermore, the EEG signal itself represents components from a large number of independently fired neurons. Therefore, large-bandwidth signal acquisition can be useful.

[0042] A useful analogy is to think of brain waves as musical notes. Like a symphony, higher and lower frequencies are interconnected and coherent through harmonics, especially when it is thought that neurons can coordinate not only based on transitions but also on phase delays. Oscillatory activity is observed at all levels of the tissue, throughout the entire central nervous system. The dominant neuronal oscillation frequencies are associated with corresponding mental states.

[0043] Brain waves have a wide range of functions and vary depending on the type of oscillatory activity. Neural oscillations also play an important role in many neurological disorders.

[0044] In standard EEG recording practice, 19 recording electrodes are evenly placed on the scalp (international 10-20 system). Additionally, one or two reference electrodes (usually placed on the earlobe) and a ground electrode (usually placed on the nose) are required to provide a reference voltage for the amplifier. However, unless supplemented by computer algorithms to reduce the raw EEG data to a manageable form, the additional electrodes may add minimal useful information. When a large number of electrodes are used, the potential at each location can be measured relative to the average of all potentials (common average reference), which generally provides a good estimate of the potential at infinity. When electrode coverage is sparse (possibly less than 64 electrodes), the common average reference is unsuitable. See Paul L. Nunez and Ramesh Srinivasan (2007), “Electroencephalogram,” *Scholarpedia* 2(2):1348, scholarpedia.org / article / Electroencephalogram. Dipole localization algorithms can be used to determine spatial emission patterns in EEG.

[0045] Scalp potentials can be expressed as the volume integral of the dipole moment per unit volume over the entire brain, provided that P(r,t) is defined in general terms rather than in columnar terms. For the important case of dominant cortical sources, the scalp potential can be approximated by the following integral over the cortical volume: Θ,VS(r,t)=∫∫∫ΘG(r,r')·P(r',t)dΘ(r'). If the volume element dΘ(r') is defined according to the cortical columns, the volume integral can be reduced to the integral over the folded cortical surface. The temporal dependence of the scalp potential is a weighted sum of the temporal variations of all dipoles in the brain, but the contribution of deep dipole volumes is usually negligible. The vector Green's function G(r,r') contains all the geometric and conductive information about the head volume conductors, and therefore the weights on the integral. Thus, each scalar component of the Green's function is essentially the inverse electrical distance between each source component and the scalp location. For the ideal case of sources in an infinite medium with constant conductivity, the electrical distance equals the geometric distance. The Green's function explains the finite spatial extent of tissues and their inhomogeneity and anisotropy. The forward problem of EEG consists of selecting a head model to provide G(r,r') and performing an integral over some hypothetical source distribution. The inverse problem consists of finding the best-fitting source distribution P(r,t) using the recorded scalp potential distribution VS(r,t) plus some constraints (typically hypothesized) on P(r,t). Since the inverse problem has no unique solution, any inverse solution depends strictly on the chosen constraints, such as only one or two isolated sources, sources confined to the cortex, or spatial and temporal smoothness criteria. High-resolution EEG uses experimental scalp potentials VS(r,t) to predict the potential VD(r,t) on the dura mater surface (the folded membrane surrounding the cerebral cortex). This can be done using the head model Green's function G(r,r') or by estimating the surface Laplacian using spherical or 3D splines. Both methods typically provide very similar dural potentials VD(r,t); the estimation of the dural potential distribution is uniquely affected by head model, electrode density, and noise issues.

[0046] In an EEG recording system, each electrode is connected to one input of a differential amplifier (one amplifier per electrode pair); a common system reference electrode (or synthetic reference) is connected to the other input of each differential amplifier. These amplifiers amplify the voltage between the active electrode and the reference (typically 1,000–100,000 times or 60–100 dB voltage gain). The amplified signal is then digitized via an analog-to-digital converter after passing through an anti-aliasing filter. In clinical scalp EEG, analog-to-digital sampling typically occurs at 256–512 Hz; in some research applications, sampling rates up to 20 kHz are used. EEG signals can be captured using open-source hardware such as OpenBCI, and the signals can be processed using freely available EEG software such as EEGLAB or the Neurophysiological Biomarkers Toolkit. The amplitude of a typical adult EEG signal is approximately 10–100 μV when measured from the scalp and approximately 10–20 mV when measured from subdural electrodes.

[0047] Delta waves (en.wikipedia.org / wiki / Delta_wave) are frequencies up to 4 Hz. They tend to have the highest amplitude and the slowest wave speed. In NREM, they are commonly found in adults (en.wikipedia.org / wiki / NREM). They are also commonly found in infants. They can occur focally with subcortical lesions and in a more generalized manner with diffuse lesions, metabolic encephalopathy, hydrocephalus, or deep midline lesions. They are typically most prominent anteriorly in adults (e.g., FIRDA-frontal intermittent rhythmic delta) and most prominent posteriorly in children (e.g., OIRDA-occipital intermittent rhythmic delta).

[0048] Theta (θ) is a frequency range from 4 Hz to 7 Hz. θ is commonly present in young children. It may be present in older children and adults during drowsiness or wakefulness; it can also be present during meditation. Excessive θ at certain ages indicates abnormal activity. It can be considered a focal lesion in focal subcortical lesions; it may be present in a more generalized manner in some cases of diffuse lesions, metabolic encephalopathy, deep midline lesions, or hydrocephalus. Conversely, this range is associated with reports of relaxation, meditation, and creative states.

[0049] Alpha is a frequency range from 7 Hz to 14 Hz. This is the “posterior fundamental rhythm” (also known as the “posterior dominant rhythm” or “posterior alpha rhythm”), present in the posterior regions on both sides of the head, with its amplitude higher on the dominant side. It appears with eye closure and with relaxation and decays with eye opening or mental exertion. In young children, the posterior fundamental rhythm is actually slower than 8 Hz (and therefore technically within the theta range). In addition to the posterior fundamental rhythm, there are other normal alpha rhythms such as sensorimotor rhythms, or μ rhythms that occur when the hands and arms are idle (alpha activity in the contralateral sensorimotor cortex), and a “third rhythm” (alpha activity in the temporal or frontal lobe). Alpha can be abnormal; for example, an EEG with diffuse alpha present in a coma and unresponsive to external stimuli is called an “alpha coma.”

[0050] Beta (β) is a frequency range from 15 Hz to approximately 30 Hz. It is typically present bilaterally in a symmetrical distribution and is most prominent anteriorly. β activity is closely associated with motor behavior and often decays during active movement. Low-amplitude β with multiple and different frequencies is often associated with active, busy, or anxious thinking and active attention. The rhythmicity of the dominant group of β frequencies is associated with various pathologies, such as Dup15q syndrome and drug effects, especially benzodiazepines. In areas of cortical lesion, it may be absent or reduced. It is the dominant rhythm in alert or anxious patients or those with their eyes open.

[0051] γ is a frequency range of approximately 30-100 Hz. Γ rhythms are considered to represent different groups of neurons coming together to form a network for performing a cognitive or motor function.

[0052] The μ range is 8-13 Hz and partially overlaps with other frequencies. It reflects the synchronized firing of motor neurons in the resting state. μ inhibition is considered to reflect the motor mirror neuron system because the pattern disappears when action is observed, which may be because the normal neuron system and the mirror neuron system are "out of sync" and interfere with each other. (en.wikipedia.org / wiki / Electroencephalography)

[0053] Table 1

[0054]

[0055]

[0056] EEG and qEEG: EEG electrodes primarily detect neuronal activity in brain regions only beneath them. However, the electrodes receive activity from thousands of neurons. For example, a square millimeter of cortical surface contains more than 100,000 neurons. Simple periodic waveforms in an EEG only become distinguishable when the input to the region is synchronized with concurrent electrical activity. Temporal patterns associated with specific brain waves can be digitized and encoded in non-transient memory, and embodied in or referenced by computer software.

[0057] EEG (electroencephalography) and MEG (magnetic-electroencephalography) are available techniques for monitoring brain electrical activity. Each typically has sufficient temporal resolution to track dynamic changes in brain electrical activity. Electroencephalography (EEG) and quantitative electroencephalography (qEEG) are electrophysiological monitoring methods that analyze the brain's electrical activity to measure and display patterns corresponding to cognitive states and / or diagnostic information. They are typically non-invasive, with electrodes placed on the scalp, but invasive electrodes are used in some cases. EEG signals can be captured and analyzed by mobile devices commonly referred to as "brain wearable devices." A variety of "brain wearable devices" are readily available on the market today. EEG can be obtained using non-invasive methods in which the convergence of oscillations of brain potentials is recorded using various electrodes attached to the scalp. Most EEG signals originate from the outer layer of the brain (the cerebral cortex), considered largely responsible for thought, emotion, and behavior. Cortical synaptic activity generates electrical signals that vary in the range of 10 to 100 milliseconds. Transcutaneous EEG signals are limited by the following factors: the relative insulating properties of the skull surrounding the brain, the conductivity of cerebrospinal fluid and brain tissue, the relatively low amplitude of individual cell electrical activity, and the distance between the cell current and the electrode. EEG is characterized by: (1) voltage; (2) frequency; (3) spatial location; (4) interhemispheric symmetry; (5) responsiveness (response to changes in state); (6) waveform occurrence characteristics (random, continuous, sustained); and (7) transient temporal morphology. EEG can be divided into two main categories: spontaneous EEG that occurs in the absence of a specific sensory stimulus and evoked potentials (EPs) associated with sensory stimuli such as repetitive flashes of light, auditory tones, finger pressure, or mild electric shocks. The latter is recorded by, for example, time averaging to remove the influence of spontaneous EEG. Non-sensory trigger potentials are also known. EPs are typically time-synchronized with the trigger and therefore have organizational principles. Event-related potentials (ERPs) provide evidence of a direct link between cognitive events and brain electrical activity in a wide range of cognitive paradigms. ERPs are generally considered to be the result of a set of discrete stimuli-evoked brain events. Event-related potentials (ERPs) are recorded in the same manner as EPs, but occur over a longer latency period from the stimulus and are more associated with endogenous brain states.

[0058] Typically, magnetic sensors with sufficient sensitivity to depolarization of individual cells or groups are superconducting quantum interference devices (SQIUDs), which require cryogenic operation at liquid nitrogen temperatures (high-temperature superconductors, HTS) or liquid helium temperatures (low-temperature superconductors, LTS). However, current research has shown the potential feasibility of room-temperature superconductors (20°C). The advantage of magnetic sensing is better potential volume localization due to the dipole nature of the source; however, this increased information also increases the complexity of signal analysis.

[0059] Typically, the detected electromagnetic signal represents an action potential, an automatic response of a nerve cell to depolarization exceeding a threshold, which briefly opens a conduction channel. Cells possess ion pumps that attempt to maintain the depolarized state. Once triggered, the action potential propagates along the two-dimensional membrane, causing a brief surge of high-level depolarized ions. A quiescent period follows depolarization, which typically prevents oscillations within a single cell. Because exons extend from the body of the neuron, the action potential typically travels along the length of the axon, which terminates at a synapse with another cell. Simultaneously with the direct electrical connection between cells, the axon typically releases neurotransmitter compounds into the synapse, leading to depolarization or hyperpolarization of the target cell. In practice, the result may also be the release of hormones or peptides, which can have local or more distant effects.

[0060] Externally detectable electric fields often do not contain signals as low-frequency signals, such as static-level polarization or cumulative depolarization or hyperpolarization between action potentials. In myelinated tracts, the current flowing at the segmentation points tends to be small, and therefore the signal from individual cells is weak. Thus, the largest signal components originate from synapses and cell bodies. In the cerebrum and cerebellum, these structures are primarily located in the cortex, which is largely adjacent to the skull, making EEG usable because it provides spatial discrimination based on electrode location. However, deep signals attenuate and are poorly localized. Magnetoencephalography (MEG) detects dipoles derived from changes in current rather than voltage. In the case of radially or spherically symmetrical current flow over short distances, dipoles tend to cancel each other out, while net current flow enhances long axons. Therefore, EEG reads signals different from MEG.

[0061] EEG-based studies of emotion specificity at the single-electrode level have demonstrated that asymmetric activity in the frontal region, particularly in the α (8-12 Hz) band, is associated with emotion. Enjoyable voluntary facial expressions like smiling produce higher left frontal activation. Reduced left frontal activity is observed in voluntary facial expressions of fear. In addition to α band activity, the theta band power at the frontal midline (Fm) has also been found to be related to emotional state. Pleasant emotions (the opposite of unpleasant) are associated with increased theta power at the frontal midline. Many studies have attempted to distinguish between various emotional states reflected in EEG using pattern classification, such as neural networks, statistical classifiers, and clustering algorithms.

[0062] EEG-based studies of emotion specificity at the single-electrode level have demonstrated that asymmetric activity in the frontal region, particularly in the α (8-12 Hz) band, is associated with emotion. Ekman and Davidson found that a pleasant voluntary facial expression, smiling, produces higher left frontal activation (Ekman P, Davidson RJ (1993), “Voluntary Smiling Changes Regional Brain Activity,” Psychol Sci 4:342-345). Another study by Coan et al. found reduced left frontal activity in voluntary facial expression fear (Coan JA, Allen JJ, Harmon-Jones E (2001), “Voluntary facial expression and hemispheric asymmetry over the frontal cortex,” Psychophysiology 38:912–925). In addition to α band activity, the theta band power at the frontal midline (Fm) has also been found to be associated with emotional state. For example, Sammler and colleagues showed that pleasant (as opposed to unpleasant) emotions are associated with increased theta power in the frontal midline (Sammler D, Grigutsch M, Fritz T, Koelsch S (2007), "Music and emotion: Electrophysiological correlates of the processing of pleasant and unpleasant music", Psychophysiology 44:293-304). To further demonstrate whether these emotion-specific EEG properties are strong enough to distinguish between various emotional states, some studies have utilized pattern classification analysis methods.

[0063] Using EEG-based functional connectivity could be more appropriate for detecting different emotional states via EEG. Various methods exist for estimating EEG-based functional brain connectivity: correlation, coherence, and phase synchronization indices between each pair of EEG electrodes have been used. The assumption is that a higher correlation plot indicates a stronger relationship between the two signals (Brazier MA & Casby JU (1952), “Cross-correlation and autocorrelation studies of electroencephalographic potentials,” *Electroencephalography and Clinical Neurophysiology* 4:201–211). Coherence provides similar information to correlation, but also includes the covariance between the two signals, which is a function of frequency. (Cantero JL, Atienza M, Salas RM, Gomez CM (1999) "Alpha EEG coherence in different brain states: an electrophysiological index of the arousal level in human subjects", Neuroscience Letters 271:167–70). The hypothesis is that a higher correlation indicates a stronger relationship between the two signals.(Guevara MA, Corsi-Cabrera M (1996) "EEG coherence or EEG correlation?", *International Journal of Psychophysiology* 23:145–153; (Cantero JL, Atienza M, Salas RM, Gomez CM (1999) "αEEG coherence in different brain states: an electrophysiological indicator of arousal levels in human subjects", *Neuroscience Letters* 271:167–70; Adler G, Brassen S, Jajcevic A (2003) "EEG coherence in Alzheimer's dementia", *Journal of Neural Transm* 110:1051–1058; Deeny SP, Hillman CH, Janelle CM, Hatfield BD (2003) "Cortico-cortical communication and superior performance in skilled marksmen: An EEG coherence analysis" *Journal of Sport Exercise Psychology* 25:188–204. Phase synchronization within a neuronal group based on the phase difference between two signals is another method for estimating EEG-based functional connectivity between brain regions. This method is described in (Franaszczuk PJ, Bergey GK (1999) "An autoregressive method for the measurement of synchronization of interictal and ictal EEG signals" *Biological Cybernetics* 81:3-9).

[0064] Many groups have used EEG-based functional brain connectivity to examine emotion specificity. For example, Shin and Park showed that the correlation coefficient between temporal and occipital pathways increased when emotional states became more negative at high room temperature (Shin JH, Park DH. (2011) "Analysis for Characteristics of Electroencephalogram (EEG) and Influence of Environmental Factors According to Emotional Changes" by Lee G, Howard D, ...). Editor D, *Convergence and Hybrid Information Technology*, Springer Berlin Heidelberg, 488–500. Hinrichs and Machleidt demonstrated that coherence in the alpha band is reduced during sadness compared to happiness (Hinrichs H, Machleidt W (1992) "Basic emotions reflected in EEG coherences", *International Journal of Psychophysiology* 13:225–232). Miskovic and Schmidt found that EEG coherence between the prefrontal and posterior cortices increased when viewing highly emotionally aroused (i.e., threatening) images compared to viewing neutral images (Miskovic V, Schmidt LA (2010) "Cross-regional cortical synchronization during affective image viewing," *Brain Research* 1362:102–111). Costa and colleagues applied synchronization indices to detect the interaction of different brain regions under different emotional states (Costa T, Rognoni E, Galati D (2006) "EEG phase synchronization during emotional response to positive and negative film stimuli," *Neuroscience Letters Letters* 406:159–164). Costa's results showed an overall increase in synchronization indices within the frontal channels during emotional stimuli, particularly during negative emotions (i.e., sadness). Furthermore, phase synchronization patterns were found to differ between positive and negative emotions. Costa also found that sadness was more synchronized than happiness across all frequency bands and was associated with broader synchronization between the left and right frontal regions and within the left hemisphere. Conversely, happiness was associated with broader synchronization between the frontal and occipital regions.

[0065] Different connectivity indices are sensitive to different characteristics of EEG signals. Correlation is sensitive to phase and polarity, but not amplitude. Changes in both amplitude and phase lead to changes in coherence (Guevara MA, Corsi-Cabrera M (1996), “EEG coherence or EEG correlation?” *International Journal of Psychophysiology* 23:145-153). Phase synchronization indices are sensitive only to changes in phase (Lachaux JP, Rodriguez E, Martinerie J, Varela FJ (1999), “Measuring phase synchrony in brain signals”, *Human BrainMapp* 8:194-208).

[0066] Many studies attempt to classify emotional states by recording and statistically analyzing EEG signals from the central nervous system.

[0067] Emotional states can be classified using a dimensional theory that posits neutral, positive, and negative emotional states, as numerous studies have shown that the central nervous system's response is related to emotional valence and arousal. (See, for example, Davidson RJ (1993), “Cerebral Asymmetry and Emotion: Conceptual and Methodological Conundrums,” *Cognition & Emotion* 7:115-138; Jones NA and Fox NA (1992), “Electroencephalogram asymmetry during emotionally evocative films and its relation to positive and negative affectivity,” *Brain Cognition* 20:280-299; Schmidt LA and Trainor LJ (2001), “Frontal brain electrical activity (EEG) distinguishes valence and intensity of musical emotion.”) “emotions”, Cognition Emotion 15:487-500; Tomarken AJ, Davidson RJ, Henriques JB (1990) “Resting frontal brain asymmetry predicts affective responses to films”, J Pers Soc Psychol 59:791-801.As Mauss and Robins (2009) suggest, "measures of emotional response appear to be structured along dimensions (e.g., valence, arousal) rather than discrete emotional states (e.g., sadness, fear, anger)."

[0068] EEG-based functional connectivity was found to differ significantly between neutral, positive, and negative emotional states. Lee YY and Hsieh S (2014), “Classifying Different Emotional States by Means of EEG-Based Functional Connectivity Patterns,” PLOS ONE, 9(4):e95415.doi.org / 10.1371 / journal.pone.0095415. Connectivity patterns could be detected using pattern classification analysis with quadratic discriminant analysis. The results indicated that the classification rate was better than randomness. They concluded that estimating EEG-based functional connectivity provides a useful tool for studying the relationship between brain activity and emotional states.

[0069] Emotions influence learning. The learner model of the Intelligent Tutoring System (ITS), initially composed of cognitive modules, has been extended to include both psychological and emotional modules. Alicia Heraz et al. introduced an emotional agent. This agent interacts with the ITS to convey the learner's emotional state based on their mental state. Mental state is obtained from the learner's brainwaves. The agent learns using machine learning techniques to predict the learner's emotions. (Alicia Heraz, Ryad Razaki, Claude Frasson, "Using machine learning to predict learner emotional state from brainwaves", *Advanced Learning Technologies*, 2007, ICALT 2007. Seventh IEEE International Conference on Advanced Learning Technologies (ICALT 2007).)

[0070] Using EEG to assess emotional states has many practical applications. One such application is that of the Singapore tourism group, which uses brainwave measurements to develop emotion-based travel guides. “By studying the brainwaves of a family on vacation, the researchers drew up the Singapore Emotion Travel Guide, which advises future visitors of the emotions they can expect to experience at different attractions” (www.lonelyplanet.com / news / 2017 / 04 / 12 / singapore-emotion-travel-guide). Joel Pearson and his team at the University of New South Wales developed a protocol for using EEG to measure travelers' brainwaves and decode specific emotional states.

[0071] Another recently released application involves virtual reality (VR) technology. Looxid Labs has launched a technology that utilizes EEG data from subjects wearing VR headsets. Looxid Labs' intention is to incorporate brainwaves into VR applications to accurately infer emotions. Other products, such as MindMaze and even Samsung, are attempting to create similar applications using facial muscle recognition (scottamyx.com / 2017 / 10 / 13 / looxid-labs-vr-brain-waves-human-emotions / ). According to its website (looxidlabs.com / device-2 / ), the Looxid Labs development kit provides a VR headset embedded with miniature eye and brain sensors. It uses six EEG channels: Fp1, Fp2, AF7, AF8, AF3, and AF4 from the International 10-20 system.

[0072] To assess a user's mental state, computers can be used to analyze EEG signals generated by the user's brain. However, emotional states of the brain are complex, and brainwaves associated with specific emotions appear to change over time. Wei-Long Zheng (Shanghai Jiao Tong University) used machine learning to identify and reliably reproduce emotional brain states. The machine learning algorithm discovered a set of patterns that clearly distinguished positive, negative, and neutral emotions, applicable to different subjects as well as the same subjects over time, with an accuracy of approximately 80%. (See Wei-Long Zheng, Jia-Yi Zhu, and Bao-Liang Lu, “Identifying Stable Patterns over Time for Emotion Recognition from EEG”, arxiv.org / abs / 1601.02197; see also “How One Intelligent Machine Learned to Recognize Human Emotions”, MIT Technology Review, January 23, 2016).

[0073] MEG: Magnetoencephalography (MEG) is a functional neuroimaging technique used to map brain activity by recording the magnetic fields generated by naturally occurring electrical currents in the brain using highly sensitive magnetometers. SQUID (superconducting quantum interference device) arrays are currently the most common type of magnetometer, while SERF (spin-free exchange relaxation) magnetometers are under investigation. Matti, Hari, Riitta, Ilmoniemi, Risto J, Knuutila, Jukka, Lounasmaa, Olli V. (1993). “Magnetoencephalography-theory, instrumentation, and applications to noninvasive studies of the working human brain”, Reviews of Modern Physics 65(2):413-497. ISSN0034-6861.doi:10.1103 / RevModPhys.65.413. As is well known, "neuronal activity causes local changes in cerebral blood flow, blood volume, and blood oxygenation" (Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation, KKKwong, JW Belliveau, D.A.C. Hesler, I.E. Goldberg, R.M. Weisskoff, B.B. Poncelet, D.N. Kennedy, B.E. Hoppel, M.S. Sohen, and R. Turner). Using a "122-channel DCSQUID magnetometer with a helmet-shaped detector array covering the subject's head," it was shown that "the system allows simultaneous recording of magnetic activity all over the head" (122-channel squid instrument for investigating the magnetic signals from the human brain, A.A. Honen, MS). MJ Kajola, JET Knuutila, PPLaine, OV Lounasmaa, LT Parkkonen, JTSimola and CD Tesche Physica Scripta, Volume 1993, T49A).

[0074] In some cases, the magnetic fields cancel each other out, and therefore the detectable electrical activity may be fundamentally different from that obtained via EEG. However, the main differences in brain rhythms can be detected by two methods.

[0075] MEGs attempt to detect the emission of magnetic dipoles from cells, such as neural action potentials. A typical sensor for MEGs is a superconducting quantum interference device (SQUID). These currently require cooling to liquid nitrogen or liquid helium temperatures. However, the development of room-temperature or near-room-temperature superconductors and miniature cryogenic coolers could allow for field deployment as well as portable or mobile detectors. Because MEGs are less affected by the conductivity and dielectric properties of the medium and because they inherently detect magnetic field vectors, MEG technology allows for volumetric mapping of brain activity and the differentiation of complementary activities that might suppress detectable EEG signals. MEG technology also supports vector mapping of fields because the magnetic emitter is inherently a dipole, and therefore inherently allows for the acquisition of a wealth of information.

[0076] EEG and MEG can monitor mental status. For example, deep sleep is associated with larger amplitude, slower EEG oscillations. Various signal analysis methods allow for robust identification of different sleep patterns (SS), depth of anesthesia, seizures, and connections to detailed cognitive events.

[0077] Positron Emission Tomography (PET) Scan: PET scans are imaging tests that help show how tissues and organs function (Bailey, DL, DWTownsend, PEValk, MNMaisey (2005). "Positron Emission Tomography: Basic Sciences," Secaucus, NJ: Springer-Verlag, ISBN 1-85233-798-2). PET scans use radiopharmaceuticals (positron emission tracers) to visualize this activity. They use this radiation to produce 3-D images that are colored for different activities in the brain.

[0078] fMRI: Functional magnetic resonance imaging, or functional MRI (fMRI), is a functional neuroimaging procedure that uses MRI technology to measure brain activity by detecting changes associated with blood flow (“Magnetic Resonance, a critical peer-reviewed introduction” European Magnetic Resonance Forum, 17 November 2014; Huettel, Song, and McCarthy (2009)). Yukiyasu Kamitani et al., “Neuron” (DOI: 10.1016 / j.neuron.2008.11.004), used images of brain activity extracted from a functional MRI scanner to reconstruct black-and-white images from scratch. See also Celeste Biever, “Mind-reading software could record your dreams.” New Scientist, December 12, 2008 (www.newscientist.com / article / dn16267-mind-reading-software-could-record-your-dreams / ).

[0079] Functional near-infrared spectroscopy (fNIRS): fNIR is a non-invasive imaging method involving the quantification of chromophore concentrations resolved by measurements of near-infrared (NIR) light attenuation or by changes in time or phase. NIR spectroscopy utilizes an optical window in which skin, tissue, and bone are almost transparent to NIR light in the 700–900 nm spectrum, while hemoglobin (Hb) and deoxyhemoglobin (deoxy-Hb) are stronger light absorbers. The difference in the absorption spectra of deoxy-Hb and oxygen-Hb allows for the measurement of relative changes in hemoglobin concentration by using light attenuation at multiple wavelengths. Two or more wavelengths are selected, one above the isoabsorption point of 810 nm and one below, at which deoxy-Hb and oxygen-Hb have the same absorption coefficient. Using the modified Beer-Lambert law (mBLL), the relative concentration can be calculated as a function of the total photon path length. Typically, the light emitter and detector are placed on the same side of the subject's skull, so the recorded measurements are due to backscattered (reflected) light following the elliptical pathway. The use of fNIR as a functional imaging method relies on the principle of neurovascular coupling, also known as hemodynamic response or blood oxygen level-related (BOLD) response. This principle also forms the core of fMRI technology. Through neurovascular coupling, neuronal activity is linked to relevant changes in local cerebral blood flow. fNIR and fMRI are sensitive to similar physiological changes and are often comparative methods. Studies involving fMRI and fNIR have shown highly correlated results for cognitive tasks. Compared to fMRI, fNIR has several advantages in terms of cost and portability, but it cannot be used to measure cortical activity deeper than 4 cm due to limitations in light emitter power and has more limited spatial resolution. fNIR includes the use of diffusion optical tomography (DOT / NIRDOT) for functional purposes. Multiplexing fNIRS channels can enable 2D topographic functional mapping of brain activity (e.g., using Hitachi ETG-4000 or Artinis Oxymon), while using multiple emitter spacings can be used to construct 3D tomographic maps.

[0080] Beste Yuksel and Robert Jacob, "Brain Automated Chorale (BACh)", ACM CHI 2016, DOI:10.1145 / 2858036.2858388, present a system to help beginners learn to play Bach chorales on the piano by measuring how hard their brains are working. This is done by estimating the brain's workload using functional near-infrared spectroscopy (fNIRS), a technique that measures oxygen levels in the brain—specifically, the prefrontal cortex. A working brain draws in more oxygen. Sensors strapped to the player's forehead communicate with a computer that delivers new music online, one line at a time. See also Anna Nowogrodzki's "Mind-reading tech helps beginners quickly learn to play Bach," New Scientist, February 9, 2016, available online at: www.newscientist.com / article / 2076899-mind-reading-tech-helps-beginners-quickly-learn-to-play-bach / .

[0081] LORETA: Low-resolution electromagnetic tomography of the brain, commonly referred to as LORETA, is a functional imaging technique that typically uses a linearly constrained minimum variance vector bundler in the time-frequency domain, as seen in Gross et al., “Dynamic imaging of coherent sources: Studying neural interactions in the human brain”, PNAS. Described in 98,694-699,2001. It allows for the induction and evoked oscillatory activity of images (mostly 3D) within a variable time-frequency range, where time is relative to the triggering event. There are three categories of imaging related to the techniques used for LORETA. See wiki.besa.de / index.php?title=Source_Analysis_3D_Imaging#Multiple_Source_Beamformer_.28MSBF.29. The Multiple Source Beamformer (MSBF) is a tool for imaging brain activity. It is applied in the time-frequency domain and is based on single-trial data. Therefore, it can image not only the evoked images but also the induced activity, which is not visible in the time-domain averaging of the data. Coherent Source Dynamic Imaging (DICS) can find the coherence between any two pairs of voxels in the brain or between an external source and a brain voxel. DICS requires time-frequency transformed data and can find the coherence between evoked and induced activity. The following imaging methods provide images of brain activity based on distributed multi-source models: CLARA is an iterative application of LORETA images. It concentrates the obtained 3D images in each iteration step. LAURA uses a spatial weighting function in the form of a local autoregressive function. LORETA has a 3D Laplacian operator implemented as a spatially weighted prior. sLORETA is an unweighted minimum norm normalized by the resolution matrix. swLORETA is equivalent to sLORETA except for additional depth weights. SSLOFO is an iterative application of normalized minimum norm images in which the source space is continuously reduced. User-defined volumetric images allow for experimentation using different imaging techniques. It is possible to specify user-defined parameters for a series of distributed source images to generate new imaging techniques. If no individual MRI is available, the minimum norm image is displayed on a standard brain surface and computed against standard source locations. If available, a single brain surface is used to construct a distributed source model and image brain activity. Unlike classical LORETA, cortical LORETA is computed on the cortical surface rather than in a 3D volume. Unlike classical CLARA, cortical CLARA is computed on the cortical surface rather than in a 3D volume. Multi-source probe scanning (MSPS) is a tool for validating discrete multi-source models.The source sensitivity image displays the sensitivity of the selected source in the current discrete source model and is therefore data-independent.

[0082] Neurofeedback: Neurofeedback (NFB), also known as neurotherapy or neurobiofeedback, is a type of biofeedback that uses a real-time display of brain activity, most commonly an electroencephalogram (EEG), to teach brain function to self-regulate. Typically, sensors are placed on the scalp to measure activity, with the results displayed using a video display or sound. Feedback can also take many other forms. Typically, attempts are made to present feedback through primary sensory input, but this is not a limitation of the technique.

[0083] The applications of neurofeedback in enhancing performance extend to art forms such as music, dance, and performance. Research by musicians at music conservatories has found that alpha-theta training benefits three vocal ranges: musicality, communication, and technique. Historically, alpha-theta training (a form of neurofeedback) was created to assist creativity by inducing hypnosis—a "boundary state of wakefulness associated with creative insights"—through promoting neuronal connectivity. Alpha-theta training has also been shown to improve novice singing in children. The combination of alpha-theta neurofeedback with heart rate variability training (a form of biofeedback) has yielded benefits in dance by enhancing performance in competitive ballroom dancing and increasing cognitive creativity among contemporary dancers. Furthermore, neurofeedback has been shown to potentially infuse a superior state of flow into the performer due to greater immersion during performance.

[0084] Several studies on the brainwave activity of experts performing tasks related to their respective professional fields have revealed certain characteristic signs of what is known as “flow” associated with top performance. Mihaly Csikszentmihalyi (University of Chicago) found that the most skilled chess players showed less EEG activity in the prefrontal cortex, which is typically associated with higher cognitive processes such as working memory and verbal expression during the game.

[0085] Chris Berka et al., *Advanced Brain Monitoring*, Carlsbad, California, *The International Journal of Sport and Society*, Vol. 1, p. 87, focused on the brainwaves of Olympic archers and professional golfers. The team found a slight increase in alpha band patterns a few seconds before an archer shoots or a golfer hits a ball. This could be related to the case-dependent negative changes observed in evoked potential studies, as well as the premotor potential or BP (from German, "preparatory potential"), a measure of activity in the brain's motor cortex and complementary motor areas that induces voluntary muscle movement. Berka also used neurofeedback to train novice archers. Each person was connected to electrodes that combed and displayed specific brainwaves along with a monitor that measured their heart rate. By controlling their breathing and learning to intentionally manipulate waveforms on a screen in front of them, the novices managed to generate alpha wave characteristics of a flowing state. This, in turn, helps improve its accuracy in hitting targets.

[0086] Low-Energy Neurofeedback Systems (LENS): LENS, or low-energy neurofeedback systems, use very low-power electromagnetic fields to deliver feedback to the recipient. The feedback travels along the same leads that carry brainwaves to amplifiers and computers. Although the feedback signal is weak, it still produces measurable changes in brainwaves without conscious effort from the individual receiving the feedback. The system is software-controlled to receive input from EEG electrodes via the scalp to control the stimulation. Neurofeedback uses a feedback frequency that is different from but related to the dominant brainwave frequency. When exposed to this feedback frequency, the power of the EEG amplitude distribution changes. In most cases, the power of the brainwaves decreases, but sometimes it increases. In either case, the result is a change in brainwave state and a greater ability for the brain to regulate itself.

[0087] Content-Based Brainwave Analysis: Memory Is Not Unique. Janice Chen, *Nature Neuroscience*, DOI:10.1038 / nn.4450, shows that when people describe scenes from Sherlock Holmes plays, their brain activity patterns are almost identical for each scene. Furthermore, there is evidence that when people tell this to someone else, they also implant the same activity into their brains. In addition, Chen et al. found in a study of people who hadn't seen the movie listening to someone else's description of it that the listeners' brain activity looked very similar to that of people who had seen the movie. See also Andy Coghlan, "Our brains record and remember things in exactly the same way," *New Scientist*, December 5, 2016 (www.newscientist.com / article / 2115093-our-brains-record-and-remember-things-in-exactly-the-same-way / ).

[0088] Brian Pasley, in *Frontiers in Neuroengineering*, doi.org / whb, developed a technique for reading thoughts. The team hypothesized that hearing speech and thinking about oneself might activate some of the same neural signatures in the brain. They conjectured that an algorithm trained to recognize spoken speech might also be able to recognize thought-provoking utterances. In experiments, a speech-trained decoder was able to reconstruct the thought-provoking utterances of several volunteers using neural activity alone. See also Helen Thomson, “Hearing our inner voice,” *New Scientist*, October 29, 2014 (www.newscientist.com / article / mg22429934-000-brain-decoder-can-eavesdrop-on-your-inner-voice / ).

[0089] Jack Gallant and colleagues were able to detect which image in a set a person was viewing based on brain scans using software that compares a subject's brain activity while viewing images with images captured while viewing "training" photos. The program then selects the most likely match from a set of previously unviewed images.

[0090] Ann Graybiel and Mark Howe used electrodes to analyze brain waves in the ventral striatum of rats while teaching them to navigate a maze. When the rats learned the task, their brain activity showed rapid bursts of gamma waves. Once the rats mastered the task, their brain waves slowed to almost a quarter of their initial frequency, becoming beta waves. Graybiel's team hypothesized that this shift reflects when learning becomes habitual.

[0091] Bernard Balleine, *Proceedings of the National Academy of Sciences*, DOI: 10.1073 / pnas.1113158108. See also Wendy Zukerman, "Habits form when brainwaves slow down," *New Scientist*, September 26, 2011 (www.newscientist.com / article / dn20964-habits-form-when-brainwaves-slow-down / ). It is hypothesized that slower brainwaves may be the brain clearing excess activity to improve behavior. It is suggested that the rate of increase may be enhanced when learning technology by increasing this beta wave activity.

[0092] US 9,763,592 provides a system for instructing changes in user behavior, the system comprising: collecting and analyzing a dataset of bioelectrical signals; and providing suggestions for behavior change based on the analysis. Stimuli may be provided to prompt the user to perform actions, which may be visual, auditory, or tactile.

[0093] Chess is a good example of a cognitive task that requires extensive training and experience. Numerous EEG studies have been conducted on chess players. Pawel Stepien, Wlodzimierz Klonowski, and Nikolay Suvorov, “Nonlinear analysis of EEG in chessplayers,” *EPJ Nonlinear Biomedical Physics*, 2015, 3:1, demonstrates the superior availability of the Higuchi Fractal Dimension method for analyzing EEG signals related to chess tasks compared to sliding window empirical mode decomposition. The paper shows that even when there is no significant difference in the contribution of different EEG bands to the total power of the signal between playing and relaxed states, the EEG signal during play is more complex, nonlinear, and unpredictable. Further data needs to be gathered from more chess experts and compared with data from novice chess players. See also Junior, LRS, Cesar, FHG, Rocha, FT, and Thomas, CE, “EEG and Eye Movement Maps of Chess Players”, Proceedings of the Sixth International Conference on Pattern Recognition Applications and Methods (ICPRAM 2017), pp. 343-441 (fei.edu.br / ~cet / icpram17_LaercioJunior.pdf).

[0094] Estimating EEG-based functional connectivity provides a useful tool for studying the relationship between brain activity and emotional states. See You-Yun Lee and Shulan Hsieh, “Classification of different emotional states using EEG-based functional connectivity patterns,” 4 / 17 / 2014 (doi.org / 10.1371 / journal.pone.0095415), which aims to classify different emotional states using EEG-based functional connectivity patterns and demonstrates that changes in EEG-based functional connectivity are significantly different between emotional states. Furthermore, connectivity patterns were detected using pattern classification analysis with quadratic discriminant analysis. The results indicate that the classification rate is superior to randomness. Estimating EEG-based functional connectivity provides a useful tool for studying the relationship between brain activity and emotional states.

[0095] Neuromodulation / Neuroenhancement: Neuromodulation alters neural activity by targeting specific neural sites within the body with stimuli such as electrical stimulation or chemical agents. Neuromodulation is performed to normalize or modulate neural tissue function. Neuromodulation is an evolving therapy that may involve a range of electromagnetic stimuli, such as magnetic fields (TMS, rTMS), electrical currents (TES, e.g., tDCS, HD-tDCS, tACS, electrosleep), or drugs injected directly into the subdural space (intrathecal drug delivery). Emerging applications involve the targeted introduction of genes or gene regulators and light (optogenetics). The most clinically experienced approach is electrical stimulation. Neuromodulation, whether electrical or magnetic, employs the body's natural biological response to stimulating nerve cell activity, which can influence the population of nerves by releasing neurotransmitters such as dopamine or other chemical messengers such as peptide P that can modulate the excitability and excitation patterns of neural circuits. More direct electrophysiological effects on nerve membranes also exist. Depending on the application, the ultimate effect is to "normalize" the perturbed state of neural network function. The hypothesized mechanisms of action for neural stimulation include depolarization blockade, random normalization of neural excitation, axonal blockade, reduction of neural excitation keratosis, and inhibition of neural network oscillations. Although the exact mechanisms of neural stimulation are unknown, its empirical effectiveness has led to considerable clinical application.

[0096] Neurostimulants (NE) refer to the targeted enhancement and extension of cognitive and emotional abilities based on an understanding of the underlying neurobiology of healthy individuals without any mental illness. Thus, it can be considered a comprehensive term encompassing both pharmacological and non-pharmacological approaches to improving cognitive, emotional, and motor functions, and is the primary ethical and legal discourse accompanying these goals. Crucially, for any drug to qualify as a neurostimulant, it must reliably produce substantial cognitive, emotional, or motor benefits beyond normal function in healthy individuals (or in a selected group with pathological conditions) without causing side effects: at most reaching levels of commonly used, comparable, legal substances or activities such as caffeine, alcohol, and sleep deprivation. NE pharmacological agents include well-proven nootropics such as racetam, vinpocetine, and phosphatidylserine, as well as drugs used to treat patients with neurological disorders. Non-pharmacological measures include non-invasive brain stimulation for improving a wide range of cognitive and emotional functions, and brain-computer interfaces with the potential to extend the broad range of available motor and cognitive actions to humans.

[0097] Brain stimulation: Non-invasive brain stimulation (NIBS) bypasses the methods of other imaging techniques, making it possible to establish causal relationships between cognitive processes and the function of specific brain regions. NIBS can provide information about when a specific process occurs. NIBS offers the opportunity to study brain mechanisms beyond process localization, providing information about when and how activity in a given brain region participates in cognitive processes. When using NIBS to explore cognitive processes, it is important not only to understand how NIBS works but also to understand the function of the neural structures themselves. Non-invasive brain stimulation (NIBS) methods, including transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (TES), are used in cognitive neuroscience to induce transient changes in brain activity and thereby alter the behavior of subjects.

[0098] The application of NIBS aims to establish the role of a given cortical region in a specific ongoing motor, perceptual, or cognitive process. Physically, NIBS techniques influence neuronal states through different mechanisms. In TMS, a solenoid (coil) is used to deliver a strong and transient magnetic field, or "pulse," to induce a transient current at the cortical surface beneath the coil. The pulse causes rapid, above-threshold depolarization of the cell membrane affected by the current, subsequently causing transsynaptic depolarization or hyperpolarization of interconnected neurons. Thus, strong TMS can induce currents that generate action potentials in neurons, while weak (subthreshold) TMS can alter the cell's susceptibility to depolarization. Complex coil arrays can provide complex 3D excitation fields. In contrast, in TES techniques, stimulation involves applying a weak current directly to the scalp via an electrode pair. As a result, TES induces subthreshold polarization of cortical neurons, which is too weak to generate action potentials. (Suprathreshold tES corresponds to electroconvulsive therapy, a currently unpromising but apparently effective treatment for depression). However, by altering intrinsic neuronal excitability, TES can induce changes in residual membrane potential and postsynaptic activity in cortical neurons. This, in turn, can alter the spontaneous firing rate of neurons and modulate their response to incoming signals, leading to changes in synaptic efficacy. Typical applications of NIBS involve different types of protocols: TMS can be delivered as a single pulse at precise time (spTMS), as pairs of pulses separated by variable intervals, or as a series of stimuli in a regular or patterned protocol of repetitive TMS (rTMS). In TES, different protocols are established by the current used and by its polarity; these protocols can be direct (anodic or cathodic transcranial direct electrical stimulation: tDCS), alternating at a fixed frequency (transcranial alternating current stimulation: tACS), oscillating transcranial direct current stimulation (oscillating tDCS), high-definition transcranial direct current stimulation (HD-tDCS), or at a random frequency (transcranial random noise stimulation: tRNS) (Nitsche et al., 2008; Paulus, 2011).

[0099] Typically, the ultimate effect of NIBS on the central nervous system depends on a long list of parameters (e.g., frequency, temporal characteristics, intensity, coil / electrode geometry, current direction) delivered as part of the experimental procedure before (offline) or during (online) the task. Additionally, these factors interact with several variables related to anatomical structures (e.g., the nature and location of brain tissue) and the physiology (e.g., sex and age) and cognitive state of the stimulated area / subject. The entrainment hypothesis suggests the possibility of inducing specific oscillatory frequencies in the brain using external oscillatory forces (e.g., rTMS, and tACS). The physiological basis of oscillatory cortical activity lies in the timing of interacting neurons; brain rhythms emerge when groups of neurons synchronize their firing activities, network oscillations are generated, and the basis for interactions between brain regions may develop. The reported studies lack consistency and inference are limited due to the various experimental protocols used for brain stimulation, the descriptions of the actual protocols employed, and the limited controls. Therefore, despite some consensus on various aspects of the effects of extracranial brain stimulation, the results achieved have a degree of uncertainty depending on the details of the implementation. On the other hand, within a specific experimental protocol, it is possible to obtain statistically significant and reproducible results. This suggests that feedback control may be effective for implementation schemes that control stimuli for a given purpose; however, existing research employing feedback control is lacking.

[0100] Changes in neuronal thresholds are caused by alterations in membrane permeability (Liebetanz et al., 2002), which affect the responses of task-related networks. The same mechanism may be responsible for both TES and TMS methods, i.e., noise-induced neural activity within the system. However, TES-induced neural activity will be highly influenced by the system state, as it is a neuromodulation method (Paulus, 2011), and its effect will depend on the activity of the stimulated region. Therefore, the final outcome will depend heavily on task characteristics, system state, and how TES interacts with such states.

[0101] In TMS, magnetic pulses cause a rapid increase in current, which in some cases may lead to above-threshold depolarization of the cell membrane under the influence of the current, thereby triggering action potentials and causing transsynaptic depolarization or hyperpolarization of connected cortical neurons, depending on their natural response to the firing of one or more stimulated neurons. Thus, TMS activates a neural population that depends on several factors and may be equivalent (promotion) or inequivalent (inhibition) to task performance. TES induces polarization of cortical neurons at a subthreshold level, weak enough not to induce action potentials. However, by inducing polarity shifts with intrinsic neuronal excitability, TES can alter the spontaneous firing rate of neurons and modulate their response to incoming signals. In this sense, the effects induced by TES are even more constrained by the state of the stimulated region determined by the condition. In short, NIBS results in stimulation-induced modulation of the state, which can essentially be defined as noise-induced modulation. The induced noise will not only be random activity but will also depend on the interaction of many parameters from the characteristics of the state stimulus.

[0102] NIBS-induced noise will be influenced by the state of the neural population in the stimulated area. Although the type and number of neurons “triggered” by NIBS are theoretically random, the changes in induced neuronal activity may be related to ongoing activity; however, even with the mention of a nondeterministic process, the introduced noise will not be a completely random element. Because it will be partially determined by experimental variables, the levels of noise introduced by the stimulus and by the context, as well as the interaction between the two noise levels (stimulus and context), can be estimated. Known transcranial stimulation does not allow for the use of concentrated and highly targeted signals to stimulate clearly defined areas of the brain to establish a unique brain-behavior relationship; therefore, the known introduced stimulus activity in brain stimulation is “noise.”

[0103] Cosmetic neuroscience has emerged as a new research area. Roy Hamilton, Samuel Messing, and Anjan Chatterjee, “Rethinking the thinking cap – Ethics of neural enhancement using noninvasive brain stimulation,” *Neurology*, January 11, 2011, Vol. 76, No. 2, pp. 187-193 (www.neurology.org / content / 76 / 2 / 187.), discuss the use of noninvasive brain stimulation techniques, such as transcranial magnetic stimulation and transcranial direct current stimulation, to enhance neural functions: cognitive skills, emotion, and social cognition.

[0104] Computerized brain stimulation (EBS), or focal brain stimulation (FBS), is a form of clinical neurobiological electrotherapy used to stimulate neurons or neural networks in the brain by directly or indirectly exciting cell membranes with electrical currents. See en.wikipedia.org / wiki / Electrical_brain_stimulation. CNS stimulation can affect motor skills.

[0105] Transcranial electrical stimulation (tES): tES (tDCS, tACS, and tRNS) is a family of non-invasive methods that use weak direct current to polarize target brain regions for cortical stimulation. The most commonly used and well-known method is tDCS, as all considerations for its use have been extended to other tES methods. The assumptions about the cognitive applications of tDCS are very similar to those of TMS, with the exception that tDCS has never been considered a dummy lesion method. tDCS can increase or decrease cortical excitability in the stimulated brain regions and correspondingly promote or inhibit behavior. tES does not induce action potentials but rather modulates the neuronal response threshold, thus it can be defined as subthreshold stimulation.

[0106] Michael A. Nitsche and Armin Kibele, “Noninvasive brain stimulation and neural entrainment enhance athletic performance—a review,” *J. Cognitive Enhancement* 1.1 (2017): 73-79, discuss noninvasive brain stimulation (NIBS) as a method that bypasses other imaging techniques, making it possible to establish causal relationships between cognitive processes and the function of specific brain regions. NIBS can provide information about when a specific process occurs. NIBS offers the opportunity to study brain mechanisms beyond process localization, providing information about when and how activity in a given brain region participates in cognitive processes. When using NIBS to explore cognitive processes, it is important not only to understand how NIBS works but also to understand the function of the neural structures themselves. Noninvasive brain stimulation (NIBS) methods, including transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (tES), are used in cognitive neuroscience to induce transient changes in brain activity and thereby alter the behavior of subjects. Applications of NIBS aim to establish the role of a given cortical region in a specific ongoing motor, perceptual, or cognitive process (Hallett, 2000; Walsh and Cowey, 2000). Physically, NIBS techniques influence neuronal states through different mechanisms. In TMS, a solenoid (coil) is used to deliver a strong and transient magnetic field, or "pulse," to induce a transient current at the cortical surface beneath the coil. (US2004078056) The pulse causes rapid and above-threshold depolarization of the cell membrane affected by the current (Barker et al., 1985, 1987), subsequently causing transsynaptic depolarization or hyperpolarization of interconnected neurons. Thus, TMS induces currents that generate action potentials in neurons. Complex coil arrays can provide complex 3D excitation fields. In contrast, in tES techniques, stimulation involves the direct application of a weak current to the scalp via electrode pairs (Nitsche and Paulus, 2000; Priori et al., 1998). As a result, tES induces subthreshold polarization in cortical neurons, which is too weak to generate action potentials. However, by altering intrinsic neuronal excitability, tES can induce changes in residual membrane potential and postsynaptic activity in cortical neurons. This, in turn, can alter the spontaneous firing rate of neurons and modulate their response to incoming signals (Bindman et al., 1962, 1964, 1979; Creutzfeldt et al., 1962), thereby leading to alterations in synaptic efficacy.Typical applications of NIBS involve different types of protocols: TMS can be delivered as a single pulse at precise time (spTMS), as pairs of pulses separated by variable intervals, or as a series of stimuli in a regular or patterned protocol of repetitive TMS (rTMS) (for a complete classification, see Rossi et al., 2009). Generally, the ultimate effect of NIBS on the central nervous system depends on a long list of parameters (e.g., frequency, timing characteristics, intensity, coil / electrode geometry, current direction) delivered as part of the experimental procedure before (offline) or during (online) the task (e.g., Jacobson et al., 2011; Nitsche and Paulus, 2011; Sandrini et al., 2011). In addition, these factors are related to several anatomical variables (e.g., the nature and location of brain tissue, Radman et al., 2007) as well as the physiological state of the stimulated area / subject (e.g., sex and age, Landi and Rossini, 2010; Lang et al., 2011; Ridding and Ziemann, 2010) and cognitive state (e.g., Miniussi et al., 2010; Silvanto et al., 2008; Walsh et al., 1998).

[0107] Transcranial direct current stimulation (tDCS): Cranial electrotherapy (CES) is a form of non-invasive brain stimulation that applies small pulses of current across the head to treat various conditions such as anxiety, depression, and insomnia. See en.wikipedia.org / wiki / Cranial_electrotherapy_stimulation. Transcranial direct current stimulation (tDCS) is a form of neurostimulation that uses electrodes placed on the scalp to deliver a constant, low current to a targeted area of ​​the brain. It was initially developed to help patients with brain damage or psychotic symptoms such as major depressive disorder. tDCS appears to have some potential for treating depression. See en.wikipedia.org / wiki / Transcranial_direct-current_stimulation.

[0108] Research is underway on tDCS for accelerated learning. Mild electric shocks (typically 2 mA) are used to depolarize neuronal membranes, making cells more energized and responsive to input. Weisend, *Experimental Brain Research*, Vol. 213, p. 9 (DARPA), shows that tDCS accelerates the formation of new neural pathways during the time a person practices a skill. tDCS appears to induce a state of fluidity. Subjects' movements become more automated, they report calmly, focus their attention, and their performance improves immediately. (See Adee, Sally, “Zap your brain into the zone: Fast track to pure focus,” *New Scientist*, No. 2850, February 1, 2012, www.newscientist.com / article / mg21328501-600-zap-your-brain-into-the-zone-fast-track-to-pure-focus / ).

[0109] Reinhart, Robert MG., “Disruption and rescue of interareal theta phase coupling and adaptive behavior,” *Proceedings of the National Academy of Sciences* (2017): This paper provides evidence for a causal relationship between interareal theta phase synchronization in the frontal cortex and multiple components of adaptive human behavior. Reinhart’s results support the idea that precise timing of rhythmic group activities spatially distributed in the frontal cortex transmits information to direct behavior. Given that previous work has shown that phase synchronization can alter plasticity related to the timing of spikes, and that Reinhart’s findings show that stimulation of neural activity and behavior can last for 20-minute electrical stimulation periods, it is reasonable to hypothesize that externally modulated interareal coupling alters behavior by inducing neuroplastic changes in functional connectivity. Reinhart suggests that it may be possible to non-invasively intervene in the temporal coupling of distant rhythmic activities in the human brain to optimize (or prevent) the spikes in one area from having a postsynaptic effect on another area, thereby improving (or weakening) the interareal communication required for cognitive action control and learning. Furthermore, these neuroplastic changes in functional connectivity are induced by 0° phase, suggesting that induced synchronization does not require detailed calculations of communication delays between regions such as the MFC and IPFC to effectively alter behavior and learning. This aligns with work demonstrating that despite long axonal conduction delays between distant brain regions, 0° phase lag-induced theta-phase synchronization can occur between these regions and underlie meaningful cognitive and motor functions. It is also possible that subcortical or posterior third cortical regions with non-zero time lags interact with these two frontal regions to drive changes in goal-oriented behavior.

[0110] High-Definition tDCS: The City University of New York has developed High-Definition Transcranial Direct Current Stimulation (HD-tDCS) by introducing the 4×1 HD-tDCS montage. The 4×1 HD-tDCS montage allows for precise targeting of cortical structures. The area of ​​current is defined by the area of ​​a 4× loop, allowing for a smaller loop radius to increase focus. 4×1 HD-tDCS allows for monofocal stimulation, meaning the polarity of the central 1× electrode will determine the direction of neural modulation below the loop. This contrasts with conventional tDCS, where the need for a single anode and cathode always results in bidirectional modulation (even when using additional head electrodes). Therefore, 4×1 HD-tDCS provides the ability not only to select the cortical brain region to target but also to modulate the excitability of said brain region using a designed polarity, without considering back-electrode backflow.

[0111] Transcranial alternating current stimulation (tACS): TACS is a non-invasive technique that uses alternating current applied to the skin and skull to deliver potential neural oscillations in the brain at a specific frequency. See en.wikipedia.org / wiki / Transcranial_alternating_current_stimulation

[0112] U.S. Publication No. 20170197081 discloses a method for using transcutaneous electrical stimulation (TES) to transcutaneously stimulate nerves to alter or induce cognitive states.

[0113] Transcranial alternating current stimulation (tACS) is a non-invasive technique that uses alternating current applied to the skin and skull to carry potential neural oscillations in the brain at a specific frequency. See en.wikipedia.org / wiki / Transcranial_alternating_current_stimulation;

[0114] Transcranial random noise stimulation (tRNS): Transcranial random noise stimulation (tRNS) is a non-invasive brain stimulation technique and a form of transcranial electrical stimulation (tES). See en.wikipedia.org / wiki / Transcranial_random_noise_stimulation. Stimulation can include transcranial pulsed electrical stimulation (tPCS).

[0115] Transcranial magnetic stimulation (TMS): TMS is a method in which a varying magnetic field is used to induce an electrical current to flow through small areas of the brain via electromagnetic induction. During a TMS procedure, a magnetic field generator or "coil" is placed near the head of the person receiving treatment. The coil is connected to a pulse generator or stimulator that delivers varying electrical currents to the coil. TMS is used diagnostically to measure the connections between the central nervous system and skeletal muscles to evaluate damage in various disease states, including stroke, multiple sclerosis, amyotrophic lateral sclerosis (ALS), motor disorders, and motor neuron disease. There is evidence that TMS can be used to treat neuropathic pain, major depressive disorder, and other conditions.

[0116] PEMF: When a pulsed electromagnetic field (PEMF) is applied to the brain, it is called transcranial magnetic stimulation (TMS) and has been FDA-approved since 2008 for people who have failed to respond to antidepressants. Weak magnetic stimulation of the brain is often referred to as transcranial pulsed electromagnetic field (tPEMF) therapy. See en.wikipedia.org / wiki / Pulsed_electromagnetic_field_therapy.

[0117] Deep Brain Stimulation (DBS): Deep brain stimulation (DBS) is a neurosurgical procedure involving the implantation of a medical device called a neurostimulator (sometimes referred to as a "brain pacemaker"). This neurostimulator delivers electrical impulses to specific targets (nuclei) in the brain via implanted electrodes to treat motor and neuropsychiatric disorders. See en.wikipedia.org / wiki / Deep_brain_stimulation.

[0118] Transcranial pulsed ultrasound (TPU): Transcranial pulsed ultrasound (TPU) uses low-intensity, low-frequency ultrasound (LILFU) as a method for stimulating the brain. See en.wikipedia.org / wiki / Transcranial_pulsed_ultrasound;

[0119] Sensory stimulation: Time patterns of brain waves can be transmitted remotely using light, sound, or electromagnetic fields.

[0120] Optical Stimulation: The functional relevance of brain oscillations within the alpha frequency range (8–13 Hz) was repeatedly investigated using rhythmic visual stimulation. Two hypotheses exist regarding the origin of steady-state visually evoked potentials (SSVEPs) measured in EEG during rhythmic stimulation: entrainment of brain oscillations and superposition of event-related responses (ERPs). The entrainment, rather than superposition, hypothesis supports rhythmic visual stimulation as a means of manipulating brain oscillations, as superposition presupposes a linear summation of individual responses, independent of the ongoing brain oscillation. Participants were measured with rhythmic flashing light stimulation of varying frequencies and intensities by comparing the phase coupling of brain oscillations induced by non-rhythmic jitter stimulation, temporal variations, stimulation frequency, and intensity conditions. Phase coupling was found to be more pronounced with increasing stimulation intensity and stimulation frequencies closer to each participant's intrinsic frequency. Even in single sequences of SSVEPs, nonlinear characteristics (intermittent phase-locking) contradicted the linear summation of individual responses presupposed by the superposition hypothesis. Therefore, the evidence suggests that visual rhythmic stimulation entrains brain oscillations, thus validating rhythmic stimulation as a method of manipulating brain oscillations. See Notbohm A, Kurths J, Herrmann CS, “Modification of Brain Oscillations via Rhythmic Light Stimulation Provides Evidence for Entrainment but Not for Superposition of Event-Related Responses,” Front Hum Neurosci, February 3, 2016; 10:10. doi:10.3389 / fnhum.2016.00010.eCollection 2016.

[0121] It is also known that periodic visual stimulation can trigger epileptic seizures.

[0122] Cochlear implants: Cochlear implants are surgically inserted electronic devices that provide sound sensation to people with profound hearing loss or severe hearing impairment in both ears. See en.wikipedia.org / wiki / Cochlear_implant.

[0123] Vagus nerve stimulation (VNS): Vagus nerve stimulation (VNS) is a medical treatment involving the delivery of electrical impulses to the vagus nerve. It is used as adjunctive therapy for certain types of refractory epilepsy and treatment-resistant depression. See en.wikipedia.org / wiki / Vagus_nerve_stimulation.

[0124] Brain-to-brain interface: A brain-to-brain interface is a direct communication pathway between the brains of one animal and the brains of another. A brain-to-brain interface was used to help rats cooperate with each other. When the second rat failed to select the correct lever, the first rat noticed (without receiving a second reward) and generated a round of task-related neuronal activation, making the second rat more likely to select the correct lever. Human studies have also been conducted.

[0125] In 2013, researchers from the University of Washington were able to send brain signals to recipients using computer recordings and some form of magnetic stimulation, causing the recipients to hit trigger buttons on a computer game. In 2015, researchers connected multiple brains from both monkeys and rats to form an "organic computer." This hypothesized that by using a brain-to-brain interface (BTBI), animal brains could be used as their computational units to construct biological computers or brain networks. Initial exploratory work demonstrated that cooperation between rats in a distant cage was connected via signals from an array of cortical microelectrodes implanted in their brains. The rats were rewarded when a "decoding rat" that met the entry signal performed an action and when a "coding rat" emitted a signal that resulted in the "desired action." In the initial experiments, the rewarded action was pushing a lever at a distant location corresponding to a glowing LED near the master position. The rats needed about a month to adapt to the entry "brainwaves." When the decoding rat failed to select the correct lever, the encoding rat noticed (without receiving the expected reward) and generated a round of task-related neuronal firing, making it more likely that a second rat would select the correct lever.

[0126] In another study, computer readings were used to trigger some form of magnetic stimulation to send brain signals to the recipient based on the subject's brain activity, which caused the recipient to hit a trigger button on a computer game.

[0127] Brain-computer interface (BCI): Sometimes also called neural control interface (NCI), mind-computer interface (MMI), direct neural interface (DNI), or brain-computer interface (BMI), it is an enhanced or wired direct communication pathway between the brain and an external device. BCI differs from neuromodulation in that it allows bidirectional information flow. BCI is typically used to study, map, assist, enhance, or repair human cognitive or sensorimotor functions.

[0128] Synthetic telepathy, also known as technological telepathy or psychotronics (geeldon.wordpress.com / 2010 / 09 / 06 / synthetic-telepathy-also-known-as-techlepathy-or-psychotronics / ), describes a process that uses a computer through a brain-computer interface that intercepts and processes human thoughts (as electromagnetic radiation) and generates return signals that can be perceived by the human brain. Dewan, EM, “Occipital Alpha Rhythm Eye Position and Lens Accommodation”, Nature 214, 975-977 (June 3, 1967) demonstrated the mental control of alpha waves, which disconnects and reconnects them to produce Morse code representations of words and phrases solely through thought. US3,951,134 describes the use of radio to remotely monitor and alter brainwaves and references demodulation of waveforms to display them to an operator for viewing and transmission to a computer for further analysis. In 1988, Farwell, LA and Donchin, E., (1988), “Talking off the top of your head: towards a mental prosthesis utilizing event-related brain potentials,” EEG and Clinical Neurophysiology, 70(6), 510-523, describes a method for transmitting language information using a P300 response system, which matches observed information with combinations of what the subject is thinking. In this case, it is possible to select letters of the alphabet that the subject is thinking of. Theoretically, any input can be used and a dictionary can be constructed. US Patent No. 6,011,991 describes a method for remotely monitoring an individual's brainwaves for communication purposes and outlines a system for monitoring an individual's brainwaves via sensors and then specifically transmitting this information via satellite to a computer for analysis. This analysis will determine whether an individual is attempting to convey “a utterance, phrase, or thought that corresponds to a matching stored normalized signal.”

[0129] Methods of synthesizing telepathy can be categorized into two main groups: passive and active. Like sonar, receivers can actively or passively receive signals. Passive reception is the ability to "read" signals without first broadcasting them. This can be roughly equated to tuning into a radio station—the brain generates electromagnetic radiation that can be received at a certain distance. This distance is determined by the receiver's sensitivity, the filters used, and the required bandwidth. Most universities have limited budgets and use receivers such as EEGs (and similar devices). A related military technology is the TEMPEST surveillance system. Robert G. Malech's method requires broadcasting a modulated signal at the target location. This method uses an active signal that interferes with the modulation of the brain. Therefore, the original brainwaves can be inferred using the returned signal.

[0130] Computer-mediated mediation falls into two basic categories: interpretive and interactive. Interpretive mediation is a passive analysis of signals from the human brain. The computer "reads" the signals and then compares them to a database of signals and their meanings. Using statistical analysis and repetition, false positives decrease over time. Interactive mediation can operate in either a passive-active or active-active mode. In this case, passive and active refer to methods of reading and writing into the brain and whether or not broadcast signals are utilized. Interactive mediation can also be performed manually or through artificial intelligence. Manual interactive mediation involves a human operator generating return signals such as speech or images. AI-mediated mediation utilizes the subject's cognitive system to identify images, pre-speech, objects, sounds, and other artifacts, rather than developing AI routines to perform such activities. AI-based systems can incorporate natural language processing interfaces that generate sensations, mental impressions, humor, and dialogue to provide a mental picture of a computerized personality. Statistical analysis and machine learning techniques, such as neural networks, can be used.

[0131] In March 1991, ITV News Service reported on ultrasound transmitted over commercial radio (100 MHz) with the intent to instill a sense of despair in Iraqi soldiers. US 5,159,703 refers to "silent communication systems in which a non-auditory carrier, in a range of very low or very high audio frequencies or in the adjacent ultrasonic spectrum, is modulated by desired intelligence and propagates acoustically or vibratoryly at an amplitude or frequency, typically induced into the brain through the use of a loudspeaker, headphones, or a piezoelectric transducer."

[0132] It is known to analyze EEG patterns to extract indications of certain volitional activities (US Patent No. 6,011,991). This technology describes the use of a computer to match EEG recordings with stored normalized signals. The matched signals are then converted into corresponding references. The patent application describes methods: "systems capable of identifying specific nodes in an individual's brain whose activation affects characteristics such as appetite, hunger, thirst, communication skills"; and "devices mounted on a person (e.g., under the scalp) that can be activated in a predetermined manner or sequence to remotely activate one or more specifically identified brain nodes to elicit a predetermined sensation or response in the individual," without technical description of the embodiments. This patent also describes "monitoring brain activity by electroencephalography (EEG), magnetoencephalography (MEG), etc."

[0133] Brain entrainment: Also known as brainwave synchronization or neural entrainment, brain entrainment refers to the brain's ability to naturally synchronize its brainwave frequencies with the rhythms of periodic external stimuli, typically auditory, visual, or tactile. Brainwave entrainment techniques are used to induce various brain states, such as relaxation or sleep, by generating stimuli that occur at regular, periodic intervals to mimic the brain's electrical circulation in a desired state, thereby "training" the brain to consciously change its state. Reproduced acoustic frequencies, flashing light, or tactile vibrations are the most common examples of stimuli used to generate different sensory responses. The assumption is that hearing these beats at a certain frequency can induce a desired state of consciousness corresponding to specific neural activity. Neural activation patterns measured in Hz correspond to states of alertness such as focused attention and deep sleep.

[0134] Neural oscillations are rhythmic or repetitive electrochemical activities in the brain and central nervous system. These oscillations can be characterized by their frequency, amplitude, and phase. Neural tissue can generate oscillatory activity driven by mechanisms within individual neurons and by interactions between them. It can also modulate its frequency to synchronize with the periodic vibrations of external acoustic or visual stimuli. The functional roles of neural oscillations are not fully understood; however, they have been shown to be associated with emotional responses, motor control, and many cognitive functions, including information transfer, perception, and memory. Specifically, neural oscillations, especially theta activity, are extensively associated with memory functions, and the coupling between theta and gamma activity is considered crucial for memory functions, including episodic memory. Electroencephalography (EEG) has been most widely used to study neural activity generated by large groups of neurons known as neural swarms, including studies of changes in EEG profiles during sleep and wakefulness cycles. EEG signals change significantly during sleep and show a transition from faster frequencies to increasingly slower frequencies, suggesting a relationship between the frequency of neural oscillations and cognitive states, including consciousness and alertness.

[0135] The term "entrainment" is used to describe a common tendency in many physical and biological systems to synchronize their periodicity and rhythm through interaction. This tendency has been identified as particularly relevant to studies of sound and music, and especially to acoustic rhythms. The most common and familiar example of neuromotor entrainment to acoustic stimuli can be observed when feet or fingers spontaneously tap in time with the rhythmic beat of a song. Exogenous rhythmic entrainment occurring outside the body has been identified and documented in various human activities, including how people adjust the rhythm of their speech patterns to those of the subjects they are communicating with, and the rhythmic coherence of clapping by an audience. Even among unfamiliar groups, respiratory rates, dynamic and subtle expressive motor movements, and rhythmic speech patterns have been observed to synchronize and entrain in response to auditory stimuli such as a rhythmically coherent piece of music. Furthermore, synchronization of movement with repetitive tactile stimuli occurs in animals, including cats and monkeys, and humans, accompanied by shifts in electroencephalogram (EEG) readings. Examples of endogenous entrainment occurring in the body include the synchronization of the human circadian sleep-wake cycle with the 24-hour light-dark cycle, and the frequency-following response of humans to sounds and music.

[0136] Brain waves, or neural oscillations, share fundamental components with acoustic and light waves, including frequency, amplitude, and periodicity. Synchronized electrical activity in cortical neural groups can be synchronized in response to external acoustic or optical stimuli, and can also entrain or synchronize its frequency and phase with the frequency and phase of a specific stimulus. Brainwave entrainment is the colloquial form of this "neural entrainment," a term used to describe how the convergent frequency of oscillations generated by synchronized electrical activity in cortical neuronal groups can be tuned to synchronize with the periodic vibrations of external stimuli, such as sustained acoustic frequencies perceived as pitch, intermittent sounds perceived as rhythm, or regularly repeating patterns of rhythmically intermittent flashing light.

[0137] Changes in neural oscillations, verifiable by electroencephalography (EEG) measurements, are induced by listening to music. Listening to music can enhance and nutritionally modulate voluntary arousal, thereby increasing and decreasing arousal levels, respectively. Auditory stimulation of music has also been shown to improve immune function, promote relaxation, improve mood, and aid in stress relief.

[0138] Frequency-following response (FFR), also known as frequency-following potential (FFP), is a specific response to heard sounds and music. Neural oscillations adjust their frequency to match the rhythm of the auditory stimulus through this specific response. The use of sound intended to influence the frequency of cortical brain waves is called auditory actuation, through which the frequency of neural oscillations is "driven" to carry the rhythm of the sound source.

[0139] Baseline correction can be performed on event-related time-frequency measurements to account for pre-event baseline activity. Typically, the baseline period is defined by the average of values ​​within a time window preceding the time-locked event. At least four commonly used methods exist for baseline correction in time-frequency analysis. These methods involve various baseline normalizations. See [link to relevant documentation]

[0140] "Electroencephalograms (EEG) and functional magnetic resonance imaging (fMRI) have been used to study specific brain activity associated with different emotional states," Mauss and Robinson noted in their review paper. "Emotional state is likely to involve circuits rather than any brain region considered in isolation" (Mauss IB, Robinson MD (2009), Measures of Emotion: A Review. CognEmot, 23:209–237).

[0141] The amplitude, latency from stimulus, and covariance of each component can be examined, either in conjunction with a cognitive task (ERP) or not, and in the case of multiple electrode sites. Steady-state visually evoked potentials (SSVEPs) are measured using a continuously sinusoidally modulated flickering light, typically superimposed on a TV monitor displaying a cognitive task. Brain responses are measured in narrow bands containing stimulus frequencies. Amplitude, phase, and coherence (in the case of multiple electrode sites) may correlate with different parts of the cognitive task. Brain entrainment can be detected by EEG or MEG activity.

[0142] The entrainment hypothesis (Thut and Miniussi, 2009; Thut et al., 2011a, 2012) suggests the possibility of inducing specific oscillatory frequencies in the brain using external oscillatory forces (e.g., rTMS, and tACS). The physiological basis of oscillatory cortical activity lies in the timing of interacting neurons; when groups of neurons synchronize their firing activities, brain rhythms emerge, network oscillations are generated, and the basis for interactions between brain regions may develop (Buzsàki, 2006). The reported studies lack consistency and inference are limited due to the various experimental protocols used for brain stimulation, the descriptions of the actual protocols employed, and the limited controls. Therefore, despite various consensuses on the various aspects of the effects of extracranial brain stimulation, the results achieved have a degree of uncertainty depending on the details of the implementation. On the other hand, within specific experimental protocols, it is possible to obtain statistically significant and reproducible results. This suggests that feedback control may be effective for implementations that control stimulation for a given purpose; however, studies employing feedback control are lacking.

[0143] Different cognitive states are associated with different oscillation patterns in the brain (Buzsàki, 2006; Canolty and Knight, 2010; Varela et al., 2001). Thut et al. (2011b) directly tested the entrainment hypothesis through parallel EEG-TMS experiments. They first identified a single source and a single alpha frequency for parietal-occipital alpha modulation (MEG study). Then, they applied rTMS with a single alpha power while recording resting EEG activity. The results confirmed three predictions of the entrainment hypothesis: induction of a specific frequency after TMS, enhancement of oscillations during TMS stimulation due to synchronization, and phase alignment between the induced frequency and the ongoing activity (Thut et al., 2011b).

[0144] If associative stimulation is the general principle of human neural plasticity in which the timing and intensity of activation are key factors, then it is possible that the use of external forces with phase-synchronized / aligned oscillations within or between areas could also facilitate effective communication and associative plasticity (or alter communication). In this regard, it has been shown that associative cortical-cortical stimulation enhances the coherence of oscillatory activity between stimulated areas (Plewnia et al., 2008).

[0145] In coherent resonance (Longtin, 1997), adding a certain amount of noise to an excitable system results in the most coherent and skilled oscillatory response. The brain's response to externally timed embedded stimuli may lead to a reduction in phase variance and enhanced alignment (clustering) of phase components of ongoing EEG activity (entrainment, phase resetting), which can alter the signal-to-noise ratio and increase (or decrease) signal power.

[0146] If we consider neuronal activity in the brain as a loosely coupled set of oscillators, controllable parameters include the size of the neuronal region, oscillation frequency, resonant frequency or time constant, oscillator damping, noise, amplitude, coupling with other oscillators, and, of course, external influences such as stimulation and / or power loss. In the human brain, pharmacological interventions can be significant. For example, altering stimulant drugs, such as caffeine, neurotransmitter release and reuptake, and neurotransmission, can affect the operation of neural oscillators. Similarly, subthreshold external stimuli (including DC, AC, and magneto-electromagnetic effects) can also affect the operation of neural oscillators.

[0147] Phase resetting or shifting can synchronize inputs and facilitate communication, ultimately benefiting Hebbian plasticity (Hebb, 1949). Therefore, rhythmic stimulation can induce a statistically higher degree of coherence in spiky neurons, which promotes (or inhibits) the induction of specific cognitive processes. Here, the perspective is slightly different (coherent resonance), but the underlined mechanism is similar to the mechanism described so far (stochastic resonance), and another key factor is the repetition of the stimulus at a specific rhythmicity.

[0148] In the 1970s, British biophysicist and psychobiologist C. Maxwell Cade monitored the brainwave patterns of advanced meditators and 300 of their students. He discovered that the specific brainwave patterns of the most advanced meditators differed from those of the rest of his students. He noted that these meditators exhibited highly active alpha brainwaves, accompanied by beta, theta, and even delta waves, which had about half the amplitude of the alpha waves. See Cade, “Awake Mind: Biofeedback and the Development of Higher Consciousness” (Dell, 1979). Anna Wise extended Cade's research and found that the brainwave patterns of high achievers, including composers, inventors, artists, athletes, dancers, scientists, mathematicians, and CEOs and presidents of large companies, differed from those of average performers and exhibited a specific balance among beta, alpha, theta, and delta brainwaves, with the alpha wave having the strongest amplitude. See Anna Wise, “The High-Performance Mind: Mastering Brainwaves for Insight, Healing, and Creativity.”

[0149] The entrainment hypothesis seems plausible due to the demonstrated characteristics of the EEG response to a single TMS pulse, the spectral composition of which resembles the spontaneous oscillations of the stimulated cortex. For example, TMS triggers alpha waves in the "resting" visual cortex (Rosanova et al., 2009) or motor cortex (Veniero et al., 2011), which are the natural frequencies of both types of cortex at rest. Using the entrainment hypothesis, the noise generation framework shifts to a more complex and extended level, where noise is synchronized with ongoing activity. Nevertheless, the model interpreting the results will not change; the stimulus will interact with the system, and the final outcome will depend on the introduction or modification of the noise level. The entrainment hypothesis makes a clear prediction of the frequency involvement and the possibility of induced phase alignment in online repetitive TMS paradigms, namely, resetting ongoing brain oscillations via external spTMS (Thut et al., 2011a, 2012; Veniero et al., 2011). The entrainment hypothesis is superior to localization methods in gaining knowledge about how the brain works, rather than where and how a single process occurs. TMS pulses can be phase-aligned with the natural, ongoing oscillations of the target cortex. When additional TMS pulses are delivered in sync with the phase-aligned oscillations (i.e., at the same frequency), further phase alignment occurs, which will cause the oscillations of the target area to resonate with the TMS training. Therefore, when TMS is frequency-tuned to the potential brain oscillations, entrainment may be expected (Veniero et al., 2011).

[0150] Binaural beats: Binaural beats are auditory brainstem responses originating in the superior olivary nucleus of each hemisphere. They are generated by the interaction of two distinct auditory impulses originating from opposing ears, below 1000 Hz, with a frequency difference between 1 and 30 Hz. For example, if a pure tone of 400 Hz is presented to the right ear and simultaneously to the left ear a pure tone of 410 Hz, a standing wave with amplitude modulation of 10 Hz will occur as the two waveforms engage in phase and out-of-phase within the superior olivary nucleus, representing the difference between the two tones. This binaural beat is inaudible in the ordinary sense of speech (the human hearing range is 20–20,000 Hz). It is perceived as an auditory beat and theoretically can be used to entrain specific neural rhythms—cortical potentials that tend to entrain or resonate with the frequency of external stimuli—through frequency-following response (FFR). Therefore, it is theoretically possible to utilize specific binaural beat frequencies as a mind-controlling technique to entrain specific cortical rhythms. Binaural beats appear to be associated with frequency-following responses in the brain via electroencephalography (EEG).

[0151] The uses of audio embedded with binaural beats, mixed with music or various popping or background sounds, are diverse. These range from relaxation and meditation to stress reduction, pain management, improved sleep quality, reduced sleep needs, superlearning, enhanced creativity and intuition, telepathy, and extracorporeal experiences and lucid dreaming. Audio embedded with binaural beats is often combined with various meditation techniques, as well as positive affirmations and visualizations.

[0152] When two signals of different frequencies are presented (one signal for each ear), the brain detects the phase difference between these signals. In a natural state, the detected phase difference provides directional information. When these phase differences are heard using stereo headphones or speakers, the brain processes this anomalous information differently. Perceptual integration of the two signals occurs, resulting in the perception of a third “beat” frequency. As the two different input frequencies engage in phase and out-of-phase engagement, the difference between the signals fluctuates. Due to these increasing and decreasing differences, amplitude-modulated standing waves—binaural beats—are heard. At the frequency of the difference between the two auditory inputs, the binaural beat is perceived as a undulating rhythm. Evidence suggests that the binaural beat originates in the superior olivary nucleus of the brainstem (the first site of contralateral integration in the auditory system). Studies also indicate that the frequency-following response originates in the inferior colliculus. This activity is transmitted to the cortex, where scalp electrodes can record it. The binaural beat can be easily heard at low frequencies (<30 Hz), a characteristic of the EEG spectrum.

[0153] Synchronized brainwaves have long been associated with meditative and hypnotic states, and audio embedded with binaural beats has the ability to induce and enhance this state of consciousness. The reason is physiological. Each ear is “hardwired” (so to speak) to the two hemispheres of the brain. Each hemisphere has its own olivary nucleus (sound processing center) that receives signals from each ear. To maintain this physiological structure, when binaural beats are perceived, there are actually two standing waves of equal amplitude and frequency, one for each hemisphere. Thus, there are two separate standing waves that tether a portion of each hemisphere to the same frequency. Binaural beats appear to contribute to hemispheric synchronization in the indicated meditative and hypnotic states of consciousness. Brain function is also enhanced by increasing transsol communication between the left and right hemispheres of the brain. en.wikipedia.org / wiki / Beat_(acoustics)#Binaural_beats.

[0154] Time-frequency analysis: Brian J. Roach and Daniel H. Mathalon, “Event-related EEG time-frequency analysis: an overview of measures and analysis of early gamma band phaselocking in schizophrenia,” *Schizophrenia Bulletin*, USA, 2008; 34:5:907-926, describe a mechanism for EEG time-frequency analysis. Fourier and wavelet transforms (and their inverses) can be performed on the EEG signal.

[0155] There are many methods for time-frequency decomposition of EEG data, including Short-Term Fourier Transform (STFT) (Gabor D., "Theory of Communication," *J. Inst. Electr. Engrs.*, 1946; 93: 429-457), Continuous (Daubechies I., "Ten Lectures on Wavelets," Philadelphia, PA: Society for Industrial and Applied Mathematics; 1992: 357, 21; Combes JM, Grossmann A, Tchamitchian P, "Wavelets: Time-Frequency Methods and Phase Space-Proceedings of the International Conference," December 14-18, 1987, Marseille, France) or Discrete (Mallat) SG. “A theory for multiresolution signal decomposition: the wavelet representation” IEEE Trans. Pattern Anal Mach Intell 1989; 11: 674-693. Wavelet transform, Hilbert transform (Lyons RG, “Understanding Digital Signal Processing” 2nd ed., Upper Saddle River, NJ: Prentice Hall PTR; 2004: 688) and matching pursuit (Mallat S, Zhang Z., “Matching pursuits with time-frequency dictionaries”, IEEE Trans. Signal Proc. 1993; 41(12): 3397-3415). The prototype analysis system can be implemented using, for example, MATLAB (www.mathworks.com / products / wavelet.html) with a wavelet toolbox.

[0156] Single instruction multiple data processors, such as graphics processing units that include the NVIDIA CUDA environment or the AMD Firepro high-performance computing environment, are known and can be used for general computing or specifically for data matrix transformations.

[0157] Statistical analysis can be represented in a form that allows for parallelization, and the statistical analysis can be efficiently implemented using various parallel processors, a common form of which is the SIMD (Single Instruction Multiple Data) processor found in a typical graphics processing unit (GPU).

[0158] Artificial neural networks have been used to analyze EEG signals.

[0159] Principal Component Analysis (PCA): Principal Component Analysis (PCA) is a statistical procedure that uses orthogonal transformations to convert a set of observations of potentially related variables into a set of linearly uncorrelated variables called principal components. If there are n observations for variable p, the number of distinct principal components is min(n-1, p). This transformation is defined such that the first principal component has the largest possible variance (that is, it takes into account as much variability in the data as possible), and each subsequent component, in turn, has the highest variance under the constraint of orthogonality with the preceding components. The resulting vectors are uncorrelated orthogonal basis sets. PCA is sensitive to the relative scaling of the original variables. PCA is the simplest of the multivariate analyses based on true eigenvectors. Typically, its operation can be viewed as revealing the internal structure of the data in a way that best explains the variance in the data. If a multivariate dataset is visualized as a set of coordinates in a high-dimensional data space (one axis per variable), PCA can provide the user with low-dimensional slivers, i.e., projections onto the object when viewed from its most informative perspective. This is done by using only the first few principal components, thus reducing the dimensionality of the transformed data. PCA is closely related to factor analysis. Factor analysis typically incorporates more domain-specific assumptions about the underlying structure and solves for the eigenvectors of slightly different matrices. PCA is also related to canonical correlation analysis (CCA). CCA defines a coordinate system that optimally describes the cross-covariance between two datasets, while PCA defines a new orthogonal coordinate system that optimally describes the variance in a single dataset. See en.wikipedia.org / wiki / Principal_component_analysis.

[0160] The general model for confirmatory factor analysis is represented as x = α + Λξ + ε. The covariance matrix is ​​represented as E[(x-μ)(x-μ)′] = ΛΦΛ′ + Θ. If the residual covariance matrix Θ = 0 and the correlation matrix Φ = 1 among the latent factors, then factor analysis is equivalent to principal component analysis, and the resulting covariance matrix is ​​simplified to Σ = ΛΛ′. When there are p number of variables and all p components (or factors) are extracted, this covariance matrix can alternatively be represented as Σ = DΛD′, or Σ = λDΑD′, where D = an orthogonal matrix of n × p eigenvectors, and Λ = λA, a p × p matrix of eigenvalues, where λ is a scalar and A is a diagonal matrix whose elements are proportional to the eigenvalues ​​of Σ. The following three components determine the geometric characteristics of the observed data: λ parameterizes the volume of the observation, D indicates the orientation, and A represents the shape of the observation.

[0161] When population heterogeneity is explicitly assumed to be in a model-based clustering analysis, the observed covariance matrix is ​​decomposed into the following general form.

[0162] Where λ k Parameterize the volume of the k-th cluster, D k Indicates the orientation of the cluster, and A k This indicates the shape of the cluster. The subscript k indicates that each component (or cluster) can have a different volume, shape, and orientation.

[0163] Suppose we substitute the value The mean matrix and covariance matrix of the random vector X are μ X and ∑ X λ1>λ2>…>λ m >0 is ∑ X The ordered eigenvalues ​​of ∑ X The i-th eigenvalue means that its i-th eigenvalue is the largest. Similarly, when vector α i Corresponding to ∑ X When the i-th eigenvalue is ∑ X The i-th eigenvector. To derive the form of the principal components (PC), consider maximizing... The optimization problem is affected by Constraints. The Lagrange multiplier method was used to solve this problem.

[0164]

[0165]

[0166] Because -φ1 is ∑ X The eigenvalues, where α1 is the corresponding normalized eigenvector, are then determined by the selected α1. Maximize to ∑ X The first eigenvector. In this case... The first PC is named X, α1 is the vector of coefficients of z1, and var(z1) = λ1.

[0167] In order to find the second Maximize It is constrained by z2, which is unrelated to z1. Because Therefore, this problem is set to the maximum value. by Constraints, and We will still use the Lagrange multiplier method.

[0168]

[0169]

[0170]

[0171]

[0172] Because -φ2 is ∑ X The eigenvalues, where α2 is the corresponding normalized eigenvector, are then determined by the selected α2. Maximize to ∑ X The second eigenvector. In this case, The second PC, named X, has α2 as a vector of coefficients of z2, and var(z2) = λ2. Continuing in this manner, it can be shown that the i-th PC... It is by α i Select as ∑ X The i-th eigenvector is constructed and has variance λ. i A key finding of PCA is that the principal components are sets of linear functions only from the original data, which are uncorrelated and have orthogonal coefficient vectors.

[0173] For any positive integer p ≤ m, such that B = [β1, β2, ..., β] p [ is a proper m×p matrix with orthogonal columns, i.e., And Y = B T X. By taking B = [α1, α2, ..., α...] p Maximize the trace of the covariance matrix of Y, where α i It is ∑ X The i-th eigenvector. Because ∑ X It is symmetric with all distinct eigenvalues, therefore {α1, α2, ..., α...} m} is an orthonormal basis, where α iFor ∑ X The i-th eigenvector of B, and the columns of B can be represented as Since i = 1, ..., p, therefore B = PC, where P = [α1, ..., α m ], C = {ci j} is an m×p matrix. Then, P T ∑ X P = Λ, where Λ is its k-th diagonal element λ. k The diagonal matrix of Y, and the covariance matrix of Y is

[0174]

[0175] in It is the i-th row of C. Therefore,

[0176]

[0177] Because C T C = B T PP T B = B T B = I, therefore Furthermore, the columns of C are orthogonal. Using the Gram-Schmidt method, C can be extended to D such that D makes its columns orthogonal. The orthonormal basis of is given, and contains C as its first p columns. D is square-shaped, therefore an orthogonal matrix, and its columns are given as... Another orthonormal basis. A row of C is part of a column of D, therefore i = 1, ..., m. Consider constraints. And the goal If (i = 1, ..., p) and (i = p + 1, ..., m), then maximize trace(∑ Y When B = [α1, α2, ..., α] p When [the value is], C is directly calculated to be the value other than c. ii =1, a matrix of all zeros except i=1,...,p. This satisfies the maximization condition. In fact, by taking B = [γ1,γ2,...,γ...] p ], where {γ1,γ2,...,γ p} is a span{α1,α2,...,α p Any orthonormal basis of the subspace of Y also satisfies the maxima condition, thus obtaining the same trace of the covariance matrix of Y.

[0178] Suppose we want to approximate a random vector X by projecting it onto a subspace spanning columns B, where B = [β1, β2, ..., β]. p ] is a column with orthogonal structure, i.e. an m×p matrix. If Let B be the residual variance of each component of X. Then, when B = [α1, α2, ..., α...] p Minimize when Where {α1, α2, ..., α p} is ∑ X The first p-eigenvector. In other words, when B = [α1, α2, ..., α... p When minimizing X-BB T The trace of the covariance matrix of X. When used as a common processing step in data analysis methods, E(X) = 0, this property states that when B = [α1, α2, ..., α...], the trace of the covariance matrix of X. p When [the goal is to minimize E||X-BB], minimize E||X-BB. T X|| 2 .

[0179] The projection of a random vector X onto a subspace spanning columns B is Then the residual vector is ε = X - BB. T X, which has a covariance matrix

[0180] ∑ ε =(I-BB) T )∑ X (I-BB T ).Then,

[0181]

[0182] In addition, it is known that:

[0183] trace(∑ X BB T =trace(BB) T ∑ X ) = trace(B T ∑ X B)

[0184] trace(BB T ∑ X BB T ) = trace(B T ∑ X BB T B) = trace(B T ∑ X B).

[0185] The final equation stems from the fact that B has an orthogonal column. Therefore,

[0186]

[0187] To minimize It is sufficient to maximize trace(B) T ∑ X B). This can be achieved by choosing B = [α1, α2, ..., α]. p To complete, where {α1, α2, ..., α] p} is as above ∑ X The first p-feature vector.

[0188] See Pietro Amenta and Luigi D'Ambra, "Generalized Constrained Principal Component Analysis with External Information," (2000). Assume that the principal components are at orders (n×p1), ..., (n×p...). K ) and (n×q1),...,(n×q S The matrix X k (k = 1, ..., K) and Y s Data on S standard variables with K explanatory variables and n statistical units were collected in the range (s = 1, ..., S). Assume that, without loss of generality, the identifier X... k and Y s D n =diag(1 / n) is a unified matrix of spatial measures of variables and a weight matrix of statistical units. Furthermore, it is assumed that X k and Y s With weight D n Centered on.

[0189] Let X = [X1|...|X] respectively. K ] and Y = [Y1|...|Y S Let the order be (n×∑) k p k ) and (n×∑ s q s The columns of the K and S matrices are connected. Also, W... Y =YY', while indicating v k For each X k The coefficient vector of the linear combination (p) k ,1), such that z k =X k v k Make C k For dimension p k×m(m≤p k The matrix is ​​associated with the external information explanatory variables of set k.

[0190] Generalized CPCA with external information (GCPCA) (Amenta, D'Ambra, 1999) lies in finding the K coefficient vector v. k (or in the same way, K-linear combination z) k Simultaneously constrained by C' k v k =0 constraint, such that:

[0191]

[0192] Or, in an equivalent way,

[0193]

[0194] Where A = Y'X, B = diag(X′1X1,…,X') K X K ), C' = [C′1|...|C' k ],v'=(v1'|...|v k ') and f = B 0.5 v, where

[0195] The biggest problem with constrained proofs is the standard. An extension of this (Sabatier, 1993), where a larger set of standard variables has extrinsic information. The solution to this constrained maximum problem leads to the solution of the characteristic equation.

[0196] (P X -P XB-1C W Y g=λg

[0197] Where g = Xv, Is with The oblique projection operator associated with the straight and decomposition.

[0198]

[0199] In respectively And P C =C(C'B -1 C) -1 C'B -1 In the case of I and B -1 To cross matrix X k The orthogonal projection operator of the column subspace of C. Furthermore, It is to cross the matrix XB -1C. Orthogonal projection operator of the subspace of the column. Starting with relation

[0200]

[0201] (It is from the expression (IP) C )X'W Y g = λBv (obtained), coefficient vector v k And maximizing the linear combination z of (1) k =X k v k They can be obtained through relationships and Provided.

[0202] The solution eigenvector g can be written as a linear combination z k The sum: g = ∑ k X k v k Note that, according to Sturm's theorem, the eigenvalues ​​associated with the feature system are less than or equal to those in the GCPCA feature system:

[0203] Spatial principal component analysis: Let J(t,i; α,s) be the current density of voxel i, as estimated by LORETA, at time frame t after stimulus initiation for subject s under condition α. ​​Let area:Voxel→fBA be a function that assigns a corresponding fBA b∈fBA to each voxel i∈voxel. In the first preprocessing step, the average current density over each FBA is calculated for each subject s.

[0204]

[0205] Where N b It represents the number of voxels in fBA b for subject s under condition α.

[0206] In the second analysis phase, the average current density x(t,b;α,s) from each fBA b under condition α for each subject s is subjected to spatial PCA analysis with correlation matrix and variance maxima rotation.

[0207] In this study, spatial PCA used the fBA defined above as a variable sampled along the time epochs (0-1000 ms; 512 time frames) of the EEG samples, and estimated the inverse solution. Spatial matrices (each matrix was set to size b×t = 36×512 elements) were collected for each subject and condition and subjected to PCA analysis, including covariance matrix calculation; eigenvalue decomposition and variance maximization rotation to maximize factor loading. In other words, in the spatial PCA analysis, the average current density per subject under each condition was approximated as...

[0208]

[0209] Here x(t;α,s)∈R 36 In the case where x is a vector representing the time-dependent activation of fBA, x0(α,s) is its average activation, and x k (α,s) and c k These are the principal components and their corresponding coefficients (factor loads) calculated using principal component analysis.

[0210] Nonlinear Dimensionality Reduction: High-dimensional data, meaning data that requires more than two or three dimensions to represent, can be difficult to interpret. One simplification approach is to assume that the data of interest resides on an embedded nonlinear manifold within a higher-dimensional space. If the manifold's dimension is low enough, the data can be visualized in a low-dimensional space. Nonlinear methods can be broadly categorized into two groups: those that provide mappings (from a high-dimensional space to a low-dimensional embedding or vice versa), and those that only provide visualizations. In the context of machine learning, mapping methods can be viewed as a preliminary feature extraction step, after which pattern recognition algorithms are applied. Typically, those that only provide visualizations are based on proximity data (i.e., distance measurements). Relevant linear decomposition methods include Independent Component Analysis (ICA), Principal Component Analysis (PCA) (also known as Karhunen-Loève Transform-KLT), Singular Value Decomposition (SVD), and Factor Analysis.

[0211] Self-organizing graphs (SOMs, also known as Kohonen graphs) and their probabilistic variant generating topology maps (GTMs) use point representations in an embedded space to form latent variable models based on a nonlinear mapping from the embedded space to a higher-dimensional space. These techniques are related to work on density networks, which are also based on roughly the same probabilistic models.

[0212] Principal curves and manifolds provide a natural geometric framework for nonlinear dimensionality reduction and extend the geometric interpretation of PCA by explicitly constructing embedded manifolds and encoding them using standard geometric projections. How to define the “simplicity” of a manifold is problem-dependent; however, it is typically measured by the manifold’s intrinsic dimension and / or smoothness. Generally, the principal curve is defined as a solution to the optimization problem. The objective function includes the quality of the data approximation and some penalty terms for the manifold’s curvature. Initial approximations of the principal curve are generated via linear PCA, Kohonen’s SOM, or an autoencoder. Elastic graph methods provide an expectation-maximization algorithm for principal curve learning, where a quadratic energy functional is minimized at the “maximization” step.

[0213] An autoencoder is a feedforward neural network trained to approximate the same function. That is, it is trained to map from a vector of values ​​to the same vector. When used for dimensionality reduction, one of the hidden layers in the network is restricted to contain only a small number of network units. Therefore, the network must learn to encode vectors into a small dimension and then decode them back into the original space. Thus, the first half of the network is a model mapping from a high-dimensional space to a low-dimensional space, and the second half is a model mapping from a low-dimensional space to a high-dimensional space. Although the concept of autoencoders is old, it was only recently possible to train deep autoencoders using restricted Boltzmann machines and stacked denoising autoencoders. Related to autoencoders is the NeuroScale algorithm, which uses a stress function inspired by multidimensional scaling and Sammon mapping (see below) to learn a nonlinear mapping from a high-dimensional space to an embedded space. The mapping in NeuroScale is based on a radial basis function network.

[0214] Gaussian Process Latent Variable Model (GPLVM) is a probabilistic dimensionality reduction method that uses Gaussian processes (GPs) to find lower-dimensional nonlinear embeddings in high-dimensional data. It is an extension of the probabilistic formula for PCA. The model is defined probabilistically, and then latent variables are marginalized, with parameters obtained by maximizing the probabilities. Like kernel PCA, it uses a kernel function to form a nonlinear mapping (in the form of a Gaussian process). However, in GPLVM, the mapping is from the embedding (latent) space to the data space (like density networks and GTMs), whereas in kernel PCA it is in the opposite direction. It was initially proposed for visualization of high-dimensional data but has been extended to construct shared manifold models between two observation spaces. GPLVM and many variants have been proposed specifically for human motion modeling, such as inversely constrained GPLVM, GP Dynamic Model (GPDM), balanced GPDM (B-GPDM), and topologically constrained GPDM. To capture the coupling effects of pose and gait manifolds in gait analysis, multilayer joint gait-pose manifolds have been proposed.

[0215] Curve Component Analysis (CCA) seeks configurations of points in the output space that preserve the original distances as much as possible while focusing on small distances in the output space (unlike Sammon's mapping, which focuses on small distances in the original space). It should be noted that CCA, as an iterative learning algorithm, actually begins by focusing on large distances (like Sammon's algorithm) and then gradually shifts the focus to small distances. Small distance information will rewrite large distance information, provided a trade-off is made between the two. The stress function of CCA relates to the sum of right Bregman divergences. Curve Distance Analysis (CDA) trains a self-organizing neural network to fit a manifold and attempts to preserve geodesic distances in its embeddings. It is based on "Curve Component Analysis" (which extends Sammon's mapping) but instead uses geodesic distances. Differential Homeomorphism Dimension Reduction, or Diffeomap learning, feeds data into smooth differential homeomorphic mappings of lower-dimensional linear subspaces. The method addresses the smoothing of time-indexed vector fields, such that the flow along the field starting at a data point terminates in a lower-dimensional linear subspace, thus attempting to preserve pairwise differences under both forward and inverse mappings.

[0216] A common algorithm for manifold learning is kernel principal component analysis (kernel PCA), which is a combination of PCA and the kernel trick. PCA begins by computing the covariance matrix of an M×n matrix X. It then projects the data onto the first k eigenvectors of that matrix. In contrast, KPCA begins by computing the covariance matrix of the data after transforming it to a higher-dimensional space. It then projects the transformed data onto the first k eigenvectors of that matrix, just like PCA. It uses the kernel trick to eliminate a large amount of computation, allowing the entire process to be performed without actually computing Φ(x). Of course, Φ must be chosen such that it has a known corresponding kernel.

[0217] Laplacian eigenmaps (also known as locally linear eigenmaps, LLE) are a concrete instance of kernel PCA performed by constructing a kernel matrix relevant to the data. KPCA has an internal model, allowing it to map points to embeddings that are not available during training. Laplacian eigenmaps use spectral techniques to perform dimensionality reduction. This technique relies on the fundamental assumption that the data resides in a low-dimensional manifold within a high-dimensional space. While this algorithm cannot embed points outside the sample points, techniques based on reproducing kernel Hilbert space regularization exist to add this capability. Such techniques can also be applied to other nonlinear dimensionality reduction algorithms. Traditional techniques, such as principal component analysis, do not consider the inherent geometry of the data. Laplacian eigenmaps construct a graph based on neighborhood information from the dataset. Each data point acts as a node in the graph, and connectivity between nodes is controlled by proximity to neighboring points (e.g., using the k-nearest neighbor algorithm). The resulting graph can be viewed as a discrete approximation of a low-dimensional manifold within a high-dimensional space. Minimizing the cost function based on the graph ensures that points close to each other on the manifold are mapped close to each other in the low-dimensional space, thus preserving local distances. The Laplace-Beltrami operator acts as the embedding dimension for the eigenfunctions on the manifold because, under mild conditions, this operator has a countable spectrum, which is the basis for square-integrable functions on the manifold (compare to Fourier series on the unit circle manifold). Attempts to place the Laplace eigenmap on a solid theoretical foundation have been somewhat successful, as the graph Laplace matrix has shown convergence to the L-Laplace-Beltrami operator under certain unconstrained assumptions as the number of points reaches infinity. In classification applications, low-dimensional manifolds can be used to model data classes that can be defined based on observed sets of instances. Each observed instance can be described by two independent factors called “content” and “style,” where “content” is an invariant factor related to the nature of the class, and “style” represents the variation of the class among instances. Unfortunately, when the training data consists of instances that vary greatly in style, the Laplace eigenmap may not produce a coherent representation of the class of interest. In the case of classes represented by multivariate sequences, a structured Laplacian feature map has been proposed to overcome this problem by adding additional constraints within the neighborhood information graph of the Laplacian feature map to better reflect the inherent structure of the class. More specifically, the graph is used to encode the sequential structure of the multivariate sequences and minimize both style variation, data points from different sequences, and even proximity within sequences (if they contain repetitions). Proximity is detected using dynamic time warping by finding correspondences between and within segments of multivariate sequences that exhibit high similarity.

[0218] Like LLE, Hessian LLE is also based on sparse matrix techniques. Compared to LLE, it tends to yield higher quality results. Unfortunately, its computational complexity is very high, making it unsuitable for manifolds with large sampling. It does not have an internal model. Modified LLE (MLLE) is another LLE variant that uses multiple weights in each neighborhood to address the local weight matrix modulation problem that causes LLE mapping distortion. MLLE produces robust projections similar to Hessian LLE, but without significantly additional computational cost.

[0219] Manifold alignment leverages the assumption that heterogeneous datasets generated by a similarity generation process will share similar underlying manifold representations. Correspondences are recovered by learning projections from each original space to the shared manifold, allowing knowledge transfer from one domain to another. Most manifold alignment techniques consider only two datasets, but the concept extends to any number of initial datasets. Diffusion mapping utilizes the relationship between thermal diffusion and random walks (Markov chains); an analogy is drawn between a diffusion operator on the manifold and a Markov transition matrix that operates on a function defined on the graph, whose nodes are sampled from the manifold. Relational perspective mapping is a multi-dimensional scaling algorithm. The algorithm finds the configuration of data points on the manifold by simulating a multi-particle dynamics system on a closed manifold, where data points are mapped to particles, and the distances (or dissimilarity) between data points represent repulsive forces. As the size of the manifold gradually grows, the multi-particle system gradually cools and converges to a configuration reflecting the distance information of the data points. Local Tangent Space Alignment (LTSA) is based on the intuition that when the manifold is correctly unfolded, all tangent hyperplanes of the manifold will become aligned. It begins by computing the k nearest neighbors of each point. It computes the tangent space at each point by calculating the top d principal components in each local neighborhood. Then, it optimizes to find an embedding aligned with the tangent space. Local multidimensional scaling is performed in the local regions, and then convex optimization is used to fit all the pieces together.

[0220] Maximum variance expansion was previously known as semidefinite embedding. The intuition of this algorithm is that the variance above a point is maximized when the manifold is correctly expanded. This algorithm also begins by finding the k nearest neighbors for each point. It then attempts to solve the problem of maximizing the distance between all non-neighboring points, constrained to preserve the distance between neighboring points. Nonlinear PCA (NLPCA) uses backpropagation to train a multilayer perceptron (MLP) to fit the manifold. Unlike typical MLP training, which only updates the weights, NLPCA updates both the weights and the input. That is, both the weights and the input are treated as latent values. After training, the latent input is a low-dimensional representation of the observed vectors, and the MLP maps from this low-dimensional representation to a high-dimensional observation space. Manifold sculpting uses progressive optimization to find embeddings. Like other algorithms, it computes the k nearest neighbors and attempts to find embeddings that preserve relations in the local neighborhood. It slowly scales the variance out to the higher dimensions while simultaneously adjusting the points in the lower dimensions to preserve those relations.

[0221] Ruffini (2015) discussed a multichannel transcranial current stimulation (tCS) system that offers the possibility of optimized, non-invasive brain stimulation guided by EEG. A brain model with a realistic tCS electric field was used to create a positive “guided field” matrix, from which an EEG inverter was employed for cortical mapping. Starting with the EEG, a 2D cortical surface dipole field was defined that could generate the observed EEG electrode voltages.

[0222] Schestatsky et al. (2017) discussed transcranial direct current stimulation (tDCS), which uses a constant current to stimulate the scalp by inducing a shift in neuronal membrane excitability, leading to secondary changes in cortical activity. Although tDCS has a largely neuromodulatory effect on the underlying cortex, its effects can also be observed in distant neural networks. Accompanying EEG monitoring of the effects of tDCS can provide valuable information about its mechanisms. EEG findings can be an important surrogate marker of the effects of tDCS and can therefore be used to optimize its parameters. This combined EEG-tDCS system can also be used for the preventative treatment of neuropathies (such as epilepsy) characterized by abnormal peaks in cortical excitability. This system would form the basis for non-invasive closed-loop devices. tDCS and EEG can be used in parallel.

[0223] EEG analysis methods have emerged, in which event-related changes in EEG dynamics are analyzed within a single event-related data record. See Allen D. Malony et al., “Computational Neuroinformatics for Integrated Electromagnetic Neuroimaging and Analysis”, PAR-99-138. Pfurtscheller reported a method for quantifying the average transient inhibition of activity in the alpha band (approximately 10-Hz) after stimulation. Event-related desynchronization (ERD, decreased spectral amplitude) and event-related synchronization (ERS, increased spectral amplitude) were observed in various narrow bands (4-40Hz), which systematically depended on task and cognitive state variables as well as stimulation parameters. Makeig (1993) reported event-related changes across the entire EEG spectrum, resulting in a 2-D time / frequency metric he termed Event-Related Spectral Perturbation (ERSP). This method avoids the problems associated with analysis of prior narrow bands, as the band of interest for analysis can be based on salient features of the complete time / frequency transformation. Rappelsburger et al. introduced Event-Related Coherence (ERCOH). Several other signal processing metrics have been tested for use with EEG and / or MEG data, including dimensionality measures based on chaos theory and bispectral methods. The use of neural networks for EEG pattern recognition applied to clinical and practical problems has been proposed, but these methods are generally not employed for explicit modeling of the neurodynamics involved. Neurodynamics, as a method for physical therapy, involves the mobilization of the nervous system. This approach relies on influencing pain and other neurophysiological processes through mechanical manipulation of neural tissue and non-neural structures surrounding the nervous system. The body provides a mechanical interface to the nervous system through the musculoskeletal system. With movement, the musculoskeletal system applies uneven stress and movement within neural tissue, depending on local anatomy and mechanical properties, as well as the pattern of body movement. This activates a range of mechanical and physiological responses within the neural tissue. These responses include changes in nerve gliding, compression, elongation, tension, and intraneural microcirculation, axonal transport, and impulse communication.

[0224] The increasing utilization and focus on a growing number of EEG (and MEG) channels directly leads to the question of how to combine data from different channels. For this purpose, Donchin advocates using linear factor analysis methods based on principal component analysis (PCA). Temporal PCA assumes that the temporal process for activating each resulting component is identical across all data conditions. Because this is unreasonable for many datasets, spatial PCA (often followed by component rotation procedures such as Varimax or Promax) is potentially more relevant. Several variants of PCA have been proposed for ERP decomposition to this end.

[0225] Bell and Sejnowski published an information-theoretic iterative algorithm that decomposes linearly mixed signals into time-independent components by minimizing their mutual information. This first method for blind source separation minimizes third- and fourth-order correlations among the observed variables and achieves limited success in simulation. A general approach uses a simple neural network algorithm that employs joint information maximization, or "infomax," as a training criterion. Ten recorded speech and music sources are demixed by transforming the data using compressed nonlinearity and following the entropy gradient of the resulting mixture. A similar method is used to perform blind deconvolution, and the "infomax" method is used to decompose the visual scene.

[0226] The first application of blind decomposition in biomedical time series analysis employed the Infomax Independent Component Analysis (ICA) algorithm to decompose EEG and event-related potential (ERP) data, and its use in monitoring vigilance was reported. This separates artifacts and EEG data into components defined by spatial stability and temporal independence. ICA can also be used to remove artifacts from continuous or event-related (single-experiment) EEG data before averaging. Vigario et al. (1997) used different ICA algorithms to support the use of ICA to identify artifacts in MEG data. Meanwhile, the widespread interest in ICA has led to various applications in biomedical data and other fields (Jung et al., 2000b). Most relevant to EEG / MEG analysis, ICA is effective in separating functionally independent components from functional magnetic resonance imaging (fMRI) data.

[0227] Since the original infomax ICA algorithm was released, several extensions have been proposed. The incorporation of the term "natural gradient" avoids matrix inversion, significantly accelerating convergence and making it usable with large EEG and fMRI datasets on personal computers. The initial "spherization" step further increases the reliability of convergence. The original algorithm assumes that the source has a "sparse" (super-Gaussian) distribution of activation values. This constraint has recently been relaxed in the "extended ICA" algorithm, which allows the identification of both super-Gaussian and sub-Gaussian sources. Many variant ICA algorithms have emerged in the signal processing literature. Typically, these make more specific assumptions about the temporal or spatial structure of the components to be separated and are generally more computationally intensive than the infomax algorithm.

[0228] Because individual electrodes (or magnetic sensors) record a mixture of brain-derived and non-brain-derived signals, it is difficult to interpret and compare spectral metrics across scalp channels. For example, an increase in coherence between two electrode signals could reflect activation of strong brain-derived sources projected to both electrodes, or deactivation of a brain-derived source primarily projected to one electrode. However, if the independent components of EEG (or MEG) data can be viewed as measures of activity within functionally dissimilar brain networks, event-related coherence between independent components can reveal transient event-related changes in their coupling and decoupling (at one or more EEG / MEG frequencies). ERCOH analysis has been applied to independent EEG components in selective attention tasks. Summary of the Invention

[0229] Sleep disorders affect a significant portion of the adult population. In the United States, between 50 million and 70 million adults have a sleep disorder. Insomnia is the most common specific sleep disorder, with approximately 30% of adults reporting short-term problems and 10% having chronic insomnia. Chronic insomnia is associated with memory decline, adverse effects on endocrine function and immune response, and an increased risk of obesity and diabetes. While managing insomnia is challenging at any age, it is a particularly critical condition in older adults due to increased age-related comorbidities and medication use, as well as age-related changes in sleep structure, which shorten sleep duration and impair sleep quality. Therefore, improving sleep quality is one of the most common health complaints among older adults. Medications are widely prescribed to relieve insomnia. However, sleep-promoting agents, such as hypnotics, can have adverse effects, especially in older adults. Even natural supplements such as melatonin can cause side effects, including headaches, depression, daytime sleepiness, dizziness, stomach cramps, and irritability.

[0230] Besides the general decline in sleep quality with age in adults, the decline in the quantity and quality of slow-wave sleep (SWS), which is non-REM deep sleep, is particularly distressing. SWS plays a crucial role in brain repair and recovery in humans. Studies have shown that a 15% reduction in SWS, along with an increase in the number and duration of wakefulness, is associated with normal aging. Experimental disruption of SWS has been shown to increase light sleep, sleep fragmentation, daytime sleepiness, and impair daytime function. Given that SWS contributes to sleep persistence, enhancing SWS could lead to improved sleep quality and daytime function in patients with insomnia and older adults. Furthermore, accumulated evidence points to SWS as the time when short-term memories are consolidated into long-term memories. Recent research has linked the decline of SWS to the early onset of Alzheimer's disease and other forms of insomnia. It has also been shown that the loss of SWS stages can play a role in these debilitating age-related diseases. Unfortunately, most standard sleep pills have little effect on improving SWS while alleviating insomnia. Some evidence suggests that some hypnotic drugs alter sleep structure, thereby adversely affecting SWS. Therefore, the need for non-pharmacological techniques to promote sleep, particularly deep non-REM SS (SWS) which are lacking in older populations, remains unmet.

[0231] One promising non-pharmacological approach to promoting sleep is neuromodulation via light, sound, and / or transcranial electrical stimulation (TES). Limited human trials conducted in collaboration with the Neuromodulation Laboratory at the City University of New York (CUNY) NE Lab have shown promise in replicating the desired sleep apnea (SS) in healthy donors in other subjects (recipients). Electroencephalograms (EEGs) of healthy volunteers were recorded as they napped into stage 1 sleep, as demonstrated by the dominance of alpha waves. These EEG recordings were then filtered from noise, inverted, and used for transcranial endogenous sleep-derived stimulation (tESD). Volunteer subjects stimulated with tESD using locally modulated EEG recordings from the sleep donor rapidly napped and entered stage 1 sleep, as demonstrated by EEG, heart rate, respiratory rate, and post-sleep cognitive tests. These results were superior to control groups in studies that included sham stimulation, tDCS, and tACS (10 Hz). These results suggest that the desired SS of a healthy donor can be replicated in another subject by utilizing tACS, which is based on locally modulated brain waves recorded from a sleeping healthy donor.

[0232] Important studies exist that use markers to identify different phases of healthy or pathological sleep; these markers allow for the classification of observed EEG into one of the sleep / wake cycles. The applicant is unaware of any studies that aim to: comprehensively identify all independent components of the EEG signal during sleep; and comprehensively analyze the statistically significant interdependence of the presence of independent components with a specific sleep phase. Comprehensive identification and analysis of sleep-associated independent components would allow for the use of these components and / or the resulting signals in a tACS protocol.

[0233] EEG recordings of brain waves during various sleep stages were obtained from healthy human subjects and preprocessed. EEG recordings from the three sleep stages and from at least 10 healthy subjects during wakefulness (e.g., via a public EEG database) were then smoothed and filtered. The EEG recordings were analyzed to identify statistically significant waveform components associated with specific sleep apnea (SS). Based on the SS / wakefulness state, a model for the component coefficients of the EEG (e.g., a linear multivariate model) was developed; and the statistical significance of the model was measured. Stimulation protocols were developed that can provide safe and effective neural stimulation to induce desired SS.

[0234] Sleep disorders, especially insomnia, impose a significant economic burden and social cost. Sleep disturbances are common in adults and are associated with a variety of factors, including caffeine, tobacco, and alcohol use; sleep habits; and comorbidities. Epidemiological studies indicate that sleep disorders affect a large segment of the adult population. In the United States, between 50 million and 70 million adults have a sleep disorder. Insomnia is the most common specific sleep disorder, with approximately 30% of adults reporting short-term problems and 10% having chronic insomnia. Chronic insomnia is associated with memory decline, adverse effects on endocrine function and immune response, and an increased risk of obesity and diabetes.3 Furthermore, there are significant economic burdens and social costs associated with insomnia due to its impact on healthcare utilization, work domain, and quality of life. Recent estimates of direct and indirect costs in the United States exceed $100 billion annually. While managing insomnia is challenging at any age, it is particularly critical in older adults due to increased age-related comorbidities and medication use, as well as age-related changes in sleep structure, which shorten sleep duration and impair sleep quality. Therefore, improving sleep quality is one of the most common health complaints among older adults.

[0235] Slow-wave sleep (SWS) decline is observed in older adults. Besides the general decline in sleep quality with age in adults, the decline in both the quantity and quality of SWS, which is deep non-REM sleep, is particularly concerning. SWS plays a crucial role in brain repair and recovery in humans. It is the most prominent EEG event during sleep and manifests as spontaneous large oscillations of EEG signals occurring approximately once per second in the deepest stage of non-REM sleep. Studies have shown that a significant reduction in the amount of SWS (approximately 15% reduction) and an increase in the number and duration of wakefulness are associated with normal aging. Given that SWS contributes to sleep sustaining, and that experimental disruption of SWS increases light sleep and sleep fragmentation, enhances daytime sleepiness, and impairs daytime function, an enhancement of SWS could lead to improved sleep maintenance and daytime function in patients with insomnia and older adults. Furthermore, accumulated evidence points to SWS as a time for consolidating short-term memories into long-term memories. Recent research has linked SWS decline to the early onset of Alzheimer's disease and other forms of insomnia. It also suggests that the loss of the SWS stage may be the culprit behind these debilitating age-related diseases.

[0236] SWS enhancement is a potential nonpathological therapy for older adults. Given the crucial role of slow waves during sleep, it is not surprising that efforts are being made to improve sleep efficacy by enhancing the SWS. Recently, many drugs have been shown to increase the SWS. Although acting at different synaptic sites, the slow waves that generally enhance the effects of these drugs are mediated by enhanced GABAergic transmission. Specifically, clinical studies have shown that both tiagabine and gaboxadol increase the duration of the SWS following sleep conditioning. Tiagabine also improved performance on cognitive tasks assessing executive function and reduced the negative effects of sleep conditioning on alertness. Despite these positive results, pharmacological approaches to sleep enhancement often raise issues related to dependence and tolerability and are frequently associated with residual daytime side effects. Some evidence suggests that some hypnotic drugs, while relieving insomnia, alter sleep structure, thus adversely affecting the SWS. Even natural supplements such as melatonin can cause side effects including headache, short-term feeling of depression, daytime sleepiness, dizziness, stomach cramps, and irritability. Therefore, the need for non-pharmacological techniques to promote deep non-REM sleep, particularly in older populations, remains unmet.

[0237] The brain activity of the first subject (the "donor" in the desired sleep state) can be captured by recording neural correlations of sleep, such as brain activity patterns represented by EEG signals. The representation of the neural correlations of the first subject is used to control stimulation of the second subject (the "recipient") in an attempt to induce the same brain activity patterns in the recipient as those of the donor, in order to help the recipient achieve the desired sleep state already achieved by the donor.

[0238] One strategy for nonpharmacologically enhancing deep sleep is based on artificial and synthetic stimulation paradigms, using light, sound, electrical current, or magnetic fields to stimulate the brain. Intermittent transcranial direct current stimulation (tDCS) applied at 0.75 Hz for 5-minute intervals (separated by 1-minute breakpoints) after SWS initiation can increase EEG power in the slow oscillation band (<1 Hz) during the unstimulated interval. Similarly, tDCS stimulation at the start of SWS accelerates the decay of SWA homeostasis in subjects. Furthermore, slow waves can be triggered by directly perturbing the cortex using transcranial magnetic stimulation (TMS) during non-REM sleep. Other studies have focused on the possibility of inducing slow waves in a more physiologically natural way. In larger studies of healthy adults, bilateral electrical stimulation with vestibular devices shortened sleep initiation latency compared to pseudo-nights without stimulation. The effects of somatosensory and auditory stimulation have also been evaluated. While changes observed with somatosensory stimulation were small, acoustic stimulation was particularly effective in enhancing sleep slow waves. Specifically, using non-intermittent stimulation in which the tone is played in 15-second blocks separated by no-stimulation intervals, slow waves appear surprisingly large and numerous within the stimulated blocks. Furthermore, high-density EEG studies (hdEEG, 256 channels) have shown that the morphology, topography, and progression patterns of the induced slow waves are distinguishable from those of spontaneous slow waves observed during natural sleep. Recent studies have found an increase in EEG SWA during non-REM sleep following tone presentation, and an increase in slow oscillatory activity (0.5–1 Hz) in response to continuous acoustic stimulation at 0.8 Hz, starting 2 minutes before light cutoff and lasting for 90 minutes. Unlike previous neurostimulation approaches with artificial and synthetic stimulation paradigms, this stimulation protocol uses waveforms from source sources extracted from local EEG recordings of brain activity in healthy subjects, processed using statistical methods (e.g., principal component analysis or spatial principal component analysis, autocorrelation, etc.) that separate the individual components of brain activity. These separate brain EEG activities are then modified or modulated, and subsequently inverted and used for transcranial endogenous sleep-derived stimulation (tESD). The application of endogenous brain waveforms should not only maintain the efficacy of triggering SWS, but also mitigate the safety concerns associated with long-term brain stimulation using synthetic paradigms.

[0239] This technology provides a method for improving sleep by transferring a sleep state—a desired sleep state (SS) or a series of SSs—from a first subject (donor) (or multiple donors) to a second subject (recipient). (In some embodiments, the first and second subjects may be the same subject at different time points or based on a protocol or algorithm.)

[0240] The method attempts to achieve brainwave patterns obtained from humans within a subject. Brainwave patterns are complex, representing superpositions of modulated waveforms. The modulation is preferably determined based on the brainwave patterns of another subject or multiple subjects.

[0241] Sleep is a natural, periodic pause in consciousness, essentially a process in its individual phases that is almost unaffected by human sleep. It is a subconscious (in a technical sense) mental state, representing a resting state, activity pattern, activity rhythm, readiness, receptivity, or other state, generally independent of specific input. Essentially, the sleep state of a specific SS or a series of different SSs of the first subject (either in the desired SS or experiencing its series of individual phases, the "donor") is captured by recording neural relevances of the sleep state, for example, through brain activity patterns represented by EEG or MEG signals. The neural relevances of the first subject, as a direct or recorded representation, can then be used to control stimulation of the second subject (the "recipient"), attempting to induce in the second subject the same brain activity pattern present in the first subject (the donor), thereby transplanting the sleep state of the first subject (the donor) to assist the second subject (the recipient) in obtaining the desired SS obtained by the donor. In alternative embodiments, signals from the first subject (donor) in a first sleep stage (SS) are used to prevent the second subject (recipient) from reaching a second sleep stage, which is an undesirable sleep phase. Furthermore, the duration and timing of the different SSs for the second subject can be controlled. This allows for alteration of the individual duration or intensity of each SS and the order in which they occur. In some embodiments, the signals from the first subject can be used to trigger sleep in the second subject or prevent sleep or drowsiness and associated symptoms such as fatigue, lack of attention, etc.

[0242] In some embodiments, before or after obtaining sleep state information, the identification SS, the first subject (donor) or the observer makes a direct report or performs automated analysis of physiological parameters (e.g., brain activity patterns, heart rate, breathing patterns, blood oxygen saturation, temperature, eye movement, skin resistance, etc.) or both.

[0243] In other embodiments, processing the brain activity patterns does not attempt to classify or characterize them, but rather filters and transforms the information into a form suitable for controlling stimulation of the second subject. Specifically, according to this embodiment, subtleties that have not been reliably classified in conventional brain activity pattern analysis are taken into account. For example, it should be understood that all brain activity is reflected in synaptic currents and other neural modulations, and therefore, theoretically, conscious and subconscious information can be obtained through brain activity pattern analysis. Since available processing techniques often fail to distinguish a large number of different brain activity patterns, the available processing techniques are necessarily inadequate but can be improved. However, the fact that computational algorithms cannot be used to extract information does not mean that the information is not present. Therefore, this embodiment uses relatively raw brain activity pattern data, such as filtered or unfiltered EEG, to control stimulation of the second subject without fully understanding or knowing exactly what important information is present. In one embodiment, brain waves are recorded and “played back” to another subject, similar to recording and playing back music. This recording and playback can be digital or analog. Typically, stimulation can include low-dimensional stimuli such as stereo-light, binaural, isotonic, tactile, or other sensory stimuli; bidirectional manipulation; and control over frequency and phase and / or waveform and / or transcranial stimulation, such as TES, tDCS, HD-tDCS, tACS, or TMS. Multiple different types of stimulation can be applied in parallel, such as visual, auditory, other sensory, magnetic, and electrical stimuli.

[0244] Similarly, the current lack of understanding of the fundamental characteristics of signal components in brain activity patterns does not prevent their acquisition, storage, communication, and processing (to some extent). Stimuli can be direct, i.e., visual, auditory, or tactile stimuli corresponding to brain activity patterns, or derivative control or feedback control based on the second subject's brain activity patterns.

[0245] In order to address, in whole or in part, the foregoing problems and / or other problems that may have been observed by those skilled in the art, this disclosure provides methods, processes, systems, devices, instruments and / or apparatuses as exemplified by the embodiments described below.

[0246] While mental states are typically considered internal and subjective, they are in reality common among individuals and possess determinable physiological and electrophysiological group characteristics. Furthermore, mental states can be externally altered or induced in ways that bypass normal cognitive processes. In some cases, the triggering of mental states is subjective, and therefore the specific subject-specific sensory or motivational protocols required to induce a particular state will vary. For example, olfactory stimuli may have different effects on different individuals based on differences in exposure history, social and cultural norms, etc. On the other hand, some mental state response triggers are normative, such as "tear-jerking" media.

[0247] Mental states are represented by brainwave patterns, and in normal individuals, brainwave patterns and metabolism (e.g., blood flow, oxygen consumption, etc.) follow prototype patterns. Therefore, by monitoring an individual's brainwave patterns, the person's state or a range of mental states can be determined or estimated. However, brainwave patterns may be correlated with context, other activities, and past history. Furthermore, while prototype patterns can be observed, individual variations also exist within the patterns. Brainwave patterns can contain characteristic spaces and event patterns indicative of mental states. These patterns can be extracted from a person's brainwave signals, which can be represented, for example, as hemispherical signals in the frequency range of 3-100 Hz. These signals can then be synthesized or modulated into one or more stimulus signals, which are then used to induce the corresponding mental state in the recipient in an attempt to achieve a brainwave pattern similar to that of the source. It is not necessary to acquire a new brainwave pattern to be introduced for each situation. Instead, signals can be acquired from one or more individuals to obtain models for various corresponding mental states. Once determined, the processed signal representation can be stored in non-volatile memory for later use. However, in cases of complex interactions between mental states and context or content or activity, it may be appropriate to derive signals from a single individual within the context or content environment or activity suited to the situation. Furthermore, in some cases, a single mental state, emotion, or mood is not described or fully characterized, and therefore deriving signals from a source is an effective exercise.

[0248] A system and method are provided using a target brainwave pattern library, in which a target subject can be immersed in a presentation that includes not only multimedia content but also a series of defined mental states, emotional states, or emotions accompanying the multimedia content. In this way, the multimedia presentation becomes fully immersive. Stimulation can be provided via a head-mounted device such as a virtual reality or augmented reality headset. This head-mounted device is equipped with a stereoscopic display, binaural audio, and EEG and transcranial stimulation electrode sets. These electrodes (if present) typically deliver painless subthreshold signals, typically AC signals, corresponding to the desired frequency, phase, and spatial location of the desired target pattern. The electrodes can also be used to counteract unwanted signals by destructively interfering with the desired pattern while applying it in parallel. The head-mounted device can also generate visual and / or auditory signals corresponding to the desired state. For example, auditory signals can induce binaural beats that result in brainwave entrainment. Visual signals can contain intensity fluctuations or other modulation patterns, especially lower-level ones, which are also tuned to induce brainwave entrainment or the induction of the desired brainwave pattern.

[0249] The head-mounted device preferably includes EEG electrodes for receiving feedback from the user. That is, the stimulation system attempts to achieve a mental state, emotion, or emotional response from the user. The EEG electrodes allow determination of whether the state has been achieved, and if not, what the current state is. It is possible that the desired brainwave pattern is state-dependent, and therefore the characteristics of the stimulation used to achieve the desired state depend on the subject's initial state. Other methods for determining mental state, emotion, or mood include facial expression analysis, electromyography (EMG) analysis of facial muscles, explicit user feedback, etc.

[0250] A creation system is provided that allows content designers to determine desired mental states and then encode these states into media, which are then interpreted by a media reproduction system to generate appropriate stimuli. As indicated above, the stimuli can be audio, visual, multimedia, other sensory, or electro- or magneto-brain stimulation, and therefore a VR headset with transcranial electrical or magnetic stimulation is not required. Further, in some embodiments, patterns can be directly encoded into audiovisual content for further encoding.

[0251] In some cases, the target mental state can be obtained from an expert, actor, or professional paradigm. The state can be read from the actor or paradigm based on facial expressions, EMG, EEG, or other means. For example, the prototype paradigm participates in activities that trigger a response, such as viewing artworks in the Grand Canyon or the Louvre. The paradigm's response is then recorded or represented, and preferably, the brainwave patterns representing the response are recorded. A representation of the same experience is then presented to the target, whose aim is to also experience the same experience as the paradigm. This is typically a voluntary and public process, so the target attempts to voluntarily conform to the desired experience. In some cases, the use of the technology is not public to the target, such as in advertising presentations or billboards. For an actor to act as a paradigm, the emotion achieved by that person must be genuine. However, a so-called "method actor" does genuinely achieve the emotion they convey. However, in some cases, such as when facial expressions are used as indicators of mental state, the actor may present the desired facial expressions with an inauthentic mental state. Making an action that corresponds to an emotion often achieves a targeted mental state.

[0252] To calibrate the system, brain patterns can be measured while the person is in the desired state. The acquired brain patterns used for calibration or feedback do not need to have the same quality, accuracy, or data depth, and can actually represent the response rather than the primary label. That is, in the system, there may be some asymmetry between the brainwave patterns representing mental states and the stimulus patterns suitable for inducing those states.

[0253] This invention generally relates to inducing a mental state in a subject by transmitting brainwave patterns to the subject's brain. These brainwaves can be artificial or synthetic, or can be obtained from the brain of a second subject (e.g., someone experiencing a real-life experience or participating in an activity). Typically, the brainwave patterns of the second subject are derived while the second subject is experiencing the real-life experience.

[0254] A special case is when the first and second subjects are the same individual. For example, brainwave patterns are recorded when the subject is in a specific mental state. The same pattern can facilitate achieving the same mental state at another time. Therefore, there may be a time delay between obtaining brainwave information from the second subject and subjecting the first subject to the corresponding stimulus. The signals can be recorded and transmitted.

[0255] Temporal patterns can be non-invasively transmitted or induced via light (visible or infrared), sound (or ultrasound), transcranial direct current or alternating current stimulation (tDCS or tACS), transcranial magnetic stimulation (TMS), deep transcranial magnetic stimulation (deep TMS or dTMS), repetitive transcranial magnetic stimulation (rTMS), olfactory stimulation, tactile stimulation, or any other means capable of transmitting frequency patterns. In a preferred embodiment, normal human senses, such as light, sound, smell, and touch, are used to stimulate the subject. Combinations of stimuli may be employed. In some cases, the stimuli or combinations are innate and therefore largely subject-specific. In others, the response to context is learned and therefore subject-specific. Therefore, feedback from the subject may be suitable for determining appropriate triggers and stimuli to achieve a mental state.

[0256] This technology can be advantageously used to enhance mental responses to stimuli or context. It also provides for alterations in mental state. The technology can be used on humans or animals.

[0257] This technique can employ event-related EEG temporal and / or frequency analysis performed on neuronal activity patterns. In temporal analysis, the signal is analyzed in both time and space, typically seeking variations relative to time and space. In frequency analysis, within an analysis epoch, data typically ordered in time for the sample is transformed into a frequency domain representation using, for example, Fourier transform (FT or an implementation such as Fast Fourier Transform, FFT), and the frequencies present during the epoch are analyzed. The analysis window can be scrolling, and therefore the frequency analysis can be continuous. In hybrid time-frequency analysis, such as wavelet analysis, wavelet transforms, such as Discrete Wavelet Transform (DWT) or Continuous Wavelet Transform (CWT), are used to transform the data during the epoch. These wavelet transforms are capable of constructing a time-frequency representation of the signal, providing good temporal and frequency localization. Variations of the transformed data over time and space can be analyzed. Typically, the spatial aspect of EEG analysis is anatomically modeled. In most cases, anatomy is considered universal, but in some cases, significant differences exist. For example, brain injury, mental illness, age, race, native language, training, sex, hand bias, and other factors can lead to differences in the spatial arrangement of brain function. Therefore, when transferring emotions from one individual to another, it is preferable to normalize the brain anatomy of the two individuals by having them experience roughly the same events and measuring the spatial parameters of EEG or MEG. Note that the spatial organization of the brain is highly persistent, without injury or disease, and therefore this only needs to be performed infrequently. However, since electrode placement may be inaccurate, spatial calibration can be performed after electrode placement.

[0258] Different aspects of EEG amplitude and phase relationships can be captured to reveal details of neuronal activity. "Time-frequency analysis" reveals the brain's parallel processing of information, where oscillations of various frequencies across different brain regions reflect multiple neural processes occurring and interacting simultaneously. See Lisman J and Buzsaki G, "A neural coding scheme formed by the combined function of gamma and theta oscillations," Schizophrenia Bulletin, June 16, 2008; doi:10.1093 / schbul / sbn060. This time-frequency analysis can take the form of wavelet transform analysis. This can be used to aid in the integration and dynamic adaptive processing of information. Of course, the transform can be substantially lossless and can be performed in any convenient information domain representation. These EEG-based data analyses reveal frequency-specific neuronal oscillations and their synchronization across brain functions ranging from sensory processing to higher-order cognition. Therefore, these patterns can be selectively analyzed to transfer to or induce in subjects.

[0259] Statistical clustering analysis can be performed in a high-dimensional space to isolate or segment regions that act as signal sources and characterize the coupling between these regions. This analysis can also be used to establish the signal types within each brain region, as well as decision boundaries characterizing transitions between different signal types. These transitions can be state-dependent, and therefore can be detected based on time analysis rather than solely on parallel oscillator states.

[0260] During spectral decomposition and / or temporal / spatial / spectral analysis, various measures utilize amplitude and / or phase angle information extracted from complex data from the EEG. Some measures estimate the amplitude or phase coherence of the EEG across a single channel during an experiment, while others estimate the coherence of amplitude or phase differences across channels during an experiment. In addition to these two families of calculations, measures exist to examine coupling between frequencies within the experimental and recording locations. Of course, in the field of time-frequency analysis, many types of relationships beyond those already mentioned can be examined.

[0261] These sensory processes of specific neuronal oscillations, such as the brainwave patterns of the subject (“source”) or those of a person trained to produce the desired state (e.g., an actor trained in the “method”), can be stored on a tangible medium and / or simultaneously transmitted to the recipient using brain frequencies, depending on the nature of the response. See Galbraith, Gary C., Darlene M. Olfman, and Todd M. Huffman, “Selective attention affects human brain stem frequency-following response,” Neuroreport 14, Vol. 5, (2003): 735-738, journals.lww.com / neuroreport / Abstract / 2003 / 04150 / Selective_attention_affects_human_brain_stem.15.aspx.

[0262] According to one embodiment, the stimulation and feedback process for the second subject is combined to verify that the second subject responds appropriately to the stimulation, for example, as if the first subject has a predetermined similarity to the SS, its SS has a predetermined difference from the first subject, or has a desired change compared to the baseline SS, not based on brain activity itself or the neural relevance of the SS, but based on physical, psychological, or behavioral effects that can be measured, reported, or observed.

[0263] Feedback is typically provided to a controller of the stimulator that has at least a partial model basis, which modifies the stimulation parameters to optimize the stimulation.

[0264] As discussed above, models are typically difficult to define. Therefore, model-based controllers are incompletely defined and are expected to contain errors and artifacts. However, by employing a model-based controller, the responses to corresponding controllers lacking models can be improved using those defined parameters.

[0265] For example, brain waves are believed to represent a form of resonance in which groups of neurons interact in a coordinated manner. The frequency of the waves is related to the responsiveness of a nerve to neurotransmitters, the distance along the neural pathway, diffusion limitations, and so on. That is, the same brainwave (SS) can be represented by slightly different frequencies in two different individuals, depending on differences in brain size, the presence of neural modulators, and other anatomical, morphological, and physiological differences. These differences can be measured in microseconds or less, resulting in minute variations in frequency. Therefore, the model composition of the controller can determine the parameters and overall characteristics of neural transmission (relative to the stimulus) and resynthesize the stimulus signal to match the correct frequency and phase of the subject's brainwaves, where optimization of the waveform is adaptively determined. This may not be as simple as speeding up or slowing down signal playback, because different elements of various brainwaves representing the neural relevance of the SS may have different relative differences between subjects.

[0266] Of course, in some cases, one or more components of the stimulus to the target subject (recipient) can be represented as an abstract or semantically defined signal, and more specifically, the processing of the signal used to define the stimulus will involve high-level modulation or transformation between source signals received from a first subject (donor) or multiple donors to define the target signal used to stimulate a second subject (recipient).

[0267] Preferably, each component represents a subset of neural correlations reflecting brain activity, which are highly autocorrelated in space and time or in a hybrid representation such as wavelets. Once the characteristics of the signal are known, these can be separated by optimal filtering (e.g., optical PCA), and it should be remembered that the signal is accompanied by a modulation pattern, and the two components themselves may have some weak coupling and interaction.

[0268] For example, if the first subject (donor) is listening to music, there will be an important component of neural relevance synchronized with the specific music. On the other hand, the music itself may not be part of the desired stimulus for the target subject (recipient). Furthermore, the target subject (recipient) may be in a different acoustic environment, and the residual signal may be modified depending on the recipient's acoustic environment so that the stimulus is suitable for achieving the desired effect and does not represent hallucinations, distractions, or irrelevant or inappropriate content. For performing signal processing, it is convenient to store the signal or a partially processed representation, but a complete real-time signal processing chain can be implemented. According to another embodiment, a specific stage of the sleep state of at least one first subject (donor) is identified, and the neural relevance of brain activity is captured, and a second subject (recipient) is stimulated based on the captured neural relevance and the identified SS. SS is typically represented as a semantic variable within a finite classification space. SS identification does not require analysis of the neural relevance signal and can be voluntary self-identification by the first subject, for example, based on other bodily signals or by an observer, or manual classification by a third party using, for example, observation, fMRI, or psychological assessment. The identified SS is useful, for example, because it indicates that it can be directed (or, in some cases, targeted) at the goal of manipulating a second subject (the recipient).

[0269] The stimulus can be one or more stimuli applied to a second subject (trainee or recipient), which can be electrical or magnetic transcranial stimulation (tDCS, HD-tDCS, tACS, oscillatory tDCS or TMS), sensory stimulation (e.g., visual, auditory or tactile), mechanical stimulation, ultrasound stimulation, etc., and can be controlled relative to waveform, frequency, phase, intensity / amplitude, duration, or through feedback, self-reporting by the second subject, manual classification by a third party, or automatic analysis of the second subject's brain activity, behavior, physiological parameters, etc.

[0270] Typically, the purpose of the process is to improve the recipient's sleep as follows: by transplanting at least one desired SS or series of phases of the first subject (donor) into the second subject (recipient), by inducing the neural correlation of at least one SS (or series of phases) of the first subject (donor) corresponding to the SS of the first subject in the second subject (recipient), by using stimulation parameters including waveforms obtained from the neural correlation of the SS of the first subject over a certain period of time.

[0271] Typically, the first subject and the second subject are spatially distant from each other and may also be temporally distant. In some cases, the first subject and the second subject are the same subject (human or animal) that are temporally displaced. In other cases, the first subject and the second subject are spatially close to each other. These different embodiments differ primarily in how signals are transferred from at least one first subject (donor) to the second subject (recipient). However, when the first subject and the second subject share a common environment, signal processing of neural relevance, and especially of real-time feedback on neural relevance from the second subject, can involve interactive algorithms for neural relevance with the first subject.

[0272] According to another embodiment, the first subject and the second subject are each stimulated. In a particularly interesting embodiment, the first subject and the second subject communicate with each other in real time, wherein the first subject receives stimulation based on the second subject, and the second subject receives feedback based on the first subject. This can lead to synchronization of neural correlations (e.g., neuronal oscillations or brainwaves), and thus synchronization of SS between the two subjects. Neural correlations can be neuronal oscillations that result in brainwaves that can be detected, for example, in the form of EEG, qEEG, or MEG signals. Conventionally, these signals are found to have a dominant frequency, which can be determined by various analyses, such as spectral analysis, wavelet analysis, or principal component analysis (PCA). One embodiment provides that the modulation pattern of the brainwaves of at least one first subject (the donor) is determined independently of the dominant frequency of the brainwaves (but typically within the same class of brainwaves), and this modulation is applied to the brainwaves corresponding to the dominant frequency of the second subject (the receiver). That is, once the second subject achieves the same brainwave pattern as the first subject (which can be achieved by means other than electromagnetic, mechanical, or sensory stimulation), the modulation pattern of the first subject is applied in a manner that guides the SS of the second subject.

[0273] According to another embodiment, the second subject (recipient) is stimulated with a stimulus signal that faithfully represents the frequency composition of a defined component of the neural relevance of at least one first subject (donor). The defined components can be determined based on principal component analysis, independent component analysis (ICI), eigenvector-based multivariate analysis, factor analysis, canonical correlation analysis (CCA), nonlinear dimensionality reduction (NLDR), or related techniques.

[0274] Stimulation can be performed, for example, using a TES device, such as a tDCS device, a high-definition tDCS device, an oscillating tDCS device, a pulsed tDCS (“electric sleep”) device, an oscillating tDCS, a tACS device, a CES device, a TMS device, an rTMS device, a depth TMS device, a light source, or a sound source configured to modulate a dominant frequency on a light signal or a sound signal, respectively. Stimulation can be a light signal, a sound signal (sound), an electrical signal, a magnetic field, an olfactory stimulus, or a tactile stimulus. The electrical signal can be a pulsed signal or an oscillating signal. Stimulation can be applied via transcranial electrical stimulation (CES), transcranial electrical stimulation (TES), depth electrical stimulation, transcranial magnetic stimulation (TMS), depth magnetic stimulation, light stimulation, sound stimulation, tactile stimulation, or olfactory stimulation. Auditory stimulation can be, for example, binaural beats or isochronous tones.

[0275] This technology also provides a processor configured to process the neural correlations of SS from a first subject (donor) and selectively generate or define stimulation patterns for a second subject (recipient) based on the waveform patterns of the neural correlations from the first subject. The processor can also perform PCA, spatial PCA, independent component analysis (ICA), eigenvalue decomposition, eigenvector-based multivariate analysis, factor analysis, autoencoder neural networks with linear hidden layers, linear discriminant analysis, network component analysis, nonlinear dimensionality reduction (NLDR), or other statistical methods of data analysis.

[0276] A signal is presented to a second device configured to stimulate the second subject (recipient). The signal can be an open-loop stimulus dependent on a non-feedback control algorithm or an algorithm dependent on closed-loop feedback. The second device generates a stimulus designed to induce a desired SS in the second subject (recipient), for example, representing an SS identical to the SS present in the first subject (donor).

[0277] A typical procedure performed on neural relevance is filtering to remove noise. In some embodiments, the noise filter may be provided, for example, at 50 Hz, 60 Hz, 100 Hz, 120 Hz, and other overtones (e.g., third and higher harmonics). The stimulator associated with the second subject (recipient) will typically perform decoding, decompression, decryption, inverse transformation, modulation, etc.

[0278] Alternatively, the real wave or its hash can be authenticated via a blockchain, and thus by an immutable record. In some cases, it may be possible to use an encrypted signal stored in its encrypted form that does not require decryption.

[0279] Due to differences in brain size and other anatomical, morphological, and / or physiological variations, the dominant frequency associated with the same sleep apnea (SS) may differ among subjects. Therefore, forcefully applying the donor's frequency to the recipient may not be optimal, and the frequency may or may not precisely correspond to the recipient's frequency associated with the same SS. Thus, in some embodiments, the donor's frequency may be used to initiate the process of inducing the desired SS within the recipient. Sometimes, stimulation is stopped or replaced by neural feedback when the recipient closes to achieve the desired sleep state, allowing the recipient's brain to find its own optimal frequency associated with the desired SS.

[0280] In one embodiment, the feedback signal from the second subject can be encoded accordingly based on the source signal, and the error between the two can be minimized. According to one embodiment, the processor can perform noise reduction differently than band filtering. According to one embodiment, the neural correlation is transformed into a sparse matrix, and in the transform domain, components with high probabilities representing noise are masked, while components with high probabilities representing the signal are preserved. That is, in some cases, the components representing important modulation may not be known a priori. However, depending on their role in inducing the desired response in the second subject (recipient), "important" components can be identified, and the remainder can be filtered or suppressed. The transformed signal can then be inversely transformed and used as the basis for the stimulation signal.

[0281] According to another embodiment, a method for modifying brainwave activity (SS), such as brain entrainment, is provided. The method includes: determining the SS of a plurality of first subjects (donors); acquiring brainwaves of the plurality of first subjects (donors), for example using one of EEG and MEG, to generate a dataset containing brainwaves corresponding to different SS. The database may be encoded using classifications of SS, activity, environmental, or stimulus patterns applied to the plurality of first subjects, and the database may contain acquired brainwaves across, for example, a large number of SS, activity, environmental, or stimulus patterns. In many cases, the database records will reflect characteristic or dominant frequencies of the corresponding brainwaves.

[0282] The database can be accessed based on, for example, SS, activity, environment, or stimulus pattern, and stimulus patterns defined by one or more subjects (donors) based on the database records of a second subject (recipient).

[0283] One or more records retrieved are used to define the stimulation pattern for the second subject (recipient). As a relatively trivial example, female recipients may be stimulated primarily based on records from female donors. Similarly, child recipients of a certain age may be stimulated primarily based on records from child donors of similar age. Likewise, various demographic, personality, and / or physiological parameters can be matched to ensure a high degree of correspondence between the source and target subjects. Within the target subject, guided or genetic algorithms can be employed to select modified parameters from various components of the signal, which, based on feedback from the target subject, optimally achieves the desired target state.

[0284] A more subtle approach would be to process the entire database and stimulate the second subject based on a global brainwave stimulation model, but this is unnecessary, and the underlying basis of the model could prove unreliable or inaccurate. In fact, it might be preferable to obtain the stimulation waveform from only a single first subject (the donor) to preserve the fine-modulation aspects of the signal, which, as mentioned above, have not yet been adequately characterized. However, the selection of one or more donors does not need to be static and can be frequently changed. The selection of donor recordings could be based on group statistics of other users of the recordings, i.e., whether the recordings have or do not have the expected effect, thereby filtering the donor whose response pattern is most relevant to a given recipient, etc. The selection of donor recordings could also be based on feedback patterns from the recipient.

[0285] The stimulation process typically attempts to target the recipient's desired sleep stage (SS), which is determined automatically or semi-automatically or manually. In one embodiment, recording is used to define the modulated waveforms of a synthetic carrier or a set of carriers, and the process may involve frequency-domain multiplexed multi-carrier signals (not necessarily orthogonal). Multiple stimuli can be applied in parallel through different sub-channels and / or through different stimulator electrodes, current stimulators, magnetic field generators, mechanical stimulators, sensory stimulators, etc. The stimulation can be applied to achieve brain entrainment between a second subject (recipient) and one or more first subjects (donors). If the multiple donors are entrained by each other, each will have a corresponding brainwave pattern, depending on the basis of the brainwave entrainment. This connection between donors may help determine the compatibility between the respective donor and recipient. For example, characteristic patterns of the entrained brainwaves can be determined, even for different target SSs, and these characteristic patterns can be correlated to find relatively close matches and exclude relatively poor matches.

[0286] This technology can also provide a foundation for social networks, dating sites, employment, missions (e.g., space or military), or career tests or other interpersonal settings where people can match each other based on hook-up characteristics. For example, people who hook up effectively with each other can have better compatibility and therefore better marriages, jobs, or social relationships compared to those who don't. Hook-up effects are not limited to SS and can occur across any context.

[0287] As discussed above, the plurality of first subjects (donors) can store their respective brainwave patterns in separate database records. A neural network is trained using data from the plurality of first subjects (donors), and then accessed by inputting target phases and / or feedback information. The neural network outputs stimulation patterns or parameters for controlling one or more stimulators. When the plurality of first subjects (donors) forms the basis of the stimulation patterns, preferably, the stimulators are then controlled using neural network output parameters derived from and including features of the brainwave patterns or other neurally relevant characteristics of the SS from the plurality of first subjects (donors). These stimulators, for example, generate one or more of their own carrier waves, which are then modulated based on the output of the neural network. The trained neural network does not require periodic retrieval of records and can therefore operate in a more sustained manner, rather than with more granular schemes of control based on records.

[0288] In any feedback correlation method, brainwave patterns or other neural correlations of sleep spectroscopy (SS) can be processed by a neural network to generate the output of guided or controlled stimulation. The stimulation is at least one of, for example, light signals, sound signals, electrical signals, magnetic fields, olfactory signals, chemical signals, and vibrational or mechanical stimuli. The process can employ a relational database of SSs and brainwave patterns, such as frequency / neuralally correlated waveform patterns associated with corresponding SSs. The relational database may include: a first table, further comprising multiple data records of brainwave patterns; and a second table, comprising multiple SSs, each of which is associated with at least one brainwave pattern. Data related to the SSs and the brainwave patterns associated with them is stored in and maintained in the relational database. The relational database is accessed by receiving a query for a selected (existing or desired) SS, and data records representing the associated brainwave patterns are returned. The brainwave patterns retrieved from the relational database can then be used to modulate a stimulator in an attempt to selectively exert an effect depending on the desired sleep stage.

[0289] Another aspect of this technology provides a computer device for creating and maintaining a relational database of brain activity (SS) and frequencies associated with the SS. The computer device may include non-volatile memory for storing SS and a relational database of neural correlations of brain activity associated with the SS, the database comprising: a first table including a plurality of data records of neural correlations of brain activity associated with the SS; and a second table including a plurality of SS, each SS associated with one or more records in the first table; a processor coupled to the non-volatile memory and configured to process relational database queries and then use the relational database queries to look up the database; RAM coupled to the processor and the non-volatile memory for temporarily holding database queries and data records retrieved from the relational database; and an I / O interface configured to receive database queries and deliver data records retrieved from the relational database. Alternatives to Structured Query Language (SQL) or SQL (e.g., NoSQL) databases may be used to store and retrieve records. The relational database described above, maintained and operated by a general-purpose computer, improves the operation of the general-purpose computer by making searches for specific SS and their associated brainwaves more efficient, thereby reducing the power requirements of the computer.

[0290] Another aspect of this technology provides a method of brain entrainment, the method comprising: determining the spinal surface area (SS) of at least one first subject (donor); recording brain waves of the at least one first subject (donor) using at least one channel of an EEG and / or MEG; storing the recorded brain waves in a physical storage device; retrieving the brain waves from the storage device; and applying a stimulation signal to a second subject (recipient) via transcranial electrical and / or magnetic stimulation, comprising a brain wave pattern obtained from at least one channel of the EEG and / or the MEG, thereby achieving the desired SS of the second subject (recipient). The stimulation may have the same dimension (number of channels) as the EEG or the MEG or a different number of channels, typically reduced. For example, the EEG or the MEG may include 64, 128, or 256 channels, while the transcranial stimulator may have 32 or fewer channels. The placement of electrodes for transcranial stimulation may be approximately the same as the placement of electrodes used for recording the EEG or MEG to preserve the recorded signals and, possibly, for spatial modulation of these signals.

[0291] One advantage of transforming data is the ability to select a transform that separates the information of interest represented in the original data from noise or other information. Some transforms preserve the spatial and state transition history and can be used for more comprehensive analysis. Another advantage of transforms is that they can present the information of interest in the form of relatively simple low-order linear or statistical functions. In some cases, it is desirable to perform an inverse transform on the data. For example, if the original data contains noise, such as 50 or 60 Hz interference, a frequency transform can be performed, followed by narrowband filtering of the interference and its higher-order intermodulation products. An inverse transform can be performed to return the data to its time-domain representation for further processing. (In the case of simple filtering, a finite impulse response (FIR) or infinite impulse response (IIR) filter can be used). In other cases, the analysis continues in the transformed domain.

[0292] Transformation can be part of an efficient algorithm that compresses data for storage or analysis by making the representation of the information of interest consume fewer bits of information (if in digital form) and / or allowing it to be communicated using lower bandwidth. Typically, the compression algorithm will not be lossless and therefore the compression is irreversible relative to the truncated information.

[0293] Typically, one or more transformations and filtering of a signal are performed using conventional computer logic according to a defined algorithm. Intermediate stages can be stored and analyzed. However, in some cases, neural networks or deep neural networks, convolutional neural network architectures, or even analog signal processing can be used. According to one set of embodiments, the transformations (if any) and analysis are implemented in a parallel processing environment, such as using a SIMD processor, like a GPU (or GPGPU). Algorithms implemented in such systems are characterized by avoiding data-dependent branch instructions, where many threads execute the same instructions in parallel.

[0294] EEG signals are analyzed to determine the location (e.g., voxels or brain regions) of the emitted electrical activity patterns and to characterize the wave patterns. Spatial processing of EEG signals typically precedes content analysis because noise and artifacts can be useful for spatial resolution. Furthermore, signals from one brain region will typically be noise or interference in the analysis of signals from another brain region; therefore, spatial analysis can be part of the interpretive analysis. Spatial analysis typically takes the form of a geometrically and / or anatomically constrained statistical model, employing all raw inputs in parallel. For example, in the case where the input data is percutaneous EEG information, 32 surrounding input channels sampled at 500 sps, 1 ksps, or 2 ksps from 32 EEG electrodes are processed in a four-dimensional or higher-dimensional matrix to allow for mapping of location and transmission of impulses over time, space, and state.

[0295] Matrix processing can be performed on a Windows 10 operating system running the Matlab (Mathworks, Woburn, MA) software platform in a standard computing environment, such as an i7-7920HQ, i7-8700K, or i9-7980XE processor. Alternatively, matrix processing can be performed in a computer cluster, grid, or cloud computing environment. The processing can also be performed in parallel in distributed and loosely coupled or asynchronous environments. A preferred embodiment employs a single-instruction multiple-data processor, such as a graphics processing unit, an NVIDIA CUDA environment, or an AMD Firepro high-performance computing environment.

[0296] Artificial intelligence (AI) and machine learning methods, such as artificial neural networks and deep neural networks, can be implemented to extract signals of interest. Neural networks act as optimized statistical classifiers and can have arbitrary complexity. So-called deep neural networks with multiple hidden layers can be employed. The processing typically depends on labeled training data, such as EEG data or various processed, transformed, or classified representations of EEG data. Labels indicate the subject's sentiment, emotion, context, or state during acquisition. To process continuous data streams represented by EEG, a recurrent neural network architecture can be implemented. The formal implementation of recursion can be avoided depending on the preprocessing prior to the neural network. A four-dimensional or higher data matrix can be derived from traditional spatiotemporal processing of EEG and fed into the neural network. Since the time parameters are represented in the input data, no temporary storage for the neural network is required; however, this architecture may require a large amount of input. Alternatively, principal component analysis (PCA, en.wikipedia.org / wiki / Principal_component_analysis), spatial PCA (arxiv.org / pdf / 1501.03221v3.pdf, adegenet.r-forge.r-project.org / files / tutorial-spca.pdf, www.ncbi.nlm.nih.gov / pubmed / 1510870), and cluster analysis (en.wikipedia.org / wiki / Cluster_analysis, see US 9,336,302, 9,607,023 and cited references) can be used.

[0297] Typically, this type of neural network implementation will be able to receive unlabeled EEG data during operation and generate output signals representing the subject's predicted or estimated task, performance, context, or state during the acquisition of unclassified EEG. Of course, a statistical classifier can be used instead of a neural network.

[0298] EEG analysis, whether through conventional processing, neural network processing, or both, serves two purposes. First, it allows for inference of which areas of the brain are affected by which types of electrical activity under what conditions. Second, it allows for feedback during training (assuming appropriate spatial and anatomical correlation between the trainer and trainee) to help the system reach a desired state, or, where appropriate, a desired series of states and / or state transitions. According to one aspect of this technique, the applied stimulus depends on the initial state or condition of the measurement (which can represent a complex contextual and historical dependency matrix of parameters), and therefore the target represents a desired complex vector change. Thus, this aspect of the technique seeks to understand the complex temporal-spatial-brain activity associated with the trainee's activity or task, and seeks a corresponding complex temporal-spatial-brain activity associated with the same activity or task in the trainee, such that the trainee's complex temporal-spatial-brain activity state differs from the corresponding state sought to be achieved in the trainee. This allows for the transfer of training paradigms from individuals of different natures to achieve different results in different contexts and to some extent.

[0299] The conditions for acquiring data from trainees will include both task data and sensory stimulus data. That is, a preferred application of this system is to acquire EEG data from a trainee or technical individual, and then use this EEG data to transfer a learning, or more likely, learning readiness state to a novice trainee. The trainee's goal is to generate a set of stimulus parameters that will induce corresponding neural activity in the trainee's EEG state or lead to task execution when or before learning a skill or task.

[0300] It should be noted that EEG is not the only available data on neural or brain activity or state, and of course any and all such data can be included within the scope of this technique; therefore, EEG is only a representative example of the types of data that can be used. Other types include fMRI, magnetoencephalography (MEG), motor neuron activity, PET, etc.

[0301] While it is not necessary to map stimulus-response patterns different from those of the task to the trainees, doing so is advantageous because the trainees have extended timeframes available, during which stimuli to trainees may influence neural activity patterns, and it is possible that the trainees will correlate stimulus-response neural activity patterns with one or more trainees. It should be noted that the foregoing has indicated that the trainee is an individual; however, in practice, the trainee can be a group of trainees or technical individuals. Therefore, the analysis and processing of brain activity data can be adaptive both for each corresponding individual and for the entire group.

[0302] For example, a system might determine that not all human subjects share common stimulus-response brain activity correlations, thus requiring group separation and clustering. If the differences can be normalized, a normalization matrix or other corrections can be used. Conversely, if the differences do not allow for feasible normalization, one or more groups can be segmented, assigning different trainees to different market segments. For instance, in some tasks, the activity patterns and capabilities of the male brain differ from those of the female brain. This, coupled with anatomical differences between sexes, suggests that the system can provide sex-specific implementation schemes. Similarly, age differences can provide a plausible and scientific basis for group segmentation. However, depending on the required information base and matrix size, as well as several other factors, each system can provide essentially all the parameters needed for the entire group, with user-specific implementation schemes based on user profiles or initial settings, calibration, and system training periods.

[0303] According to one aspect of the invention, the source subject instrument is equipped with sensors for determining local brain activity during the experienced event. The aim is to identify the brain regions involved in processing this response.

[0304] Sensors typically attempt to determine neuronal firing patterns and brain region activation patterns, which can be detected using implanted electrodes, percutaneous electroencephalography (EEG), magnetoencephalography (MEG), fMRI, and other techniques. Percutaneous EEG is preferred where appropriate because it is non-invasive and relatively simple.

[0305] The source is observed using a sensor in a quiet state, where the source is experiencing an event and is in various control states, either at rest or engaged in different activities that lead to different states. Data can be obtained over sufficiently long periods and on repeated trials to determine the effect of duration. Data can also be population statistics and do not need to be obtained from a single individual in a single instance.

[0306] The sensor data is then processed using a 4D (or higher) model to determine characteristic location-related patterns of brain activity associated with the state of interest over time. This arousal state variable dimension is maintained even when the data is from a group with varying levels of arousal.

[0307] Next, the recipient prepares to undergo a mental state assessment. This assessment can include responses to questionnaires, sales evaluations, or other psychological evaluation methods. Furthermore, transcutaneous EEG (or other brain activity data) can be obtained to determine the recipient's initial state and activity during the expected mental state.

[0308] In addition, a set of stimuli, such as visual patterns, acoustic patterns, vestibular, olfactory, gustatory, tactile (light touch, deep touch, proprioceptive, stretching, heat, cold, pain, pleasure, electrical stimulation, acupuncture, etc.), and vagus nerve stimulation (e.g., parasympathetic), optionally within the baseline brain state, are applied to the subjects to obtain data defining the individual effects of sensory stimuli and the effects of various combinations on the subjects' brain states. Population data can also be used in this regard.

[0309] Then, data from the source or source group (see above) can be processed in combination with the recipient or recipient data group to extract information defining the optimal sensory stimulation over time to achieve the desired brain state and thus produce the desired mental state.

[0310] Typically, data processing tasks are daunting for both the source and receiver populations. However, statistical analysis often takes the form of allowing the parallelization of mathematical transformations used to process the data. This parallelization can be efficiently implemented using various parallel processors, with SIMD (Single Instruction Multiple Data) processors being a common form found in typical graphics processing units (GPUs). Due to the cost-effectiveness of GPUs, it is preferable to implement the analysis using efficient parallelizable algorithms, even if the computational complexity is calibrated to be greater than that of CISC-type processor implementations.

[0311] During stimulation of the recipient, EEG patterns can be monitored to determine whether the desired state has been achieved through sensory stimulation. A closed-loop feedback control system can be implemented to modify the stimulus seeking to achieve the goal. Evolutionary genetic algorithms can be used to develop user models that correlate mental state, arousal and valence, sensory stimulation, and brain activity patterns to optimize both the current stimulus and the learning period, as well as the future period, and to allow the system to be used for a range of mental states, provided the recipient's mental state is further enhanced.

[0312] Stimuli may include chemical messengers or stimuli used to alter a subject’s level of consciousness or otherwise alter brain chemistry or function. Chemicals may include hormones or endocrine analogs (such as adrenocorticotropic hormone [ACTH] (4-11)), stimulants (such as cocaine, caffeine, nicotine, phenylethylamine), psychoactive drugs, psychotropic substances, or hallucinogens (chemicals that alter brain function, thereby causing temporary changes in perception, mood, consciousness, and behavior, such as pleasure (e.g., euphoria) or dominance (e.g., increased alertness)).

[0313] While controlled or “illegal” substances are typically avoided, there are situations where they may be appropriate for use. For example, various drugs can alter the state of the brain to enhance or selectively amplify stimuli. Such drugs include stimulants (e.g., cocaine, methylphenidate (Ritalin), ephedrine, phenylpropanolamine, amphetamine), narcotics / opioids (opium, morphine, heroin, methadone, oxymorphine, oxycodone, codeine, fentanyl), hallucinogens (LSD), PCP, MDMA (psychedelic drugs), mescaline, psilocybin, magic mushrooms (Amanita muscaria), fly agaric mushrooms, cannabis / Hanax notoginseng), sage, diphenhydramine (Benadryl), cyclobenzaline hydrochloride, tobacco, nicotine, bupropion (Zyban). ), opioid antagonist inhibitors, gamma-aminobutyric acid (GABA) agonists or antagonists, NMDA receptor agonists or antagonist inhibitors (e.g., alcohol, Xanax, Valium, Halcion, Librium, other benzodiazepines, Ativapam, Klonopin, Amytal, Nembutal, Seconal, Phenobarbital, other barbiturates), psychedelics, dissociative agents, and delirium-inducing drugs (e.g., specific types of acetylcholine-depressant hallucinogens). For example, Carhart-Harris used fMRI to show that LSD and psilocybin synchronize different parts of the brain that are normally functioning by simultaneously firing neurons. This effect can be used to induce synchronization in various brain regions to improve mental state.

[0314] It has been noted that a wide range of natural and artificial substances can alter mood or arousal, and thus may affect affective or non-targeted mental states. Typically, such substances cross the blood-brain barrier and exert their psychoactive effects. However, this may not always be desired or appropriate. For example, painful stimuli can alter mood without acting as psychotropic drugs; on the other hand, anesthetics can also alter mood by dulling emotions. Furthermore, sensory stimuli, such as olfactory, visual, auditory, various types of tactile and proprioceptive sensations, balance, and vestibular stimulation, can induce mood and / or affective changes. Therefore, substances that alter sensory perception or the peripheral action of stimuli may be associated with mood. Similarly, pharmacological psychotropic drugs may alter alertness, perception, memory, and attention, which may be associated with task-specific mental state control.

[0315] Mental states can be associated with learning or performing skills. These skills can include mental skills such as cognition, alertness, attention, focus, concentration, memorization, visualization, relaxation, meditation, speed reading, creative skills, "whole-brain thinking," analysis, reasoning, problem-solving, critical thinking, intuition, leadership, learning, patience, balance, perception, linguistics or language, language comprehension, quantitative skills, "fluid intelligence," pain management, skills to maintain a positive attitude, foreign languages, music, music composition, writing, poetry, mathematics, science, art, visual arts, rhetoric, emotional control, empathy, compassion, motivational skills, people, calculation, scientific skills, or inventor skills. See U.S. Patents and Publications 6,435,878, 5,911,581, and 20090069707. The skills mentioned can include: motor skills, such as fine motor skills, muscle coordination, walking, running, jumping, swimming, dancing, gymnastics, yoga; sports or exercise; massage skills; martial arts or combat; shooting; self-defense; public speaking; singing; playing musical instruments; penmanship; calligraphy; painting; oil painting; visual, auditory, olfactory skills; playing games; sculpting; craftsmanship; massage or assembly skills. In cases where the goal is to enhance skills and to achieve (or suppress) emotions, stimuli to the recipient can be combined in a manner that achieves the desired outcome. In some cases, the components are general, while in others they are subjective. Therefore, the combination of components needs to be adapted based on the characteristics of the recipient.

[0316] This technology can be embodied in a device for acquiring brain activity information from a source, processing the brain activity information to reveal a target brain activity state and seeking a set of stimuli to achieve said state in a recipient, and generating stimuli for the recipient to achieve the target brain activity state and potential state transitions and to maintain the target brain activity state and potential state transitions for a period of time. The generated stimuli can be feedback-controlled. A general-purpose computer can be used to process the information, which is a microprocessor, FPGA, ASIC, system-on-a-chip, or special-purpose system, with a customized configuration to efficiently achieve the desired information transformation. Typically, the source and recipient operate asynchronously, with the brain activity of the source being recorded and subsequently processed. However, real-time processing and brain activity transfer are also possible. In the case of a general-purpose programmable processor implementation or part of the technology, computer instructions can be stored on a non-transient computer-readable medium. Typically, the system will have dedicated components, such as a transcranial stimulator or a modified audio and / or display system, and therefore the system will not be a general-purpose system. Furthermore, even in a general-purpose system, the operation itself can be enhanced according to this technology.

[0317] The mental state of the subject can be non-invasively induced by light, sound, transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tDAS) or HD-tACS, transcranial magnetic stimulation (TMS) or other means that can transmit frequency patterns.

[0318] The transmission of brain waves can be accomplished either through direct electrical contact with electrodes implanted in the brain or remotely using light, sound, electromagnetic waves, and other non-invasive techniques. Light, sound, or electromagnetic fields can be used to remotely transmit pre-recorded temporal patterns of brain waves to a subject by modulating the time frequency encoded on the light, sound, or electromagnetic field signals exposed to the subject.

[0319] Every activity, mental or motor, and emotional state is associated with unique brainwave patterns with specific spatial and temporal characteristics—that is, characteristic frequencies or frequency distributions over time and space. These waves can be read and recorded using several known techniques, including electroencephalography (EEG), magnetoencephalography (MEG), precise low-resolution electromagnetic tomography (eLORETA), sensory evoked potentials (SEP), fMRI, functional near-infrared spectroscopy (fNIRS), and others. The cerebral cortex is composed of neurons interconnected in a network. Cortical neurons constantly send and receive neural impulses—electrical activity—even during sleep. The electrical or magnetic activity measured by an EEG or MEG (or other device) device reflects the intrinsic activity of neurons in the cerebral cortex, as well as information sent to them by subcortical structures and sensory receptors.

[0320] It has been observed that playing back brainwaves in another animal or human with decoded temporal patterns delivered via transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), high-definition transcranial alternating current stimulation (HD-tDCS), transcranial magnetic stimulation (TMS), or via electrodes implanted in the brain allows the recipient to achieve an upcoming or accelerated mental state. For example, if the brainwaves of a mouse navigating a familiar maze are decoded (via EEG or via implanted electrodes), playing this temporal pattern to another mouse unfamiliar with the maze will allow it to learn to navigate the maze more quickly.

[0321] Similarly, recording brainwaves associated with a specific response in one subject and then “replaying” that response in another subject will induce a similar response in the second subject. More specifically, when a mental state is hypothesized for an animal, a part of the brain will exhibit characteristic activity patterns. Further, by “artificially” inducing the same pattern in another animal, the other animal will have the same mental state, or be more easily induced into said state. The pattern of interest may be deeply ingrained in the brain and thus submerged in the EEG signal by cortical potentials and patterns. However, other techniques besides surface electrode EEG can also be used to determine and spatially distinguish deep brain activity, for example, distinguishing it from the limbic system. For example, various types of magnetic sensors can sense deep brain activity. See, for example, 9,618,591; 9,261,573; 8,618,799; and 8,593,141.

[0322] In some cases, cortical activation patterns-dominated EEG can be used to sense mental states because cortical patterns may be associated with lower levels of brain activity. Note that it is not necessary to determine the state representation of a mental state every time the system is used; instead, once the spatial and temporal patterns of brain activity and synchronization states associated with a particular mental state are identified, these patterns can be used for multiple targets and over time.

[0323] Similarly, while the goal is, for example, to trigger a target based on the assumed pattern of brain activity, this can be achieved in various ways, and these methods for inducing the desired pattern do not need to be invasive. Furthermore, user feedback, especially in the case of a human subject, can be used to tune the process. Finally, the use of various senses, particularly visual, auditory, vestibular, tactile, proprioceptive, gustatory, olfactory, vagus nerve afferents, and other cranial nerve afferents, can be used to trigger high-level mental activity that achieves a desired mental state, emotion, or mood in a specific subject.

[0324] Therefore, in experimental subjects, which may involve laboratory-scale and / or invasive monitoring, a set of brain electrical activity patterns corresponding to specific emotional or mental states are identified. Preferably, these are also correlated with surface EEG findings. For the recipient, a harmless and non-invasive stimulation system is provided. For example, audiovisual stimulation alone may be used. EEG electrode sets are provided to measure brain activity, and adaptive or genetic algorithmic schemes are provided to optimize the audiovisual presentation in an attempt to induce the target patterns found in the experimental subjects within the recipient. After identifying stimulation patterns that may be path-related, it is possible that these patterns will persist, but over a longer period of time, there may be some degree of desensitization to one or more stimulation patterns. In some cases, audiovisual stimulation is insufficient, and TMS or other electromagnetic stimulation (overthreshold or preferably subthreshold) is used to assist in achieving the desired state and maintaining it for the desired period of time.

[0325] Brainwave transmission can be accomplished either through direct electrical contact with electrodes implanted in the brain or remotely using light, sound, electromagnetic waves, and other non-invasive techniques.

[0326] Light, sound, or inaccessible electromagnetic fields can be used to remotely transmit pre-recorded temporal patterns of brain waves to subjects by modulating the time frequencies encoded on the light, sound, or electromagnetic field signals they are exposed to.

[0327] The time pattern of brain waves (possibly pre-recorded) is transmitted to the subject remotely by modulating the encoded time frequency on the light, sound, or electromagnetic field signals to which the subject is exposed.

[0328] When a group of neurons are fired simultaneously, the activity manifests as brain waves. Different brain wave frequencies are associated with different mental states in the brain.

[0329] By providing selective stimulation based on temporal patterns, desired mental states can be induced in target individuals (e.g., humans, animals), where the temporal patterns are correlated with or represent shifts in the target's EEG patterns when the desired mental state is present, indicating a transition toward achieving the desired mental state. Temporal patterns can be targeted to discrete spatial regions of the brain via the physical arrangement of stimulators or the natural neural pathways transmitted by the stimulation (or its results).

[0330] EEG patterns can be obtained from one or more other individuals, the same individual at different times, or in vivo animal models of the desired mental state. Therefore, the method can replicate the mental state of the first subject in a second subject. A mental state is typically not a state of consciousness or ideation, but rather a subconscious (in a technical sense) state representing emotion, readiness, receptivity, or other states, usually independent of a particular thought or idea. Essentially, the mental state of the first subject (the “trainer” or “donor” in the desired mental state) is captured by recording the neural relevance of the mental state, for example, through brain activity patterns such as EEG or MEG signals. The neural relevance of the first subject, as a direct or recorded representation, can then be used to control stimulation of the second subject (“trainee” or “recipient”) in an attempt to induce in the second subject (recipient / trainee) the same brain activity patterns present in the first subject (donor / trainer) to assist the second subject (recipient / trainee) in acquiring the desired mental state already acquired by the donor / trainee. In an alternative embodiment, a signal from the first subject (donor / trainer) in a first mental state is used to prevent the second subject (recipient / trainee) from reaching a second mental state, wherein the second mental state is an undesirable sleep stage.

[0331] Source brainwave patterns can be acquired from a person in a desired brain state via multichannel EEG or MEG. Computational models of brain states are difficult to create. However, according to this technique, such models are unnecessary. Instead, the signal can be processed through statistical processes (e.g., PCA or related techniques) or statistically trained processes (e.g., neural networks). The processed signal preferably retains information about the specific location, frequency, and phase of the signal source. When stimulating the recipient's brain, the source can be modified to address issues such as differences in brain size and electrode placement. Therefore, the retained characteristics are normalized spatial features, frequencies, phases, and modulation patterns.

[0332] Normalization can be based on feedback from the target subject, such as comparing the target subject's current state with the corresponding state of the source subject, or other comparisons of known states between the target and the source. Typically, the excitation electrode in the target subject does not correspond to the feedback electrode or the electrode in the source subject. Therefore, other types of normalization are needed, which can also be based on statistical or statistically trained algorithms.

[0333] According to one embodiment, stimulation of the second subject is associated with a feedback process to verify that the second subject responds appropriately to the stimulus, for example, by exhibiting a predetermined similarity to the mental state of the first subject, a predetermined difference in mental state from the first subject, or a desired change compared to a baseline mental state. Advantageously, the stimulus can be adapted to the feedback. In some cases, the feedback can be functional, i.e., not based on the neural relevance of brain activity itself or mental state, but on reported or observed physical, psychological, or behavioral actions.

[0334] Feedback is typically provided to a computational model-based controller of the stimulator, which modulates stimulation parameters to optimize the stimulation based on a brain and brain state model applicable to the target.

[0335] For example, brain waves are believed to represent a form of resonance in which groups of neurons interact in a coordinated manner as a set of coupled or interacting oscillators. The frequency of the waves is related to the neural responsiveness to neurotransmitters, distance along the neural pathway, diffusion limitation, and perhaps pacemaker neurons or neural pathways. That is, the same mental state can be represented by different frequencies based on differences in brain size, the presence of neural modulators, physiological differences, etc., between two different individuals. These differences can be measured over microseconds or less, resulting in fractional variations in frequency. However, if the stimulus differs from the natural or resonant frequency of the target process, the result may differ from the expectation. Therefore, a model-based controller can determine parameters of neural transmission and group characteristics (relative to the stimulus) and resynthesize the stimulus wave to match the correct waveform, where waveform optimization is adaptively determined. This may not be as simple as speeding up or slowing down signal playback, because different elements of various brain waves representing the neural relevance of mental states may have different relative differences between subjects. Therefore, according to one set of embodiments, the stimulator automatically calibrates the target based on the correspondence (error) between the measured response to the stimulus and the desired mental state sought by the stimulus. In cases where existing data results are chaotic or unpredictable, genetic algorithms are employed to explore the range of stimulus parameters and determine the target's response. In some cases, based on a model maintained within the system, the target exhibits anomaly or unexpected responses to stimuli. In such cases, when deviations from the expected response are identified, the system can seek new models, such as from an online model library, like via the internet. If the model is predictable, a translation can be provided between the applicable model from the source or trainer and the applicable model for the target to resolve the discrepancy. In some cases, the desired mental state is relatively universal, such as sleep and wakefulness. In these cases, the brain response model can be a statistical model rather than an implementation of a neural network or deep neural network type.

[0336] Therefore, in one embodiment, a hybrid approach is provided that, on one hand, utilizes donor-sourced brainwaves extracted from brain activity readings (e.g., EEG or MEG) of at least one first subject (the donor), preferably processed by principal component analysis or spatial principal component analysis, autocorrelation or other statistical processing techniques (clustering, PCA, etc.) or statistically trained techniques (backpropagation of errors, etc.), wherein the donor-sourced brainwaves separate components of brain activity, which can then be modified or modulated based on high-level parameters (e.g., abstractions). See ml4a.github.io / ml4a / how_neural_networks_are_trained / . Thus, stimulators can be programmed to induce a sequence of brain states defined by names (e.g., SS1, SS2, etc.) or as a series of "abstract" semantic labels, icons, or other representations, each brain state corresponding to a technical sequence of brain states or substates. This sequence can be automatically defined based on biological and systemic training, thereby alleviating the low-level task for the programmer. However, in general, this technique maintains the use of components or subcomponents of the donor’s brain activity readings, such as EEG or MEG, and does not attempt to represent or abstract them to a semantic level.

[0337] According to this technique, neural network systems or statistical classifiers can be used to characterize brainwave activity and / or other data from the subject. In addition to classification or abstraction, reliability parameters are presented, which predict the accuracy of the output. At high accuracy, a model-based stimulator can be provided to select and / or parameterize the model and generate stimulation for the target subject. At low accuracy, a filtered signal representation can be used to control the stimulator, bypassing one or more models. The advantage of this hybrid approach is that, when using a model-based stimulator, many different parameters can be explicitly controlled independently of the source subject. On the other hand, if data processing fails to yield highly useful predictions of the correct model-based stimulator parameters, the model itself can be avoided, thus benefiting direct stimulation-type systems.

[0338] Of course, in some cases, one or more components of the stimulus to the target subject may be represented as an abstract or semantically defined signal, and more specifically, the processing of the signal used to define the stimulus will involve high-level modulation or transformation between source signals received from the first subject to define the target signal used to stimulate the second subject.

[0339] Preferably, each component represents a subset reflecting the neural correlations of brain activity, which is highly spatially and temporally autocorrelated or in a hybrid representation such as wavelets. For example, one signal could represent a modulated 10.2 Hz signal, while another signal could represent a superimposed modulated 15.7 Hz signal with distinct spatial sources. Once the spatial and temporal characteristics of the signals are known, they can be separated by optimal filtering, while remembering that signals are accompanied by modulation patterns and that the two components themselves may have some weak coupling and interaction.

[0340] In some cases, fundamental frequency, modulation, coupling, noise, phase jitter, or other signal characteristics can be substituted. For example, if the first subject is listening to music, there will be an important component of neural relevance synchronized with the specific music. On the other hand, the music itself may not be part of the stimulus desired by the target subject. Therefore, through signal analysis and decomposition, components of the signal from the first subject that have high temporal relevance to the music can be extracted or suppressed from the generated signal. Furthermore, the target subject may be in different acoustic environments, and the residual signal can be modified according to the target subject's acoustic environment so that the stimulus is suitable for achieving the desired effect and does not represent hallucinations, distractions, or irrelevant or inappropriate content. For performing processing, it is convenient to store the signal or a partially processed representation, but a complete real-time signal processing chain can be implemented. This real-time signal processing chain is typically characterized by a constant average buffer size, i.e., a relatively constant hysteresis between the output and input, considering that the processing may be periodic.

[0341] The mental state of a first subject can be identified, and neural correlations of brain activity can be captured. Based on the captured neural correlations and the identified mental state, a second subject is stimulated. Within a limited classification space, mental state can be represented as a semantic variable. Mental state identification does not require analysis of neural correlation signals and can be voluntary self-identification by the first subject, manual classification by a third party, or automatic determination. For example, the identified mental state is useful because its representation can be directed towards (or against) the manipulation of the second subject.

[0342] The stimulus can be one or more inputs to the second subject, such as electrical or magnetic transcranial stimulation, sensor stimulation, mechanical stimulation, ultrasound stimulation, etc., and can be controlled relative to waveform, intensity / amplitude, duration, feedback, self-reported effects by the second subject, manual classification by a third party, and automatic analysis of the second subject's brain activity, behavior, physiological parameters, etc.

[0343] The process can be used to induce neural correlations of a desired mental state in a target subject, the desired mental state being obtained from the same person at different times, simultaneously, or at different times from different people. For example, an attempt can be made to induce neural correlations of a first subject in a desired mental state in a second subject by using stimulation parameters that include waveforms derived from the neural correlations of a first subject's mental state over a period of time.

[0344] The first and second subjects may be spatially distant from each other and temporally distant. In some cases, the first and second subjects are the same subject, displaced in time (e.g., humans). In other cases, the first and second subjects are spatially close to each other. In some cases, the neural correlations of the desired mental state are derived from mammals with simpler brains, and these neural correlations are then extrapolated to the human brain. (Animal brain stimulation is also possible, for example, to enhance training and performance.) When the first and second subjects share a common environment, signal processing of neural correlations, and especially of real-time feedback on neural correlations from the second subject, may involve interactive algorithms for neural correlations with the first subject.

[0345] The first and second subjects can each receive stimulation independently. They can communicate with each other in real time, with the first subject receiving stimulation based on the second subject's signals, and the second subject receiving feedback based on the first subject's signals. This can lead to synchronization of mental states between the two subjects. However, the first subject does not need to receive stimulation based on real-time signals from the second subject, as the stimulation can be obtained from a third subject or from the first or second subject at different points in time.

[0346] Neural relevance can be, for example, EEG, qEEG, or MEG signals. Traditionally, these signals have been found to have dominant frequencies that can be determined through various analyses. One embodiment provides that the modulation pattern of a first subject's brainwaves is determined independently of the dominant frequency of the brainwaves (but typically within the same class of brainwaves), and this modulation is applied to a wave corresponding to the dominant frequency of a second subject. That is, once the second subject achieves the same brainwave pattern as the first subject (which can be achieved by means other than electromagnetic, mechanical, or sensor stimulation), the modulation pattern of the first subject is applied in a manner that guides the mental state of the second subject.

[0347] The second subject can be stimulated with a stimulus signal that faithfully represents the frequency composition of the defined components of the neural relevance of the first subject.

[0348] The stimulation can be performed, for example, using a tDCS device, a high-definition tDCS device, a tACS device, a TMS device, a depth TMS device, and a source of an optical signal and an audio signal configured to modulate a dominant frequency on one of the optical signal and an audio signal. The stimulation can be at least one of an optical signal, an audio signal, an electrical signal, and a magnetic field. The electrical signal can be a direct current signal or an alternating current signal. The stimulation can be transcranial electrical stimulation, transcranial magnetic stimulation, depth magnetic stimulation, optical stimulation, or audio stimulation. The visual stimulation can be ambient light or direct light. The auditory stimulation can be binaural beats or isochronous tones.

[0349] This technology can also provide a processor configured to process neural correlations from the mental state of a first subject and selectively generate or define stimulation patterns for a second subject based on waveform patterns of the neural correlations from the first subject. Typically, the processor performs signal analysis and calculates at least the dominant frequencies of the brainwaves of the first subject, and preferably the spatial and phase patterns within the brain of the first subject.

[0350] The signal is presented to a second device configured to stimulate the second subject. This stimulation may be an open-loop stimulus dependent on a non-feedback control algorithm or an algorithm dependent on closed-loop feedback. In other cases, analog processing is employed in part or in whole, wherein the algorithm includes an analog signal processing chain. The second device receives information from the processor (first device), typically including a representation of a portion of a waveform in terms of neural relevance. The second device generates a stimulus designed to induce a desired mental state in the second subject, the stimulus representing, for example, the same mental state present in the first subject.

[0351] A typical procedure performed on neural relevance is filtering to remove noise. For example, notch filters can be provided at 50Hz, 60Hz, 100Hz, 120Hz, and other overtones. Other ambient signals can also be filtered in a frequency-selective or waveform-selective (time-based) manner. Higher levels of filtering, as known in the art, can also be employed. The noise-filtered neural relevance can then be encoded, compressed (lossless or lossless), encrypted, or otherwise processed or transformed. The stimulator associated with the second subject will typically perform decoding, decompression, decryption, inverse transformation, etc.

[0352] Information security and copy protection technologies similar to those used for audio signals can be employed to protect neurally correlated signals from copying or content analysis before use. In some cases, it may be possible to use encrypted signals stored in their encrypted form without requiring decryption. For example, using asymmetric encryption schemes that support distance determination. See US7,269,277; Sahai and Waters (2005), Annual International Conference on the Theory and Applications of Cryptographic Techniques, pp. 457-473, Springer, Berlin, Heidelberg; Bringer et al. (2009), IEEE International Conference on Communications, pp. 1-6; Juels and Sudan (2006), Designs, Codes and Cryptography, 2:237-257; Thaker et al. (2006), IEEE International Conference on Workload Characterization, pp. 142-149; Galil et al. (1987), Conference on the Theory and Application of Cryptographic Techniques, pp. 135-155.

[0353] Because the system could potentially be intrusive, it might be desirable to authenticate the stimulator or the parameters it employs before use. For example, the stimulator and its parameters could be authenticated via a distributed ledger, such as a blockchain. Alternatively, in a closed system, digital signatures and other layered authentication schemes could be employed. Permissions for certain processes can be defined and executed according to smart contracts, whose automatic permissions (i.e., cryptographic authorization) are provided from a blockchain or distributed ledger system. Of course, centralized management can also be used.

[0354] In practice, the feedback signal from the second subject can be encoded accordingly based on the source signal, minimizing the error between the two. In this algorithm, the signal attempting authentication is typically fed into the error tolerance of the encrypted signal before available feedback becomes available. One way to achieve this is to provide a predetermined range of acceptable authenticable signals and then encode these acceptable authenticable signals such that authentication occurs when the presumed signal matches any predetermined range. In the case of neural correlations, a large set of numerical hash patterns representing different signals can be provided. The net result is relatively weak encryption, but the encryption strength can still be high enough to mitigate risk.

[0355] The processor can perform noise reduction differently than band filtering. Neural correlations can be transformed into a sparse matrix, and in the transform domain, components representing high-probability noise are masked while components representing high-probability signals are preserved. This distinction can be optimized or adaptive. In some cases, the components representing important modulations may not be known a priori. However, depending on their role in inducing the desired response in a second subject, "important" components can be identified, and the remainder can be filtered or suppressed. The transformed signal can then be inversely transformed and used as the basis for the stimulus signal.

[0356] Mental state modifications, such as brain entrainment, can be provided to determine the mental state of multiple first subjects; for example, using either EEG or MEG to acquire brainwaves of the multiple first subjects to create a dataset containing representations of brainwaves of the multiple first subjects. The database can be encoded using classifications of mental states, activities, environments, or stimulus patterns applied to the multiple first subjects, and the database can contain brainwaves acquired across, for example, a large number of mental states, activities, environments, or stimulus patterns. In many cases, database records will reflect characteristic or dominant frequencies of the corresponding brainwaves. As discussed above, the trainer or first subject is a convenient source of stimulus parameters, but not the only available source. The database can be accessed based on its index, such as mental state, activity, environment, or stimulus pattern, and stimulus patterns defined by the database records of one or more subjects for second subjects.

[0357] One or more records retrieved are used to define the stimulation pattern for the second subject. The selection and use of records can depend on the second subject and / or feedback from the second subject. As a relatively trivial example, female second subjects may be stimulated primarily based on records from female first subjects. Of course, a more subtle approach would be to process the entire database and stimulate the second subject based on a global EEG stimulation model, but this is unnecessary, and the underlying basis of the model may prove unreliable or inaccurate. In fact, it may be preferable to obtain stimulation waveforms from only a single first subject to preserve the micro-modulation aspects of the signal, which, as mentioned above, have not yet been adequately characterized. However, the selection of one or more first subjects does not need to be static and can be changed frequently. The selection of first subject records can be based on group statistics of the records from other users (i.e., collaborative filtering, i.e., whose response patterns have the highest correlation, etc.). The selection of first subject records can also be based on feedback patterns from the second user.

[0358] The stimulation process may attempt to target a desired mental state within a second subject, which is determined automatically or semi-automatically or manually input. The target then represents part of a query against a database to select one or more desired records. Record selection can be a dynamic process, and reselection of records can be feedback-relevant.

[0359] The recording can be used to define the modulated waveform of a synthesized carrier or a set of carriers, and the process can involve frequency-domain multiplexed multi-carrier signals (which are not necessarily orthogonal). Multiple stimuli can be applied in parallel through the affected subchannels and / or through different stimulator electrodes, magnetic field generators, mechanical stimulators, sensory stimulators, etc. Stimulations for different subchannels or modes do not necessarily have to be derived from the same recording.

[0360] The stimulation can be applied to achieve a desired mental state, for example, through brain entrainment of a second subject with one or more first subjects. Brain entrainment is not the only possible outcome of this process. If the multiple first subjects entrain each other, each subject will have a corresponding brainwave pattern depending on the basis of the brainwave entrainment. This connection between the first subjects may help determine the compatibility between the corresponding first subject and the second subject. For example, characteristic patterns of the entrained brainwaves can be determined, even for different target mental states, and these characteristic patterns can be correlated to find relatively close matches and exclude relatively poor matches.

[0361] This technology can also provide a foundation for social networks, dating sites, employment or career testing, or other interpersonal settings where people can match each other based on coupling characteristics. For example, people who effectively couple with each other can have better compatibility compared to those who do not couple. Therefore, instead of seeking to match people based on personality profiles, matching can be based on each party's ability to effectively couple with the other's brainwave patterns. This enhances nonverbal communication and assists in achieving a corresponding state during activities. This can be assessed by monitoring each individual's neural response to video and by providing test stimuli based on the brainwave correlation of the other party's mental state to understand whether coupling is effectively achieved. On the other hand, the technology can be used to assist coupling when natural coupling is inefficient, or to block coupling when it is undesirable. An example of the latter is hostility; when two people couple in a hostile environment, emotional escalation is ensured. However, if coupling weakens, undesirable escalation can be resisted.

[0362] As discussed above, the multiple first subjects can have their respective brainwave patterns associated with individual database records. However, they can also be combined into a more global model. One such model is a neural network or deep neural network. Typically, this network will have recursive features. The neural network is trained using data from the multiple first subjects, and then accessed by inputting a target state and / or feedback information, and the neural network outputs a stimulation pattern or parameters for controlling the stimulator. When the multiple first subjects form the basis of the stimulation pattern, it is preferable that the stimulator is then controlled using the neural network output parameters derived from brainwave patterns or other neurally relevant features from the mental states of the multiple first subjects and including said features, such as generating one or more of its own carrier waves, which are then modulated based on the output of the neural network. The neural network does not need to periodically retrieve records and can therefore operate in a more sustained manner, rather than a more granular scheme of control based on records.

[0363] In any feedback correlation method, brainwave patterns or other neural correlations of mental states can be processed through neural networks to generate outputs that guide or control stimuli. The stimulus is at least one of the following: a light (visual) signal, a sound signal, an electrical signal, a magnetic field, vibration, or mechanical stimulation, or other sensory input. These fields can be static or dynamically changing.

[0364] The process can employ a relational database of mental states and brainwave patterns, such as frequency / neural correlation waveform patterns associated with corresponding mental states. The relational database may include: a first table, further comprising multiple data records of brainwave patterns; and a second table, comprising multiple mental states, each associated with at least one brainwave pattern. The mental states and the associated brainwave patterns are stored and maintained in the relational database. The relational database is accessed by receiving a query for a selected mental state, and data records representing the associated brainwave patterns are returned. The brainwave patterns retrieved from the relational database can then be used to modulate a stimulator in an attempt to selectively exert an effect based on the mental state in question.

[0365] A computer device may be provided for creating and maintaining a relational database of mental states and frequencies associated with those mental states. The computer device includes: non-volatile memory for storing a relational database of neural correlations between mental states and brain activity associated with those mental states, the database including: a first table further including multiple data records of neural correlations between brain activity associated with the mental states; and a second table including multiple mental states, each of which is associated with one or more records in the first table; a processor coupled to the non-volatile memory, the processor being configured to process relational database queries and then use the relational database queries to look up the database; RAM coupled to the processor and the non-volatile memory for temporarily holding database queries and data records retrieved from the relational database; and an I / O interface configured to receive database queries and deliver data records retrieved from the relational database. SQL or NoSQL databases may also be used to store and retrieve records.

[0366] Another aspect of this technology provides a method of brain entrainment, the method comprising: determining the mental state of a first subject; recording brain waves of multiple subjects using at least one single channel of an EEG and a MEG; storing the recorded brain waves in a physical memory device; retrieving the brain waves from the memory device; and applying a stimulation signal to a second subject via transcranial stimulation, comprising a brain wave pattern derived from at least one single channel of the EEG and the MEG, thereby achieving the desired mental state in the second subject. The stimulation may have the same order (number of channels) as the EEG or the MEG or a different number of channels, typically reduced. For example, the EEG or the MEG may include 128 or 256 channels, while the transcranial stimulator may have 8 or fewer channels. Various modalities and patterns of sensory stimulation may accompany the transcranial stimulation.

[0367] The at least one channel may be fewer than six channels, and the placement of the electrodes for transcranial stimulation may be approximately the same as the placement of the electrodes for recording one of the EEG and MEG.

[0368] This technique can respond to chronobiology and specifically to subjective time perception. For the subject, this can be determined subjectively and voluntarily, but it can also be determined automatically, for example, by judging the span of attention through eye movements and analyzing the persistence of brain wave patterns or other physiological parameters after discrete stimuli. Furthermore, the brain's time constant can be analyzed through delay and phase. Further, situational negative changes (CNVs) prior to voluntary actions can be used to determine (or measure) both the timing of conscious actions and, more generally, the temporal relationship between thought and action.

[0369] Typically, brainwave activity is measured using a large number of EEG electrodes, each receiving signals from cells on the scalp or, in the case of MEG, via multiple sensitive magnetic field detectors. These EEG electrodes are responsive to local field differences. Typically, brainwave capture is performed over a relatively high number of spatial dimensions, e.g., corresponding to the number of sensors. Given that brainwaves are generated by billions of neurons connected by axons over long distances, it is generally impractical to process the brainwave signals to create a source model. Furthermore, these neurons are typically nonlinear and interconnected. However, a source model is not required.

[0370] Various types of artificial intelligence (AI) techniques can be used to analyze the neural correlations of SS (Spiritual Discrimination) represented in the brain activity data of both the first subject (donor) (or multiple donors) and the second subject (recipient). While in some cases the algorithms or implementation schemes do not need to be identical, it is useful to ensure consistency in the methods of source processing and feedback processing so that the feedback does not reach or seeks a suboptimal target SS. However, given the potential differences in conditions, resources, equipment, and objectives, it is not necessary to harmonize these processes. AI can take the form of neural networks or deep neural networks, but rule-based / expert systems, hybrid approaches, and more classical statistical analyses can be used. Typically, AI processes will have an aspect that is nonlinear in their output response to input signals and therefore violates at least one aspect of the principle of linear superposition. Such systems tend to permit discrimination because the process of decision-making and making decisions is ultimately nonlinear. AI systems require a basis of experience or information for training. This can be supervised (external labels applied to the data), unsupervised (self-discrimination of classes), or semi-supervised (part of the data is externally labeled).

[0371] The system can be tuned using self-learning or genetic algorithms, involving signal processing at both the donor and recipient systems. In a genetic algorithm-based feedback-relevance self-learning system, the responders (e.g., the target) to various types of stimuli can be determined in the stimulus space. This stimulus can be provided in the context of use, along with a specific target SS, or it can be unrestricted. The stimulator can operate using a library of stimulus patterns or attempt to generate synthetic patterns or modifications of patterns. Over time, the system will learn the optimal context-relevance parameters for mapping desired SSs to stimulus patterns.

[0372] This technique can be used to create a desired SS within a recipient's body, or to eliminate an existing SS within the recipient's body. In the latter case, deciding which final state to achieve is subject to fewer constraints, and therefore the optimization differs. For example, in the former case, achieving a specific desired SS might be difficult, requiring a set of transitions to initiate / prepare the recipient's brain to enter the target state. When the system attempts to eliminate an unwanted SS, the problem primarily concerns which path to take to most effectively leave the current state, considering various costs such as the comfort / discomfort of the stimulus, time-value costs, etc. Therefore, even with the same endpoints, a series of states differ in achieving these different goals; that is, the optimal algorithm for reaching state B from state A may differ from the optimal algorithm that exists in state A and terminates in state B.

[0373] This technique can be used to address the SS or its segments associated with dreaming. Typically, dreaming is associated with many different brain regions. Thus, the biology of dreaming is diverse. Dreams typically have biochemical or hormonal components as well as physiological components that may decay from or be absent from the cognitive state. Dreaming has long been considered to occur largely during rapid eye movement (REM) sleep, but it has been reported that dreams also occur during non-REM sleep. However, if the dreamer wakes people during the REM phase of sleep, the dream is usually remembered. Additionally, it has been shown that, for example, dreams about the face are associated with increased high-frequency activity in specific brain regions involved in facial recognition, while dreams involving spatial perception, movement, and thinking are similarly associated with brain regions that process such tasks when awake. Therefore, although general brainwaves or other neural relevance obtained from the donor may be similar or identical, the stimulation used for the second subject (recipient) may differ in modality, spatial location, intensity / waveform, other stimulation parameters and types, and the type of feedback application employed.

[0374] It is known that people with more REM sleep and stronger theta (4Hz–7Hz) activity during REM sleep are better able to consolidate emotional memories. It has been shown (Blagrove) that attempting to disrupt dreams by artificially increasing theta waves could result in incorporating more lucid experiences into the dream. (See “Dreams act as overnighttherapy,” New Scientist, May 5, 2018). Transplanting theta frequency brainwaves from people who are vividly dreaming may also help achieve the same effect. Furthermore, instead of stimulating the subject's brain with synthetic theta frequencies (e.g., isotonic tones or ambient sound beats), stimulating the recipient's brain with donor brainwaves carrying second harmonics (and higher) in addition to the dominant theta frequency can induce the same type of dream; that is, if the donor dreams of people, the recipient will be more likely to dream of people, albeit different people, because the donor's brainwaves stimulate the recipient's visual cortex. This may be helpful in treating PTSD, stress management, phobias, and certain mental illnesses.

[0375] In some cases, it may be appropriate to administer medications or pharmacological agents that aid in achieving the target SS and are used for emotional states and / or dreams, such as melatonin, hypnotics or sedatives, sedatives (e.g., barbiturates, benzodiazepines, non-benzodiazepine hypnotics, appetite peptide antagonists, antihistamines, general anesthetics, cannabis and other herbal sedatives, methaqualone and analogues, muscle relaxants, opioids), which may include certain psychotropic drugs such as adrenaline, norepinephrine reuptake inhibitors, serotonin reuptake inhibitors, peptide endocrine hormones such as oxytocin, ACTH fragments, insulin, etc. Combining medications with stimulants can reduce the desired dosage of the medication and associated drug side effects.

[0376] Therefore, the aim is to provide a method for inducing sleep in a second subject, the method comprising: recording the brain activity patterns of a first subject (donor) who has fallen asleep; and inducing sleep in the second subject (recipient) by replicating the donor's brain activity patterns in the recipient's body.

[0377] The objective is also to provide a method for preventing a second subject (recipient) from falling asleep, the method comprising: recording the brain activity patterns of a conscious first subject (donor); and preventing the second subject (recipient) from falling asleep by replicating the donor's brain activity patterns in the recipient's body.

[0378] Another objective is to provide a method for preventing a subject from falling asleep, the method comprising: recording the brain activity patterns of the subject while asleep in a training phase; determining the brain activity patterns of the subject during an operational phase; and preventing the subject from falling asleep by disrupting the recorded brain activity patterns of the subject during the operational phase corresponding to the recorded brain activity of the subject while asleep.

[0379] Another objective is to provide a method for inducing sleep in a second subject (recipient), the method comprising: identifying the mental state of a first subject (donor); recording the donor's brain activity patterns if the donor is asleep; and inducing sleep in the recipient by replicating the donor's brain activity patterns within the recipient's body. The method may further include verifying that the recipient has fallen asleep.

[0380] Another objective is to provide a method for preventing a second subject (recipient) from falling asleep, the method comprising: identifying the mental state of a first subject (donor); recording the brain activity pattern of the first subject if the donor is awake; and preventing the second subject from falling asleep by replicating the brain activity pattern of the second subject. The method may further include verifying that the second subject is awake.

[0381] Another objective is a method for transferring a desired mental state from a first subject (donor) to a second subject (recipient), the method comprising: identifying the mental state of the donor; capturing the mental state of the donor by recording brain activity patterns; storing the brain activity patterns in a non-volatile memory; retrieving the brain activity patterns from the non-volatile memory; and transferring the donor's desired mental state to the recipient by inducing the brain activity patterns in the recipient, wherein the desired mental state is one of a sleep state and a wakeful state.

[0382] Another objective is a method for transplanting a desired SS from a first subject (donor) to a second subject (recipient), the method comprising: identifying the donor's SS; capturing the donor's SS by recording brain activity patterns; storing the brain activity patterns in a non-volatile memory; retrieving the brain activity patterns from the non-volatile memory; and transplanting the donor's desired SS to the recipient by inducing the brain activity patterns in the recipient, wherein the desired SS is one of SS1, SS2, and SS3.

[0383] Another objective is a method for transplanting a desired SS from a first subject (donor) to a second subject (recipient), the method comprising: identifying the donor's SS; capturing the donor's SS by recording brain activity patterns; storing the brain activity patterns in a non-volatile memory; retrieving the brain activity patterns from the non-volatile memory; and transplanting the donor's desired SS to the recipient by inducing the brain activity patterns in the recipient, wherein the desired SS is one of a REM SS and a non-REM SS.

[0384] Another ...

Claims

1. A system for inducing a desired state of mental arousal in a second subject, the system comprising: a. EEG input, used to receive the brainwave activity patterns of a first subject with a corresponding state of mental arousal; b. A processor configured to process received brainwave activity patterns to classify the brainwave patterns according to a statistical classifier or neural network, and to at least store waveform phase information of the brainwave activity patterns. c. A neurostimulator for inducing a state of mental arousal in a second subject that corresponds to the brainwave activity pattern of the first subject by stimulating the second subject, based on waveform phase information of the classified brainwave pattern and the brainwave activity pattern of the first subject.

2. The system of claim 1, wherein the desired state of mental arousal is sleep.

3. The system of claim 1, wherein the desired state of mental arousal is wakefulness.

4. The system of claim 1, wherein brain photography includes an electroencephalograph.

5. The system of claim 1, wherein the brainwave activity pattern of the first subject stimulates the second subject by performing at least one of visual and auditory stimulation on the second subject based on the frequency-related and modulation-related brainwave patterns of the first subject.

6. The system of claim 1, wherein the desired mental state of wakefulness comprises a series of mental states including at least one sleep cycle.

7. The system of claim 1, wherein the stimulus selectively responds to a determined mental state of the second subject before or during the stimulus.

8. The system of claim 1, wherein the stimulus is provided to the second subject depending on the predicted mental state of the second subject.

9. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. Input, used to receive information identifying the mental state of the first subject based on the brainwave activity pattern of the first subject; b. Non-volatile memory, configured to store the brainwave activity patterns of the first subject; c. At least one automated processor is configured to control a neurostimulator to replicate the desired mental state of the first subject in the second subject by introducing a brainwave activity pattern, including the frequency and phase patterns of the first subject's brainwave activity pattern, into the second subject using the brainwave activity pattern stored in the first subject as the modulation pattern of the stimulation.

10. The system of claim 9, wherein the desired mental state is one of a sleep state and a wakeful state.

11. The system of claim 9, wherein the mental state of the first subject is identified by automated brain activity classification, and the brainwave activity pattern is recorded in a non-volatile memory in the form of at least one of magnetoencephalography (MEG) activity and electroencephalography (EEG) activity.

12. The system of claim 9, wherein the brainwave activity pattern is recorded in the non-volatile memory in the form of a set of compressed waveforms, the set of compressed waveforms preserving frequency and phase relationship characteristics of multiple signal acquisition channels, and wherein the stimulation of the second subject depends on the preserved frequency and phase relationship characteristics of the multiple signal acquisition channels.

13. The system of claim 9, wherein replicating the desired mental state of the first subject in the second subject by inducing the brainwave activity pattern in the second subject comprises selectively applying at least one of the visual and auditory stimuli to the second subject based on a determined brainwave activity pattern prior to or concurrent with at least one of the visual and auditory stimuli.

14. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. A non-volatile digital data storage medium configured to store data representing frequency and phase patterns of multiple channels of the brainwaves of the first subject; b. A stimulator configured to induce a brainwave pattern in the body of the second subject, wherein the brainwave pattern mimics the mental state of the first subject when the brainwaves of the first subject are acquired using a stimulation pattern comprising frequency and phase patterns of multiple brainwave channels of the first subject. c. A sensor configured to determine the brainwave pattern of the second subject in parallel with stimulation of the second subject by the stimulator; as well as d. A control configured to read the non-volatile memory and selectively control the stimulator based on the stored data and the determined brainwave patterns of the first subject, to provide stimulation to the second subject using the stimulation patterns, including the frequency and phase patterns of multiple brainwave channels of the first subject.

15. The system of claim 14, wherein the mental state is a range including sleep and wakefulness.

16. The system of claim 14, wherein the stored data originates from at least one of a magnetoencephalography (MEG) sensor and an electroencephalography (EEG) sensor.

17. The system of claim 14, wherein the stimulator is configured to provide at least one of visual and auditory stimulation to the second subject based on a frequency-related brainwave pattern of the first subject's brainwaves.

18. The system of claim 14, wherein the sensor is configured to determine the mental state of the second subject during stimulation.

19. The system of claim 14, wherein the control is configured to control the stimulator to induce a series of mental states including at least one sleep cycle in the second subject.

20. The system of claim 14, wherein the stimulus is provided to the second subject depending on the predicted mental state of the second subject.

21. A system for inducing sleep in a second subject, the system comprising: a. A memory configured to record brainwave activity patterns, including frequency and phase patterns, of a first subject who has fallen asleep; as well as b. A neurostimulator configured to induce sleep in the second subject by introducing the frequency and phase patterns of the first subject's brainwave activity into the second subject's body through stimulation containing a recorded brainwave activity pattern of a sleeping first subject.

22. A system for inducing sleep in a second subject, the system comprising: a. A memory configured to store brainwave activity patterns, including frequency and phase patterns, of a conscious first subject; as well as b. At least one automatic controller configured to stimulate a second subject by means of a pattern modulated with a stored pattern of brainwave activity including frequency and phase patterns representing an unawakened first subject, to introduce the brainwave activity pattern of the first subject into the second subject, thereby inducing a recorded brainwave activity pattern including frequency and phase patterns of the unawakened first subject during the sleep of the second subject.

23. A system for inducing sleep in a subject, the system comprising: a. A memory configured to store brain activity patterns, including frequency and phase patterns, of the subject who has fallen asleep during a training phase; b. At least one automated processor configured to determine a pattern of brain activity, including brain wave frequency and phase pattern, of the subject during sleep in the operation phase; as well as c. A neurostimulator, under the control of an automated processor, configured to induce and maintain sleep in the subject by presenting the subject with stimulation selectively dependent on a determined pattern of brain activity and a recorded pattern of brain activity including brainwave frequency and phase patterns, during the operating phase, disrupting the determined pattern of brain activity, including brainwave frequency and phase patterns, corresponding to the recorded brain activity of the subject who is already asleep.

24. A system for inducing sleep in a second subject, the system comprising: a. At least one automated processor configured to read electroencephalogram (EEG) data and identify the mental state of a first subject; b. At least one memory configured to store data representing patterns of brain activity, said brain activity patterns including frequency and phase patterns of the first subject's brainwaves when the first subject is asleep; and c. At least one neurostimulator configured to induce sleep in a second subject by generating a neurostimulation pattern for stimulating the second subject, said neurostimulation pattern being modulated by a brain activity pattern including frequency and phase patterns of the brain waves of a first subject.

25. The system of claim 24, wherein the at least one automatic processor is further configured to verify that the second subject has fallen asleep.

26. A system for inducing sleep in a second subject, the system comprising: a. At least one automated processor configured to identify the mental state of the first subject; Depending on whether the first subject is awake, the brainwave activity of the first subject, including frequency and phase patterns, is recorded; as well as b. A stimulation generator configured to stimulate the first subject by means of neural stimulation comprising computer wave activity including frequency and phase patterns of the first subject, and to generate neural stimulation for the second subject by replicating the brain wave activity including frequency and phase patterns of the first subject during sleep of the first subject.

27. The system of claim 26, wherein the at least one automated processor is further configured to determine the mental state of the second subject.

28. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. At least one automatic processor configured to identify the mental state of the first subject based on brainwave frequencies and phase patterns, including brainwave data from an electroencephalogram (EEG) of the first subject; b. A non-volatile memory configured to store the brain activity frequencies and phase patterns, including brain waves; c. A stimulator configured to stimulate a second subject by using brain activity frequencies and phase patterns including the brainwaves of the first subject, thereby replicating the desired mental state of the first subject within the second subject, wherein the desired mental state is one of a sleep state and a wakeful state.

29. A system for replicating the mental state of a first subject in a second subject, the system comprising: a. A non-volatile memory configured to record the brainwaves of the first subject; b. A stimulator configured to deliver sensory stimulation to a second subject, the sensory stimulation comprising the frequency and phase patterns of recorded brain waves of the first subject, to replicate the mental state of the first subject in the second subject by entraining the brain waves of the first subject in the second subject, wherein the mental state is one of a wakeful state and a sleep state.

30. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. At least one automated processor configured to identify the mental state of the first subject; b. At least one memory configured to simultaneously record the brainwaves of the first subject under the desired mental state; as well as c. At least one multichannel neurostimulator configured to replicate the desired mental state of the first subject in the second subject by using a stimulation pattern including recorded brain waves to provide sensory stimulation to the second subject, the stimulation pattern maintaining the frequency and phase pattern of the recorded brain waves, carrying the recorded brain waves of the first subject in the second subject, wherein the desired mental state is one of a sleep state and a wakeful state.

31. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. At least one automated processor configured to identify the mental state of the first subject; b. Non-volatile memory configured to record brainwaves of the first subject in a desired mental state; c. At least one automatic processor configured to generate a stimulation signal for a second subject, adapted to stimulate the second subject by selectively depending on the frequency and phase of brain waves and the simultaneous mental state of the second subject, inducing the brain waves of the first subject within the second subject to replicate the desired mental state of the first subject within the second subject, wherein the desired mental state is one of a sleep state and a wakeful state.

32. The system according to any one of claims 28 to 31, wherein the at least one automatic processor is further configured to identify the mental state of the second subject to verify that the second subject has the desired mental state.

33. The system according to any one of claims 29 to 31, wherein the recorded brainwaves are one of EEG, qEEG, and MEG.

34. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. A first device, the first device being used to record brainwaves of the first subject in a desired mental state; b. A non-volatile memory coupled to the first device for storing the recording of the brain waves; as well as c. A second device for inducing the brainwaves in the second subject to replicate the frequency and phase patterns of the brainwaves of the first subject in the desired mental state of the first subject, the second device being configured to receive the recording of the brainwaves of the first subject from the non-volatile memory, wherein the desired mental state is one of a sleep state and a wakeful state.

35. The system of claim 34, wherein the first device is one of an electroencephalogram (EEG) and a magnetoencephalogram (MEG).

36. The system of claim 34, wherein the second device is one of: a tDCS device, a tACS device, an HD tDCS device, a TMS device, a depth TMS device, a source of an optical signal and an audio signal, the source being configured to upmodulate a brainwave frequency on the one of the optical signal and the audio signal.

37. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: a. At least one automated processor configured to identify the mental state of the first subject; b. A non-volatile memory configured to record at least one of the EEG and MEG of the first subject, the first subject being in a desired mental state; c. At least one automated processor is configured as follows: At least one of the EEG and MEG signals is processed to generate a processed signal, including the frequency and phase of at least one of the EEG and MEG of the first subject; The processed signal is stored in non-volatile memory; The processed signal is retrieved from the non-volatile memory; The processed signal is modulated on at least one stimulus; as well as The desired mental state of the first subject is replicated in the second subject by stimulating the second subject with the at least one stimulus, wherein the desired mental state is one of a sleep state and a wakeful state.

38. The system of claim 37, wherein the at least one automatic processor is further configured to remove noise from at least one of the EEG and MEG signals.

39. The system of claim 37, wherein the at least one automatic processor is further configured such that the processing includes at least one of compressed EEG and MEG signals.

40. The system of claim 39, wherein the at least one automatic processor is further configured to decompress the at least one processed EEG and MEG signals retrieved from the non-volatile memory.

41. The system of claim 37, wherein the at least one stimulus is at least one of an optical signal, an acoustic signal, an electrical signal, and a magnetic field.

42. The system of claim 37, wherein the at least one stimulus is a DC unipolar signal.

43. The system of claim 37, wherein the at least one stimulation comprises transcranial electrical stimulation.

44. The system of claim 37, wherein the at least one stimulus is tACS.

45. The system of claim 37, wherein the at least one stimulation comprises transcranial magnetic stimulation.

46. ​​The system of claim 45, wherein the transcranial magnetic stimulation is depth magnetic stimulation.

47. The system of claim 37, wherein the at least one stimulus is either ambient light or direct light.

48. The system of claim 37, wherein the at least one stimulus is one of binaural beat and isochronous tone.

49. A system for replicating a desired mental state of a first subject in a second subject, the system comprising: At least one of an electroencephalogram (EEG) and a magnetoencephalogram (MEG) is used to record brain waves of the first subject, who is in a desired mental state. A processor coupled to one of an electroencephalogram (EEG) and a magnetoencephalogram (MEG), the processor being configured to perform signal analysis and calculate at least one dominant frequency of the brain waves of the first subject; A non-volatile memory coupled to a first processor for storing at least one dominant frequency and phase pattern of the brainwaves of the first subject; as well as A second device is configured to induce the brainwaves in a second subject to replicate the desired mental state of the first subject in the second subject. The second device is configured to receive at least one dominant frequency and phase pattern of the brainwaves of the first subject from the non-volatile memory, and to present a stimulus to the second subject containing the at least one dominant frequency and phase pattern, wherein the desired mental state is one of a sleep state and a wakeful state.

50. The system of claim 49, wherein the second device is at least one of: a tDCS device, a high-definition tDCS device, a tACS device, a TMS device, a depth TMS device, a light source capable of modulating the at least one dominant frequency in light, and a sound source capable of modulating the at least one dominant frequency in sound.

51. The system of claim 50, wherein the light source is one of a binaural beat source and an isochronous tone source.

52. A system for replicating the circadian rhythm of a first subject in a second subject, the system comprising: a. A memory configured to record one of the EEG and MEG of the first subject, the first subject having the desired phase of the circadian rhythm; b. At least one automated processor for: One of the records in EEG and MEG is processed to remove noise and preserve frequency and phase patterns; The processed EEG and MEG data are stored in non-volatile memory; Retrieve one of the processed EEG and MEG from the non-volatile memory; as well as The desired phase of the first subject's circadian rhythm is replicated in the second subject by replaying one of the processed EEG and MEG of the first subject via transcranial stimulation.

53. The system of claim 52, wherein the at least one automation processor is further configured to: Compress one of the records in the EEG and MEG before storing it in the non-volatile memory; and After retrieving one of the compressed EEG and MEG from the non-volatile memory, one of the records in the EEG and MEG is decompressed.

54. The system of claim 52, wherein the transcranial stimulation is one of tDCS, HD-tDCS, TMS, and deep TMS.

55. A system for replicating the circadian rhythm of a first subject in a second subject, the system comprising: a. At least one of an electroencephalogram (EEG) and a magnetoencephalogram (MEG), wherein the EEG and the MEG are respectively used to record one of EEG and MEG; b. A first processor, the first processor being coupled to one of the electroencephalogram (EEG) and magnetoencephalogram (MEG) instruments and configured to process one of the records in the EEG and MEG while preserving frequency and phase patterns, in order to remove noise by digital signal processing; c. Non-volatile memory, coupled to the processor for storing one of the processed EEG and MEG; as well as d. A transcranial stimulation device coupled to the non-volatile memory for replaying the processed EEG and MEG in the second subject to induce the circadian rhythm of the first subject in the second subject.

56. The system of claim 55, wherein the transcranial stimulation device is one of tDCS, HD-tDCS, TMS, and deep TMS.

57. The system of claim 55, wherein the first processor is further configured to compress one of the processed EEG and MEG.

58. The system of claim 55, further comprising a second processor coupled to the non-volatile memory and the transcranial stimulation device, the second processor being configured to decompress one of the compressed EEG and MEG.

59. A system for inducing brain activity cycles in the human body, the system comprising: A memory configured to record a set of brain activity cycles from a first subject and to tag the corresponding brain activity cycles with contexts preceding or accompanying the brain activity cycles. At least one automation processor is configured as follows: A set of recorded brain activity cycles is processed to normalize at least one of the amplitude, frequency, or time delay of the brain wave patterns represented in the set of brain activity cycles, while preserving at least one frequency and phase modulation pattern applied to the synchronous brain wave patterns representing coordination among neuronal groups. A set of brain activity cycles recorded based on the tags selected after processing; A stimulus is generated for a second subject, the stimulus comprising at least one frequency and phase modulation pattern applied to the synchronized brainwave pattern representing coordination between neuronal groups, the stimulus for the second subject being adaptively feedback-controlled based on at least the brainwave state of the second subject, the generation being controlled to synchronize the brainwave state of the second subject with the synchronized brainwave pattern representing coordination between neuronal groups of the first subject as reflected in a selected recording.

60. A system for modifying the mental state of a subject, the system comprising: The user interface is configured to receive the user's selection of the desired mental state; and At least one automated processor is configured to determine the subject's current brainwave pattern based on brain imaging input; Extract the characteristic frequencies and phases of the subject's current brainwave pattern; Determine the desired changes in the frequency and phase of the subject's current brainwave pattern in order to change the subject's current mental state to the desired mental state; A stimulation pattern is generated for the subject, which is time-synchronized with the characteristic frequencies and phases of the extracted subject's current brainwave pattern. The input is configured to receive brainwave data to monitor the subject's brainwave patterns after at least one stimulus. as well as Stimulus was modified based on the monitored brainwave patterns of the subjects; and A neurostimulator is configured to stimulate the subject with stimulation that is synchronized with the characteristic frequencies and phase time of the extracted subject's current brainwave pattern.

61. A system for altering the mental state of a subject, the system comprising at least one automated processor for controlling a neurostimulator, for: Determine the temporal relationship of the monitored neural or motor patterns of the subject, including frequency and phase patterns derived from electroencephalogram (EEG) recordings; The application of at least one stimulus to the subject depends on at least a determined temporal relationship; and The change in the subject's mental state is determined based on at least the change in the temporal relationship of the subject's monitored neural or motor patterns.

62. A system for enhancing non-REM deep sleep, the system comprising: A signal processor configured to statistically separate slow-wave sleep components from acquired brainwave patterns; At least one automated processor is configured to define stimulation patterns based on statistically separated slow-wave non-REM sleep components and the frequency and phase patterns of the acquired brainwave patterns; and A stimulator, which is configured to stimulate a subject with a defined stimulation pattern.

63. A system for enhancing deep non-REM sleep, the system comprising at least one automation processor configured to: Brain wave patterns, including frequency and phase, representing deep non-REM sleep states including slow-wave sleep were extracted from EEG recordings of endogenous brain activity of at least one subject. Statistical processing algorithms were used to separate the slow-wave non-REM sleep component from the EEG recordings of the endogenous brain activity of the at least one subject. The extracted brainwave pattern, including frequency and phase, after inversion processing; as well as The stimulator is configured to stimulate the subject with a reversed, processed, extracted brainwave pattern.

64. A neurostimulator comprising: A memory configured to store acquired brainwave patterns; At least one processor, said at least one processor being configured to: The slow-wave non-REM sleep component was statistically separated from the acquired brainwave patterns; as well as Based on the above, brain stimulation patterns are defined by statistically separating slow-wave non-REM sleep components; as well as An output signal generator is configured to present a defined brain stimulation pattern to a subject in order to introduce slow-wave non-rapid eye movement sleep components into the subject's brain.

65. A system for enhancing deep sleep, the system comprising: The memory is configured to store brain wave patterns, including frequency and phase, representing endogenous brain activity EEG recordings from at least one subject, including slow-wave non-REM sleep states and deep non-REM sleep states. At least one processor, the at least one processor being configured to process brainwave patterns including frequency and phase using a statistical processing algorithm to separate slow-wave non-REM sleep components from the endogenous brain activity EEG recordings of the at least one subject; as well as A stimulator configured to generate stimulation signals based on processed brainwave patterns including frequency and phase.

66. The system of claim 65, wherein the stimulator comprises a transcranial electrical stimulator.

67. A system for improving sleep in a second subject, the system comprising: The memory is configured to record brainwave activity patterns, including frequency and phase patterns, of a sleeping first subject. as well as A neurostimulator is configured to improve sleep by replicating the brainwave activity patterns, including frequency and phase patterns, of the first subject in the second subject.

68. The system of claim 67, wherein at least one of EEG and MEG is used to record the brainwave activity patterns of a sleeping first subject.

69. The system of claim 67, wherein the replication of the brainwave activity pattern of the first subject is accomplished by modulating at least one frequency of the brainwave activity pattern of the sleeping first subject on at least one stimulus presented to the second subject.

70. The system of claim 69, wherein the at least one stimulus is one of an optical signal, an acoustic signal, an electric current, and a magnetic field.

71. The system of claim 70, wherein the optical signal is either ambient light or parallel light.

72. The system of claim 70, wherein the sound signal is one of binaural beat and isochronous pitch.

73. A brain stimulator, comprising: The input is configured to receive brainwave activity data from the subject, including frequency and phase brainwave patterns; An analyzer configured to determine the sleep-wake state represented in the electroencephalogram (EEG) activity data, which includes frequency and phase brainwave patterns; A classifier configured to classify the EEG activity data, including frequency and phase brainwave patterns, relative to the sleep-wake state; The processor is configured as follows: The desired changes in the sleep-wake state represented in the EEG activity data, which includes frequency and phase brainwave patterns, are determined based on at least one cycle model of the sleep-wake state. The brain stimulation mode of the brain stimulator is controlled based on EEG activity data including frequency and phase brainwave patterns with corresponding classifications to achieve the desired change in the sleep-wake state without substantially directly waking the subject through stimulation.

74. The brain stimulator of claim 73, further comprising at least one of an auditory stimulator and a visual stimulator, wherein the auditory stimulator and the visual stimulator present the subject with signals that are substantially devoid of semantic, musical, or object content.

75. The brain stimulator of claim 73, wherein the brain stimulation mode is adapted to synchronize the subject's brainwave pattern with the modulated waveform.

76. The brain stimulator of claim 73, wherein the desired change in sleep-wake state is hemisphere-specific.

77. The brain stimulator of claim 73, wherein the processor is further configured to model the response of the brainwave activity data, including frequency and phase brainwave patterns, to the brain stimulation pattern and to tune the brain stimulation pattern to optimally achieve the desired change in the classification-related sleep-wake state.

78. The brain stimulator of claim 73, wherein the processor is further configured to normalize the EEG activity data including frequency and phase brainwave patterns relative to a population constant and to access the stimulation pattern database depending on the population constant.

79. The brain stimulator of claim 78, wherein the processor is further configured to denormalize stimulation patterns retrieved from the stimulation pattern database based on the difference between the subject's brainwave activity data and the population constant.

80. The brain stimulator of claim 73, wherein the processor is further configured to introduce a noise pattern having a random component into the brain stimulation pattern.

81. A system for inducing mental states corresponding to a predetermined sequence in a subject, the system comprising at least one automated processor configured to: Determine the predetermined order of mental states and the subject's current mental state; and At least one record from a database is processed based on the predetermined sequence of mental states and the past history of the subject's mental states to generate an optimal brain stimulation pattern, including at least one brainwave frequency and its phase relationship, for achieving the subject's target mental state; and A stimulator configured to selectively stimulate the subject with at least one of a direct brain stimulator and an indirect sensory input brain stimulator, depending on the optimal brain stimulation pattern.

Images