APPARATUS AND METHOD FOR "TRANSPLANTING" BRAIN STATES VIA BRAIN ENTRAINMENT

Abstract

Brain states, which correlate with specific motor, cognitive, and emotional states, are non-invasively monitored, representing macroscopic cortical activity manifested as oscillatory network dynamics. Sensory and/or transcranial stimulation, entraining brain rhythms, effectively induce desired brain states correlated with a state of sleep or state of attention. Brain waves are recorded from a donor which are then inverted by processing and used to entrain the brain of a recipient. Brain states may thus be transferred between people by acquiring an associated cortical signature from a donor, which, following processing, is applied to a recipient through sensory or transcranial stimulation. This technique provides an effective neuromodulation approach to the noninvasive, non-pharmacological treatment of a variety of psychiatric and neurological disorders for which current treatments are mostly limited to pharmacotherapeutic interventions.

Claims

1. A method of brain entrainment for transplanting a desired brain state, comprising: extracting a dominant brainwave frequency of at least one donor in the desired brain state; detecting an endogenous dominant brainwave frequency of a subject desirous of achieving the desired brain state; stimulating the subject's brain with at least one external stimulus having a frequency corresponding to the endogenous dominant brainwave frequency of the subject to lock phase and establish an Arnold tongue; and after establishing the Arnold tongue, incrementally changing the frequency of the external stimulus from the endogenous dominant frequency to the dominant brainwave frequency of said at least one donor's brain, thereby assisting the subject in achieving the desired brain state.

2. The method of claim 1, wherein the desired brain state is a sleeping state.

3. The method of claim 1, wherein the respective brain state is one of a waking state, a relaxation state, a state of hyper-focus, a state of flow, a state of altered consciousness, a meditative state, and an emotional state.

4. The method of claim 1, wherein said at least one donor is a plurality of donors, the method further comprising a step of: computing an average dominant frequency of brainwaves of said plurality of donors, wherein said plurality of donors are of the same biological sex and having an age within 5 years of the subject's age.

5. The method of claim 1, wherein the step of stimulating the subject comprises stimulating the subject with at least one of a visual, an auditory, a tactile, a transcranial electric stimulation, a transcranial direct current stimulation (tDCS), a transcranial alternating current stimulation (tACS), and a transcranial magnetic stimulation.

6. The method of claim 1, wherein the step of extracting the dominant frequency comprises performing one of a statistical pattern classification, a principal component analysis, a multilinear principal component analysis, a support vector machine analysis, an independent component analysis, a linear discriminant analysis, a quadratic discriminant analysis, a maximum entropy Markov model analysis, or a multinomial logistic regression.

7. The method of claim 1, wherein the step of extracting the dominant frequency over time comprises performing a clustering.

8. The method of claim 1, wherein the step of extracting the dominant frequency over time comprises performing a factor analysis.

9. The method of claim 1, wherein the step of extracting the dominant frequency over time comprises performing a EEG multifractal analysis.

10. The method of claim 1, wherein the step of extracting the dominant frequency over time comprises using a recurrent neural network or a deep neural network.

11. The method of claim 1, wherein the at least one external stimulus has a frequency defined using a recurrent neural network or a deep neural network.

12. The method of claim 1, wherein the external stimulus comprises at least one of binaural beats and isochronic tones.

13. The method of claim 1, wherein the at least one external stimulus comprises a visual stimulus selected from at least one of a red light and a near-infrared light.

14. A system for transplanting a desired brain state through brain entrainment, comprising: at least one memory configured to store an extracted dominant brainwave frequency of at least one donor in the desired brain state; at least one automated processor configured to: detect an endogenous dominant brainwave frequency of a subject desirous of achieving the desired brain state; stimulate the subject's brain with at least one external stimulus having a frequency corresponding to the endogenous dominant brainwave frequency of the subject to lock phase and establish an Arnold tongue; and after establishing the Arnold tongue, incrementally change the frequency of the external stimulus from the endogenous dominant frequency to the dominant brainwave frequency of said at least one donor's brain, to thereby assist the subject in achieving the desired brain state.

15. The system of claim 14, wherein said at least one donor is a plurality of donors, wherein the extracted dominant brainwave frequency of at least one donor in the desired brain state comprises a computed average dominant frequency of brainwaves of said plurality of donors, said plurality of donors being of the same biological sex and having an age within 5 years of the subject's age.

16. The system of claim 14, further comprising a stimulator configured to stimulate the subject's with at least one of a visual, an auditory, a tactile, a transcranial electric stimulation, a transcranial direct current stimulation (tDCS), a transcranial alternating current stimulation (tACS), and a transcranial magnetic stimulation.

17. The system of claim 14, further comprising an automated processor configured to extract the dominant frequency through one of a statistical pattern classification, a principal component analysis, a multilinear principal component analysis, a support vector machine analysis, an independent component analysis, a linear discriminant analysis, a quadratic discriminant analysis, a maximum entropy Markov model analysis, a multinomial logistic regression, a clustering, a factor analysis, and a EEG multifractal analysis.

18. The system of claim 14, further comprising a recurrent neural network or a deep neural network, configured to control the frequency of the external stimulus.

19. The system of claim 14, wherein the external stimulus comprises at least one of binaural beats, isochronic tones, a red light, and a near-infrared light.

20. A method of enhancing memory, comprising: providing a transcranial brain stimulation to a subject; determining a range of endogenous frequencies of brain oscillation of the subject after the transcranial brain stimulation; after the transcranial brain stimulation, entraining brainwaves of the subject with a first sensory stimulation pattern having frequencies within the determined range of endogenous frequencies of brain oscillation of the subject; after entraining brainwaves of the subject with the first sensory stimulation pattern, entraining brainwaves of the subject with a second sensory stimulation pattern having a distinct range of frequencies from the range of endogenous frequencies of brain oscillation of the subject, the second sensory stimulation pattern comprising a temporal pattern derived from a human brainwave pattern corresponding to a cognitive task; and engaging the subject in the cognitive task concurrent with exposure of the subject to the second sensory stimulation pattern.

21. The method according to claim 20, wherein the second sensory stimulation pattern is derived from the human brainwave pattern corresponding to a cognitive task through a deconvolution operation.

22. A system for enhancing memory, comprising: a stimulator configured to provide a transcranial brain stimulation to a subject; a user interface configured to engage the subject in a cognitive task; at least one processor configured to: entrain brainwaves of the subject with a first sensory stimulation pattern having frequencies within a range of endogenous frequencies of brain oscillation of the subject; entrain brainwaves of the subject with a second sensory stimulation pattern having a distinct range of frequencies from the range of endogenous frequencies of brain oscillation of the subject, the second sensory stimulation pattern comprising a temporal pattern derived from a human brainwave pattern corresponding to a cognitive task; control a sequence of the first sensory stimulation pattern and the second sensory stimulation pattern; and engage the subject through the user interface in the cognitive task concurrent with exposure of the subject to the second sensory stimulation pattern.

23. The system according to claim 19, wherein the at least one processor is further configured to deconvolve the human brainwave pattern corresponding to the cognitive task.

24. An apparatus for brain entrainment, comprising: a sleeping mask having an internal compartment for housing electronics; a motherboard having a memory module, a video controller, an audio controller, a wireless receiver, and a power connector, said motherboard positioned inside the sleeping mask with the power connector protruding outside the mask for connecting to a power cable; a first printed circuit board (PCB) containing a plurality of light-emitting diodes (LED) capable of emitting light in at least one of a red and a near-infrared portion of the spectrum, and controlled by the video controller, the first PCB position on the inner surface of the sleeping mask opposite a first eye; a second printed circuit board (PCB) containing a plurality of light-emitting diodes (LED) capable of emitting light in at least one of a red and a near-infrared portion of the spectrum, and controlled by the video controller, the second PCB position on the inner surface of the sleeping mask opposite a second eye, wherein light emitted by the LEDs on the first PCB and LEDs on the second PCB are synchronized with each other and light emitted by LEDs is modulated by a waveform having a frequency corresponding to a frequency characteristic of a desired brain state being extracted from brainwaves of at least one donor in the desired brain state; two earphones capable of emitting one of binaural beats and isochronic tones, said earphones controlled by the audio controller modulating the waveform on one of a frequency of sound and amplitude of sound, wherein binaural beats has the waveform frequency modulated on the frequency difference in each audio channel and the isochronic tones have the waveform frequency modulated on sound amplitude; a battery contained in a compartment inside the sleeping mask; wherein the apparatus being configured to stimulate a subject wearing the sleeping mask to achieve brain entrainment via Arnold tongue in order to achieve the desired brain state.

Description

BRIEF DESCRIPTION OF THE DRAWING

[0980] The FIGURE is an illustration of a system for transplantation of brain states. Electroencephalography (EEG) is used to register a cortical signature associated with a distinct brain state from a donor. This signal is digitally processed and inverted before being applied to a recipient via transcranial electric stimulation (TES).

DETAILED DESCRIPTION DF THE PREFERRED EMBODIMENTS

[0981] Brain States

[0982] Brain states refer to the synthesis of endogenous activity in part shaped by previous experience and current sensory input to create an overall state accessible to objective measurement. A brain state is sometimes described as a snapshot in time of the central nervous system (CNS). However, such a static picture taken in a particular moment in time does not reflect temporal patterns that are critical for describing the brain state. Brain states are expressed as patterns of active neurons, active synapses, and neural oscillations. When speaking of brain states, measurable patterns of neural activity and the ability to actively manipulate these spatiotemporal patterns in order to alter behavior are of concern. Some define brain states as patterns of synchronous neural firing (Brown, 2006). Brain states are interpreted here as macroscopic patterns of electrical activity in the brain that repeatedly occur as a function of endogenous neuronal dynamics and their response to sensory input. The neural correlates of brain states are rhythmic activity patterns resulting from neuronal oscillations. These rhythmic activity patterns represent measurable cortical signatures.

[0983] The concept of brain states has also been referred to as mental states. (Poltorak 2019) There is poor consensus about what mental states are and the nature of their relationship with the brain. According to identity theory, mental states are identical with brain states (Payne 2021). Others disagree (Chalmers 1995). Herein, the phrase brain states is employed to encompass the field, which concerns brain states as defined by measurable patterns of neural activity.

[0984] Brain states include conscious and unconscious states. Conscious states include inter alia the state of focus, the state of flow, and various emotional, motor, and cognitive states as well as altered states of consciousness. Unconscious states include various stages of sleep and the state of general anesthesia (Gervasoni et al., 2004).

[0985] An example of a state that is not a pure brain state is a state of relaxation, which is predominantly a state of the peripheral nervous system (PNS), not the central nervous system (CNS). The PNS regulates the relaxation state, which includes relaxation of the skeletal muscles regulated by the somatic nervous system, relaxation of smooth muscles as well as vasodilation, the slowing of the heartbeat and respiration, etc., all regulated by the autonomous nervous system (ANS). To induce this state, one must stimulate the vagus nerve, not the brain. Indeed, it is the excitation of the parasympathetic nervous system that causes the state of relaxation.

[0986] Sleep states are of particular interest, because each stage of sleep is (at least partially) defined by prominent rhythmic network activity in specific frequency bands (Steriade, 2006). The first non-REM stage of sleep, N1, is referred to as relaxed wakefulness. During this transitional stage the alpha waves (8-13 Hz) are replaced with the theta waves (5-7 Hz). The second non-REM stage of sleep, N2, is characterized by the appearance of sleep spindles (short bursts of high-frequency neural activity further subdivided into fast spindles of 11-13 Hz and slow spindles of 13-15 Hz (Cline, 2011)) and K-complexes (single long delta waves that last for only a second). The third non-REM stage of sleep, N3, is deep sleep, during which theta waves are replaced with delta waves (4 Hz and below). (See also (Lin et al., 2020).) Previously, N3 was subdivided into N3 and N4. For a summary of the four-stage sleep classification, see for example (Purves et al., 2001). It is worth pointing out that classification of sleep stages is not based exclusively on the detection of a single, dominant frequency but rather accounts for a series of criteria about frequency, amplitude, waveform shape, the occurrence of transient synchronized events (such as spindles and K-complexes), auxiliary signals about muscle tone and eye movements, and so forth.

[0987] The complexity of neural activity suggests that the reduction of brain states to an oscillatory signal of a single frequency can be expected to fail to capture the full complexity of brain states. Indeed, the conventional focus on individual oscillations and frequency bands as signatures of brain states is limiting. Nevertheless, the research literature reveals an almost exclusive focus on single-frequency stimulation waveforms for the modulation of brain states with noninvasive brain stimulation (such as the use of a sine wave in transcranial alternating current stimulation, tACS). Taking into account the rich temporal structure of brain activity beyond a single (dominant) frequency in the spectrum should enable more effective brain entrainment through stimulation. Thus, a naturalistic modulation waveform based on cortical signatures extracted from endogenous brain waves of a sleeping subject contains all attendant rhythms, which dynamically change in their prominence as the subject moves through various stages of the healthy sleep cycle. Using such a naturalistic, multifrequency dynamic waveform for stimulation is expected to be more efficacious than using a single static frequency, typically 5 Hz, that is traditionally used to induce sleep. Thus, neuromodulation with such naturalistic waveforms may prove beneficial in treating insomnia and other sleep disorders (Gebodh et al., 2019; Gebodh et al., MO).

[0988] Brain Waves as Neural Correlates of Brain States

[0989] Oscillating neurons produce emergent meso- and macroscale rhythmic electric signals often referred to as brain rhythms or brain waves, which can be detected and recorded using noninvasive techniques. Brain waves can be considered the neural correlates of brain states.

[0990] To be sure, brain waves are not identical with brain states and may not contain all information encoded in brain states. Measurements of brain waves, of course, are subject to all the limitations (such as noise or signal averaging across vast numbers of neurons) of the instruments being utilized, and loss of information is inevitable. However, a bilateral correlation between brain waves and brain states is posited to exist.

[0991] There is no doubt that brain states cause (via neuronal oscillations) brain waves. The question is, do brain waves cause brain states? Specific spatiotemporal patterns of neural activity in the brain that are correlated with particular brain states can also cause these brain states. So, if brain state A causes brain waves X (characterized by their frequencies and spatial distribution), replicating the same brain waves X (at least in the same person) results in brain state A. Therefore, the causal nature of the relationship can be experimentally investigated. This is more likely to hold true for brain states that are predominantly cortical states, because cortical brain states are more likely to be the product of cortical activity patterns such that, when a specific cortical activity is induced through stimulation, the brain switches to the corresponding state.

[0992] A growing number of studies have recently shown that modulating brain rhythms causes changes in cognitive performance, indicative of a causal role of brain oscillations in brain states (Grover et al., 2021; Romei et al., 2016; Frohlich et al., 2015). Sensory entrainment was shown to result in behavioral (perceptual) changes (Mathewson et al., HID). Therefore, a causal relationship that can be experimentally investigated is posited to exist. This is more likely to hold true for brain states that are predominantly cortical states, because cortical brain states are more likely to be the product of cortical activity patterns such that, when a specific cortical activity is induced through stimulation, the brain switches to the corresponding state.

[0993] On the other hand, this conjecture is less likely to hold for brain states that involve deep brain structures and specialized nuclei or the brain stem. For example, it can be imagined that certain spatiotemporal patterns are indicative of an underlying brain state not accessible with noninvasive measurements of brain activity. Imposing the corresponding activity patterns may fail to induce the corresponding brain state, since controlling (e.g., stimulating) inaccessible subcortical brain areas (or, more broadly, physiological processes) would be necessary. The state of sleep, for example, is regulated by the hypothalamus (including suprachiasmatic nuclei (SCN) responsible for circadian rhythms or the ventrolateral preoptic nucleus (VLPO)), brain stem, thalamus, pineal gland, basal forebrain, and amygdala. There is little reason to expect that the cortical signatures of the sleeping donor would entrain subcortical structures (such as the SEN or VLPO) of the recipient. Nevertheless, recent research confirms that cortical structures play a role in regulating sleep homeostasis and global sleep-wake dynamics (Krone et al., 2021), leaving room for the possibility that the state of sleep may be induced by entrainment.

[0994] This distinction results from the fact that neither recording nor stimulation techniques have access to all physiological processes that (at least in theory) contribute to a brain state.

[0995] Functional neuroimaging such as EEG or MEG can capture the neuronal activity of localized brain regions, which correlate with distinct cognitive or behavioral states (brain states). EEG recordings have demonstrated, for example, that the pattern of brain activity changes during meditative acts, and frontal cortex EEG activity has been associated with emotion induction and regulation (Yu et al., 2011; Dennis and Solomon, MD). EEG recordings reflect ionic fluctuations resulting from neuronal communication in the cortex arising from dendritic depolarizations (Nunez and Srinivasan, 2006). Alternatively, MEG measurements reflect intracellular ionic fluctuations resulting from action potentials (Hamalainen et al., 1993). In both cases, output measures correlate with localized cortical activity. EEG signals are not easy to localize. However, the overall macroscopic EEG signal reflects smaller scale endogenous rhythmic processes.

[0996] EEG or MEG recordings can be used to extract cortical signatures. This can be done using statistical techniques such as principal components analysis (PCA), independent components analysis (ICA), spectral analysis and similar techniques, or machine-learning pattern recognition.

[0997] Deconvolution of Brain States

[0998] As noted above, the stimulation pattern is defined based on brain states, e.g., EEG signals, but the stimulation needs to be inverted in order to have the desired effect. This is based on the fact that the EEG signals are produced by waves of depolarization of neurons in the brain, in a cyclic pattern. The EEG measures the convoluted pattern at the scalp. With a large number of signals, and knowledge of skull and brain anatomy, inferences may be drawn on the location and state of the population of neurons that generate components of the signal. For example, using statistical separation techniques such as principal component analysis and related techniques, independent sources may be inferred from complex superpositions. However, because neurons are linked, and share blood supply, cerebrospinal fluid, etc., the independence of sources is incomplete. Further, the superpositions of sources may be non-linear, further complicating the analysis. However, various techniques are known for addressing each type of deviation from the presumptions of independent sources, normal distribution statistics, and linearity.

[0999] The essential process for inversion is to determine what kind of stimulus leads to a desired brain wave pattern. To the extent that both desired and undesired waves are generated from a single stimulus, it is possible to suppress or cancel, or otherwise distinguish desired and undesired waves.

[1000] The system is generally considered underactuated, with, for example, two eyes and/or two ears available for sensory stimulation, and feasible skin or proprioceptive stimulation on the order of <16 stimuli, while the EEG may have 64 or 128 electrodes, and the brain model may have hundreds of relevant sources or voxels. See: [1001] Abutaleb, A., Aya Fawzy, and Khaled Sayed. Blind deconvolution of EEG signals using the stochastic calculus. In 2012 Cairo International Biomedical Engineering Conference (CIBEC), pp. 175-178. IEEE, 2012. [1002] Edlinger, Gtinter, Paul Wach, and Gert Pfurtscheller. On the realization of an analytic high-resolution EEG. IEEE transactions on biomedical engineering 45, no. 6 (1998): 736-745. [1003] Ehinger, Benedikt V., and Olaf Dimigen. Unfold: an integrated toolbox for overlap correction, non-linear modeling, and regression-based EEG analysis. PeerJ 7 (2019): e7838. [1004] Grandchamp, Romain, Claire Braboszcz, Scott Makeig, and Arnaud Delorme. Stability of ICA decomposition across within-subject EEG datasets. In 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 6735-6739. IEEE, 2012. [1005] Huang, Lu, and Hong WANG. Blind separation of EEG based on blind deconvolution. Journal of Northeastern University (Natural Science) 37, no. 8 (2016): 1100. [1006] Kazemipour, Abbas, Ji Liu, Krystyna Solarana, Daniel A. Nagode, Patrick D. Kanold, Min Wu, and Behtash Babadi. Fast and stable signal deconvolution via compressible state-space models. IEEE Transactions on Biomedical Engineering no. 1 (2017): 74-86. [1007] Nat-Ali, A. M., D. Adam, and J-F. Motsch. The brainstem auditory evoked potentials estimation using a bayesian deconvolution method. In Proceedings of 18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, vol. 4, pp. 1516-1517. IEEE, 1996. [1008] Nunez, Paul L. Spatial analysis of EEG. In Images of the Twenty-First Century. Proceedings of the Annual International Engineering in Medicine and Biology Society, pp. 691-692. IEEE, 1989. [1009] Panwar, Sharaj, Paul Rad, John Duarles, and Yufei Huang. Generating EEG signals of an RSVP experiment by a class conditioned wasserstein generative adversarial network. In 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC), pp. 1304-1310. IEEE, 2019. [1010] Penny, W., N. Trujillo-Barreto, and E. Aubert. Spatia-temporal models for EEG. Friston et al. (Eds.), Statistical Parametric Mapping: Models for Brain Imaging. Elsevier (2006). [1011] Sanei, S., and A. R. Leyman. EEG brain map reconstruction using blind source separation. In Proceedings of the 11th IEEE Signal Processing Workshop on Statistical Signal Processing (Cat. No. DITH8563), pp. 233-236. IEEE, 2001. [1012] Satherley, Brenda Lee. Zero-based ensemble deconvolution and EEG spectral topography. (1994). [1013] Sjontoft, Erik. A deconvolution procedure for use in extracting information in average evoked response EEG signals. IEEE Transactions on Biomedical Engineering 4 (1980): 227-230. [1014] Trujillo-Barreto, Nelson J., Eduardo Aubert-Vzquez, and William D. Penny. Bayesian M/EEG source reconstruction with spatia-temporal priors. Neuroimage 39, no. 1 (2008): 318-335. [1015] Wen, Tingxi, and Zhongnan Zhang. Deep convolution neural network and autoencoders-based unsupervised feature learning of EEG signals. IEEE Access B (2018): 25399-25410.

[1016] Neuromodulation

[1017] These cortical signatures may be inverted in order to stimulate cortical activity. Endogenous electric fields (i.e., brain waves) directly entrain neuronal network activity (Frohlich and McCormick, 2010). This fundamental discovery enables the design of low-amplitude electromagnetic brain stimulation modalities to modulate and enhance oscillatory network dynamics. Indeed, transcranial electric stimulation (TES) (Annarumma et al., 2018), including transcranial alternating current stimulation (tACS) (Antal et al., 2008) and transcranial direct current stimulation (tDCS) (Nitsche, 2000; see also Utz et al., 2010), are used to electrically stimulate cortical activity, whereas transcranial magnetic stimulation (TMS) (Barker, 1985; Lawson McLean, 2019) uses a spatially targeted magnetic field to achieve a similar endpoint of controlling electric activity in brain networks. Typical brain stimulation methods use constant stimulation as a waveform (e.g., tDCS) or a synthetic waveform, which may be a step function modulated on a direct current, such as electrosleep (Robinovitch, 1911), a sinusoid modulated on an oscillatory direct current (osc-tDCS) (D'Atri et al., 2016), or a single fixed frequency of alternating current (tACS) (Rosa and Lisanby, 2012; Herrmann, 2013). (Helfrich et al., 2014) utilized simultaneous tACS stimulation combined with EEG recordings to show that, when tACS was applied to the parieto-occipital lobe of the brain, alpha oscillations increased and became synchronized with the entrainment frequency. Similarly, brain entrainment can be achieved by stimulating the brain via sensory pathways using visual or auditory stimuli (Notbohm et al., 2016).

[1018] Whereas TES and TMS modulate neuronal activity, entrainment using sensory pathways can also change the brain state. Indeed, the endogenous circadian rhythm responsible for the transition from the sleeping state to the waking state is primarily entrained by light (Bedrosian et al., 2013).

[1019] Endogenous circadian rhythms are entrained-both in terms of the firing of action potentials and gene expression-to the diurnal cycle. Although such entrainment happens on a distinctly slower timescale than the neuronal oscillations discussed here so far, this serves as an example of how sensory input alters rhythmic processes in the brain. Rhythmic brain activity is a reflection of rhythmic processes in our environment (from the slow timescales of the day-night cycle to the fast oscillations in auditory input) across an extremely wide band of frequencies.

[1020] It is noteworthy that suprachiasmatic nuclei (SEN) owe their ability to sense the cycle of light and darkness to the retinohypothalamic tract (RHT), connecting the intrinsically photosensitive retinal ganglion cells (ipRGC) in the retina to SCN via the optic nerve (Cooley et al., 2001). The RHT provides a natural pathway to use photobiomodulation for stimulating the hypothalamus and, through it, other structures of the brain.

[1021] The brain is very efficient at processing light and sound stimuli. Therefore, stimulation with light (photobiomodulation) or sound (isochronic tones or binaural beats) holds significant promise for entraining a brain state. These natural stimuli have a greater likelihood of reproducing naturalistic brain rhythms than single frequency stimulation waveforms do. (See, for example, Prez et al., 2017).

[1022] Sensory stimulation is particularly promising since it uses established neuronal circuits for the brain to respond to external input and avoids the pitfalls inherent in recent debates about the amount of energy actually delivered to the brain by transcranial current stimulation. However, it is also conceivable that certain structural and functional brain networks are configured to be robust to perturbations caused by sensory input. These circuits may be preferentially modulated by electromagnetic stimulation.

[1023] These techniques may be used to transplant a desirable brain state from one subject (the donor), who is in the desired brain state, to another subject (the recipient, who is either another person or the same subject at another time), who wishes to attain this brain state. This may be achieved by recording and subsequently inducing desired brain states. Thus, EEG or MEG may be used to record the brain state of the first subject (the donor), from which cortical signatures may be extracted and inverted to create modulation waveforms. Such waveforms may then be modulated on various physical signals, such as direct or alternating current, magnetic field, light, sound, or vibration and applied via TES (tDCS, osc-tDCS, tACS), TMS, or sensory (visual, auditory, or tactile) stimulation to entrain the brain of the recipient with the cortical signatures of the donor, thereby inducing the cognitive-behavioral state of the donor within the recipient.

[1024] This technique (transplanting brain state) was tested in the domain of sleep (Gebodh et al., 2019; Gebodh et al., 2020). In these studies, TES (tACS modulated with endogenous sleep-derived waveform, tESD) was used. For the reasons explained above, light and sound appear to be more promising modalities for inducing the desired brain state than direct brain stimulation with TES would be.

[1025] Generally, the notion of transplanting brain states, including sleep, attention, and learning as well as emotional valence is of particular interest. Attention states in the brain primarily result from the cognitive control process of the selective direction of information processing resources to behaviorally relevant stimuli and from active suppression of the detection and processing of irrelevant stimuli (Gulbinaite et al., 2017). Thus, functionally, attention serves as both a filter, eliminating less relevant stimuli from conscious perception, and an amplifier, increasing the salience of behaviorally relevant stimuli. This cognitive state is associated with specific neuronal oscillations (Schroeder et al., 2010), which may be captured by EEG or MEG. The neural oscillations associated with attention have been shown to be disrupted in a number of conditions, including epilepsy (Besle et al., 2011), dyslexia (Thomson and Goswami, 2008; Leong and Goswami, 2014; Soltsz et al., 2013), and schizophrenia (Lakatos et al., 2013). Therefore, acquiring a brain wave signature during states of attention in a healthy donor may prove valuable when applied to a recipient exhibiting attention deficits associated with disrupted or otherwise irregular cortical oscillations.

[1026] Stimulating the brain using a waveform with a fixed frequency that significantly differs from the frequency of the endogenous brain waves (regular or irregular), as frequently done in TES and TMS studies, may cause interference and is unlikely to change the rhythm of endogenous brain waves. In contrast, brain entrainment should optimally start where the brain is, not where it is desired to be.

[1027] The Arnold tongue is a theoretical framework for entrainment, essentially suggesting that it is easier to entrain oscillations closer to endogenous frequencies within a given subject. Thus, neuromodulation with a dynamic waveform used for entrainment should start at the current frequency of the endogenous brain waves since such close matching of stimulation and endogenous frequency is required for phase locking as indicated by the Arnold tongue (Ali, 2013) for entrainment of neural oscillations. Once entrainment is achieved at this initial frequency, the frequency of stimulation can be gradually changed to move the endogenous rhythm toward the desired frequency in order to achieve successful entrainment at the desired target frequency (Notbohm et al., 2016; Thut et al., 2011). This approach also avoids interference issues.

[1028] Previous research shows that memory functions are acutely sensitive to neural entrainment and may be disrupted via TMS (Hanslmayr et al., 2014), indicating the possibility of an inverse, positive entrainment of these oscillations.

[1029] Similarly, emotional arousal and valence are correlated with distinct cortical signatures observable through EEG (Allen et al., 2018). Previous data indicate that happiness resulting from musical experience, for instance, is associated with increased theta frequency oscillations in the left frontal cortical hemisphere (Rogenmoser et al., 2016). Cortical oscillations associated with negative affect conversely correlate with decreased theta frequency oscillations in this same region. Notably, aberrant cortical oscillations have been observed in a range of affective disorders, including major depression (Van der Vinne et al., 2017). Indeed, the left frontal hemisphere exhibits disrupted cortical rhythms in patients diagnosed with major depression when compared with healthy controls (Nusslock et al., 2018). Similar data have highlighted cortical asymmetry of frontal lobe oscillations in post-traumatic stress disorder (PTSD) (Meyer et al., 2018). Simple cortical entrainment via binaural beat stimulation has already proven adequate for inducing specific emotional states (Chaieb et al., 2015). More directly, cranial electrotherapy has been demonstrated as an effective treatment for depression, anxiety, and certain forms of insomnia (Kirsch and Nichols, 2013). In fact, certain forms of depression may respond better to transcranial approaches, such as TMS, as has been demonstrated in early data on patients with treatment-resistant major depression (George, 2000; Rosenberg et al., 2010).

[1030] Transplanting Brain States

[1031] Thus, the transplanting (transferring) of brain states by replicating neural correlates of the donor's state in a recipient (who may be a different person or the same person at a different time) is founded on two primary principles. First, cortical signatures found in brain waves are neural correlates of brain states. A large body of literature has identified distinct, measurable cortical signatures associated with specific brain states, ranging from those defining the sleep/wake cycle to those underlying emotional experiences. Second, TES, TMS, and sensory stimulation by light and sound have been repeatedly demonstrated as efficacious, safe means by which cortical rhythms may be entrained with a high degree of location-specificity, with sensory stimulation using light or sound holding particular promise for brain entrainment. Third, brain waves resulting from brain entrainment causally induce the desired brain state associated with the cortical signatures that are encoded in the modulation waveforms used for entrainment. Thus, a bidirectional causal relationship is apparent between brain states and cortical signatures found in brain waves.

[1032] To be sure, brains differ. While the donor and the recipient may on occasion be the same person, when the donor and the recipient are not the same individual, they can be expected to differ in skull structure, brain size, and brain morphology. Moreover, the phase lag between oscillations and stimulus can differ across individuals during neural entrainment. The same is true for the optimal timing of tACS when it is applied to modulate entrainment. Although it is possible to adjust the waveform based on the specifics of the recipient brain obtained by fMRI and using computational models of the brain, this problem can be sidestepped entirely by using sensory stimulation that does not act on the brain itself but rather acts on the sensory organs and allows the brain to assimilate the signals it receives from these sensory organs on its own terms. This is yet another reason why sensory stimulation using visual and/or auditory pathways seems highly preferable to using transcranial electric (TES) or magnetic (TMS) stimulation.

[1033] The complexity of identifying and transplanting brain states should not be underestimated. The brain may be investigated on microscopic neurochemical levels or as a complex and widely distributed network. It is not immediately obvious which is the correct level to be acquiring and transplanting patterns that would affect changes in behavior (that is, mental states). For example, the macroscopic representations measured by EEG may be sufficient to transplant generic physiological states, such as a state associated with a specific sleep stage. However, more refined states, such as specific cognitive states, may require a higher spatial resolution to be fully captured and transplanted by non-invasive methods. Cortical signatures and representation patterns may be carefully investigated before transplanting or replicating subjective contents of cognitive processes.

[1034] It is also possible that cortical activity signals as measured by EEG do not fully capture certain brain states. For example, REM sleep can be distinguished from waking activity by the absence of muscle tone in major muscle groups as determined by electromyography (EMG). Therefore, additional physiological signals can be used to capture brain states more fully. In particular, capturing the status of the autonomic nervous system, such as, for example, the dynamic balance of sympathetic and parasympathetic activity reflected in measures such as heart rate variability, may turn out to be necessary for high-fidelity identification of brain states. These signals are also open to noninvasive modulation, such as through the stimulation of the vagus nerve.

[1035] Combining vagus nerve stimulation (VNS) with brain entrainment is promising. It has been shown that VNS significantly increased and decorrelated spontaneous activity and suppressed entrainment at 6-8 Hz (Nichols et al., 2011).

[1036] Finally, it is worth discussing the difference between online and offline effects of stimulation. The current brain stimulation literature suggests that cortical states can be effectively enhanced during stimulation, what are referred to as online effects. Transplanting brain states focuses on augmenting, restoring, and inducing brain states with stimulation, and is not dependent on a full understanding of the mechanism by which more long-lasting effects occur in brain networks after the conclusion of stimulation. Brief perturbations may be sufficient to switch to another state either with or without continued stimulation. The main mechanisms are based on the fundamental property of neuronal oscillations to respond to (weak) perturbations through entrainment.

CONCLUSION

[1037] Together these findings provide the basis for the transplant (transfer) of brain states and provide the means by which a cortical signature may be obtained via EEG or EMG associated with the desired brain state of a donor that may, in turn, be processed, inverted or deconvoluted, and subsequently applied to a recipientwho may be another person, or the same person at another time-to induce this state through cortical rhythm entrainment using preferably sensory stimuli, such as light or sound, a combination thereof, or, possibly, TES or TMS. Using cortical signatures acquired from a donor, rather than a fixed-frequency synthetic waveform stimulation as is currently typical for TES techniques, offers the distinct advantage of replicating multiphasic, multifrequency, temporally dynamic, naturalistic signals, which is more likely to modulate neuronal network activity effectively (Frohlich and McCormick, 2010) and, more important, induce naturalistic brain states due to the additional information contained in the complete spectrum of macroscale brain signals. Therefore, this technique provides a novel and effective neuromodulation approach to the noninvasive, non-pharmacological treatment of a variety of psychiatric and neurological disorders for which current treatments are mostly limited to pharmacotherapeutic interventions.