WO2020061517A1 - Electroencephalography-based technologies for repetitive transcranial magnetic stimulation - Google Patents

Abstract

A method and system for effective magnetic stimulation of electrophysiological activity achieved through phase-amplitude coupling are disclosed. Effective magnetic stimulation is achieved by identifying an excitable phase of a target oscillation in a living organism. A phase, or the fraction of the oscillation that has elapsed relative to the origin, may be excitable by its capacity to amplify the target oscillation, generate gamma wave activity, or produce other neurological responses. Upon identifying an excitable phase, a repetitive magnetic pulse is calculated with a phase that matches the excitable phase of the target oscillation.

Description

ELECTROENCEPHALOGRAPHY-BASED TECHNOLOGIES FOR REPETITIVE TRANSCRANIAL MAGNETIC STIMULATION

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This patent document claims priority to U.S. Provisional Application No. 62/734,944 entitled “REPETITIVE TRANSCRANIAL MAGNETIC STIMULATION” filed on September 21, 2018. The entire content of the above application is incorporated by reference as part of this disclosure of this patent document.

TECHNICAL FIELD

[0002] The present document generally relates to magnetic stimulation of electrophysiological oscillations.

BACKGROUND

[0003] Transcranial magnetic stimulation (“TMS”) is a non-invasive procedure that stimulates the nerve cells in the brain. Examples of TMS include a changing magnetic field applied near the cranium, causing an electric current to flow in a small taprgeted region of the brain. TMS can be applied in a series of repeated pulses to enhance electrophysiological oscillations, leading to changes in brain activity. TMS is known for providing many therapeutic applications for treating neurological and psychiatric disorders.

[0004] The stimulation frequency and the endogenous oscillatory phase at which TMS pulses are delivered play a crucial role in impacting brain activity. However, endogenous network dynamics constrain the effects of non-invasive brain stimulation. Currently, the TMS interventions do not consider intrinsic brain activity and do not adjust for stimulation parameters of the underlying brain dynamics.

[0005] Thus, an efficient means for adjusting stimulation parameters according to intrinsic brain activity is needed.

SUMMARY

[0006] A method and system for effective magnetic stimulation of electrophysiological activity achieved through phase-amplitude coupling are disclosed. In some embodiments, effective magnetic stimulation is achieved by identifying an excitable phase of a target oscillation in a living organism. A phase, or the fraction of the oscillation that has elapsed relative to the origin, may be excitable by its capacity to amplify the target oscillation, generate gamma wave activity, or produce other neurological responses. Upon identifying an excitable phase, a repetitive magnetic pulse is calculated with a phase that matches the excitable phase of the target oscillation.

[0007] The method of more effective electromagnetic brain stimulation may be used to enhance depression treatment, obsessive-compulsive disorder treatment, and addiction treatments, and other neurological and neurocognitive studies. The present inventors recognize that a fast estimation of the optimal phase also leads to a greater induction of neural plasticity and behavioral enhancements. The method allows for longer periods of increased spontaneous excitability of a certain neurological area. The method also allows for a quick estimate of the high excitability phases of a target oscillation in an individual in a specific brain location.

[0008] In one example aspect, a system for effective magnetic stimulation of

electrophysiological activity achieved through phase-amplitude coupling is disclosed. The system configured to identify, from endogenous oscillatory data, an excitable phase of a target oscillation. In response to identifying an excitable phase of a target oscillation, a repetitive magnetic pulse is calculated based on the excitable phase of the target oscillation.

[0009] In another example aspect, a method for effective magnetic stimulation of electrophysiological activity achieved through phase-amplitude coupling is disclosed. In the method, an amplitude increase in the target oscillation is induced. The inducement is the result of the identification of a first excitable phase from the first set of endogenous oscillatory data, calculating a repetitive magnetic pulse based on the first excitable phase of the target oscillation, and applying the first repetitive magnetic pulse based on the first excitable phase of the target oscillation.

[0010] In at least some embodiments, the calculation of the optimal phase considers interindividual variability, variability between brain areas, variability or differences between oscillations in distinct and overlapping frequency bands, and the existence of high-and-low excitability phase periods in each oscillatory cycle. The calculation also considers the frequency, instantaneous phase, and cortical locations of intrinsic brain oscillations.

[0011] In another example aspect, a method for magnetic stimulation includes identifying, from endogenous oscillatory data, an excitable phase of a target oscillation, calculating a repetitive magnetic pulse based on the excitable phase of the target oscillation, and sending an instruction for generating the repetitive magnetic pulse with an external electromagnetic stimulus. [0012] In another example aspect, a method for magnetic stimulation includes identifying, from a first set of endogenous oscillatory data of a living organism, a first excitable phase of a target oscillation, calculating a first repetitive magnetic pulse based on the first excitable phase of the target oscillation, applying the first repetitive magnetic pulse to the living organism, and receiving, in response to applying the first repetitive magnetic pulse to the living organism, a second set of endogenous oscillatory data from the living organism, wherein the second set of endogenous oscillatory data includes an amplitude increase in the target oscillation. The method may further include identifying, from the second set of endogenous oscillatory data, a second excitable phase of the target oscillation, calculating a second repetitive magnetic pulse based on the second excitable phase of the target oscillation, and applying the second repetitive magnetic pulse to the living organism based on the response of the target oscillation of the first repetitive magnetic pulse to the living organism.

[0013] In another example aspect, a method of a non-invasive brain stimulation includes applying, to nerve cells in a brain, pulses to induce changes in oscillatory brain activity, estimating an excitable phase of a target oscillation of the brain based on a coupling between the applied pulses and the oscillatory brain activity, and adjusting stimulation parameters at least based on the estimated excitable phase of the target oscillation.

BRIEF DESCRIPTION OF THE DRAWINGS

[0014] So that the manner in which the above-recited features of the present document can be understood in detail, a more particular description of the embodiments, briefly

summarized above, may be had by reference to implementations, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical implementations of this disclosure and are therefore not to be considered limiting of its scope, for the disclosure may admit to other equally effective implementations.

[0015] FIG. 1 illustrates a method of magnetic stimulation to an electrophysiological oscillation according to one implementation described herein.

[0016] FIG. 2 illustrates a chart demonstrating the effect of frequencies and excitability phases on an oscillation according to at least one implementation described herein.

[0017] FIG. 3 illustrates a chart demonstrating the effect of frequencies and excitability phases on gamma bursts in at least one implementation described herein. [0018] FIG. 4 illustrates a system of a scalp EEG used for modeling spectral modulations in at least one implementation described herein.

[0019] FIG. 5 shows possible scenarios of frequency and phase tuning of repetitive transcranial magnetic stimulation (rTMS) to underlying brain dynamics.

[0020] FIG. 6 shows possible mechanism of how alpha oscillations act on gating neural excitability.

[0021] FIG. 7 shows an example method for magnetic stimulation based on some embodiments of the disclosed technology.

[0022] FIG. 8 shows another example method for magnetic stimulation based on some embodiments of the disclosed technology.

[0023] FIG. 9 shows an example method of a non-invasive brain stimulation based on some embodiments of the disclosed technology.

[0024] Identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements disclosed in one implementation may be beneficially utilized in other implementations without specific recitation.

DETAILED DESCRIPTION

[0025] A method and system for effective magnetic stimulation of electrophysiological activity achieved through phase-amplitude coupling are disclosed. Effective magnetic stimulation is achieved by identifying an excitable phase of a target oscillation in a living organism. A phase, or the fraction of the oscillation that has elapsed relative to the origin, may be excitable by its capacity to amplify the target oscillation, generate gamma wave activity, or produce other neurological responses. Upon identifying an excitable phase, a repetitive magnetic pulse is calculated with a phase that matches the excitable phase of the target oscillation.

[0026] FIG. 1 illustrates a method of magnetic stimulation to an electrophysiological oscillation according to one implementation described herein.

[0027] In step 110, an excitable phase of a target oscillation is identified from endogenous oscillatory data. The endogenous oscillatory data may include frequency, phase, amplitude, type of brainwave, emission area of the cranium, neural emitter, or other electrophysiological data. Endogenous oscillatory data may include spatial, spectral, and temporal information about the underlying brain dynamics to maximize the effectiveness and consistency of repetitive magnetic stimulation.

[0028] In at least one implementation, endogenous oscillatory data is obtained from an electroencephalography (EEG). In at least one embodiment, the endogenous oscillatory data is obtained from functional magnetic resonance or positron emission tomography.

[0029] In at least one implementation, the endogenous oscillatory data is the result of a phase sweep across a sinusoidal parameter of an endogenous oscillation bandwidth. In at least one implementation, the endogenous oscillatory data includes a frequency sweep across an endogenous oscillation bandwidth.

[0030] In at least one implementation, an excitable phase is determined based upon the endogenous oscillation data. An excitable phase may also be determined based upon interindividual variability, variability between brain areas, variability or differences between oscillations in distinct and overlapping frequency bands. In at least one implementation, the excitable phase depends on the cortical location of an intrinsic brain oscillation. The excitable phase may differ over frequency bands and brain areas. In at least one implementation, the endogenous oscillatory data is the result of a phase sweep across a sinusoidal parameter of an endogenous oscillation bandwidth.

[0031] In at least one implementation, the excitable phase may be predicted using autoregressive modeling. In at least one implementation, the identifying of an excitable phase of a target oscillation is performed in real-time with the calculating of a repetitive magnetic pulse based on the excitable phase of the target oscillation. In at least one implementation, the excitable phase of the target oscillation is determined in real time with increased neural activity. The estimation of excitable phase in real time allows for quick targeting of high excitability phases of a target oscillation. The estimated excitable phase is then specifically and precisely targeted with magnetic pulses of a matching phase by an external electromagnetic stimulus. The stimulation may occur for long periods (e.g., seconds to minutes) to maintain increased spontaneous excitability of a certain brain area. In at least one implementation, the excitable phase is a negative peak value. In at least one implementation, the excitable phase is identified using data-driven models of spectral dynamics.

[0032] In at least one implementation, an excitable phase is first identified using data-driven models. Subsequently, an excitable phase and amplitude is more precisely identified in real time using phase-amplitude coupling and autoregressive models. [0033] A target oscillation may be identified in order to induce variable degrees of brain plasticity on a subject. Target oscillation may be linked to affect changes in behavior, such as depression, OCD, or an addiction. All frequency bands of endogenous oscillations may be targeted, including theta, alpha, beta, and gamma bands.

[0034] In at least one implementation, independent component analysis (ICA) is used to decompose the endogenous oscillatory data into to maximally independent component (IC) processes with a subsequent joint decomposition of IC log spectrograms. The decomposition of IC log-spectral activities can reveal contributions of independent spectral modulators to oscillatory activity in these cortical sources. Independent modulator analysis can detect spectral co-modulation between various combinations of cortical sources. Such Independent Modulator processes might derive from coordinated action of modulatory facts (e.g., thalamocortical feedback loops or brainstem-based neuromodulatory systems). Independent sources with variable peak alpha frequencies can be revealed when mixed signals at are decomposed using ICA. This allows for inter-subject variability in peak frequencies, but also a variety of peak frequencies that are exhibited by independent sources in the same brain area

[0035] In step 120, a repetitive magnetic pulse based on the excitable phase of the target oscillation is calculated. Calculation of the repetitive magnetic pulse in step 120 is enhanced to very precise entrainment and enhancement of brain oscillatory activity at a specific frequency. This may also result in consistent and effective stimulation results. In at least one implementation, calculation utilizes source estimation, frequency transform, and a priori information on underlying brain dynamics. Additional repetitive transcranial magnetic stimulation parameters may be used, including location, frequency, and phase.

[0036] In at least one implementation, the calculation of the repetitive magnetic pulse in step 120 utilizes a recursive-predictive scheme for determining whether a pulse will be delivered ahead of time. The calculation of the repetitive magnetic pulse may utilize a Kalman filter-like scheme to track the phase and amplitude of the target oscillation and then factoring in a predicted phase and amplitude evolution. Through results in the calculation of the excitable phase of the target oscillation.

[0037] In at least one implementation, the repetitive magnetic pulse has a phase corresponding to the excitable phase of the target oscillation. In at least one implementation, the repetitive magnetic pulse has a phase at or about a negative peak of the target oscillation. In one implementation, magnetic pulse coupling near a negative peak of an alpha oscillation in an underlying EEG electrode results in neural plasticity inducement. Different phases of different frequency bands and brain areas may have magnetic pulse stimulation applied to entrain the brain oscillation with the external electromagnetic stimulus.

[0038] In at least one implementation, blind source separation for detecting target frequency peaks in electrophysiological signals. Independent spectral modulators may contribute to cortical oscillatory activity for precisely identifying the target oscillation.

[0039] In at least one implementation, transcranial alternating current stimulation (tACS) and/or transcranial direct current stimulation (tDCS) is applied to cause phase-coupling (e.g. entrainment) between oscillatory brain activity and the external electromagnetic stimulus. Direct evidence for entrainment of neural oscillations may come from transcranial alternating current stimulation (tACS). In at least one implementation, a weak electric field enhances endogenous oscillations when stimulation frequencies match the endogenous oscillation. In at least one implementation, the oscillatory phase of a TMS pulse induces neural plasticity in a brain oscillation.

[0040] In at least one implementation, the phase matching at slower rhythms with endogenous oscillations causes amplitude modulation of high frequency endogenous oscillations known as phase amplitude coupling.

[0041] An important aspect of the magnetic stimulation is that the effects of every single pulse or train can summate with repeated application leading to long-lasting neuromodulatory effects. The magnetic stimulations pulses can be tuned or adjusted to the optimal phase of endogenous oscillations by using a novel method employing phase-amplitude coupling.

[0042] FIG. 2 illustrates a chart demonstrating the effect of frequencies and excitability phases on an oscillation according to at least one implementation described herein.

[0043] The vertical lines represent single magnetic stimulation pulses applied to the target oscillation. The sine waves represent the target oscillations. The stars indicate time points at which the magnetic stimulation pulse hits the endogenous oscillation.

[0044] The repetitive TMS delivered either at the wrong frequency, which results in stimulation at random phases of ongoing oscillations, or at low excitability phases of the underlying oscillation does not result in entrainment while frequency tuned repetitive TMS triggered at phases of the endogenous oscillation results in entrainment and enhancement of the endogenous oscillations. The electroencephalography (EEG) oscillations are hypothesized to reflect cyclical variation in the excitability of neuronal ensembles. Delivering the TMS pulse at high excitability phases of a target cortical oscillation makes these underlying neuronal ensembles more likely to fire, inducing long term potentiation. [0045] As shown in the top-left figure, repetitive magnetic stimulation delivered at either the wrong frequency does not result in entrainment of the target oscillation. As shown in the top- right figure, repetitive magnetic stimulation delivered at low excitability phases does not result in entrainment of the target oscillation.

[0046] As shown in the bottom-left figure, frequency -tuned repetitive magnetic stimulation triggered at high excitability phases of the endogenous oscillation results in entrainment and enhancement of the target oscillation. As shown in the bottom-left figure, frequency -tuned repetitive magnetic stimulation triggered at high excitability phases of the endogenous oscillation results in entrainment and enhancement of the target oscillation.

[0047] Transcranial magnetic pulse is considered a form of magnetic pulsing. Transcranial magnetic stimulation (TMS) of the brain is currently applied in clinical treatments. Clinical treatments with repetitive transcranial magnetic stimulation (rTMS) and experimental research findings show mixed effects, with rTMS protocols inducing variable degrees of brain plasticity over subjects and sessions. The repetitive transcranial magnetic stimulation method implemented based on some embodiments of the disclosed technology can use spatial and temporal information about the underlying brain dynamics to maximize the effectiveness and consistency of rTMS.

[0048] The repetitive transcranial magnetic stimulation method implemented based on some embodiments of the disclosed technology takes into account 1) interindividual variability 2) variability between brain areas 3) variability or differences between oscillations in distinct and overlapping frequency bands, 4) existence of high- and low-excitability phase periods in each oscillatory cycle. The unique aspect of this innovation is that we are taking into account not only the frequency of intrinsic brain oscillations, but also their instantaneous phase and cortical locations. We deliver repetitive TMS stimulation in specific phases of the ongoing intrinsic brain activity indicated by our unpublished pilot research data.

[0049] The embodiments of the disclosed technology, taking into account these factors, allow for a very precise and constant entrainment and enhancement of brain oscillatory activity in a specific frequency and ensures consistent and more effective stimulation results.

[0050] Some embodiments of the disclosed technology utilize EEG to map spatiotemporal patterns of functional reorganization induced by rTMS using an online approach, and following stimulation. EEG recording also allows measurement of longer term changes in brain activity following rTMS treatment. [0051] Non-invasive transcranial brain stimulation is widely used in both basic science and clinical medical applications. Common non-invasive brain stimulation techniques include transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), and transcranial magnetic stimulation (TMS).

[0052] In the transcranial electrical stimulation techniques (e.g., tDCS or tACS), two electrodes are placed on the scalp. Then, either direct current (DC) is made to flow from anode to cathode or an alternating current (AC) is passed between the two electrodes.

Because these non-magnetic techniques produce higher levels of electric current in the scalp and superficial tissues than compared to the actual stimulation of brain tissue, stimulation occurs mostly in superficial layers of the cortex at a low intensity.

[0053] In TMS, high intensity electromagnetic pulses are produced in one or more wire coils (transducers) placed tangential to the scalp. These pulses induce electrical currents both in the underlying brain areas near the coil, but also to some extent stimulate deeper brain regions depending on the intensity of stimulation. Since magnetic fields are not affected by volume conduction and the electric currents are generated in the underlying cortical area,

TMS allows for the generation of much stronger and deeper stimulation, while sparing the scalp and superficial structures, as compared to the non-magnetic stimulation techniques of tDCS and tACS. TMS delivered magnetic fields are roughly of the same magnitude as those produced by magnetic resonance scanners with a 1 Tesla or higher rating. TMS may be used to target and stimulate specific brain regions (e.g., motor cortex), and may alter neuronal excitability for some time following stimulation.

[0054] TMS can be applied as single, isolated pulses (single pulse TMS), as trains of stimuli delivered at a fixed frequency (repetitive TMS or“rTMS,” typically at repetition rates of 1 Hz-20 Hz), or as more complex trains of stimulation that combine different repetition frequencies. An important aspect of TMS is that the effects of each single pulse or pulse train can accumulate with repeated applications, leading to effects on the brain outlasting the period of stimulation.

[0055] The effects of rTMS, in particular on brain metabolic activity, can be observed near the cortical site of stimulation as well as at remote, but anatomically and functionally connected cortical and subcortical areas. For example, TMS applied to the left dorsolateral prefrontal cortex (DLPFC) modulates dopamine release and blood oxygen level dependent (BOLD) activity in the dorsal striatum due to the functional and the structural connections between these two structures. It follows that rTMS may not just stimulate brain regions underlying the transducer but may also affect entire neuronal circuits. These secondary effects can be both rapid and long term.

[0056] The behavioral enhancements produced by high-frequency, repetitive TMS is mediated by affecting endogenous oscillations in cortical field potentials. The effect is mediated by phase-coupling (so called entrainment) between oscillatory brain activity and the external electromagnetic stimulus. The entrainment of neural oscillations may be achieved using transcranial alternating current stimulation (tACS) or transcranial direct current stimulation (tDCS) or both. The phase of the alpha oscillation in an underlying EEG electrode in real time shows that stimulating at the negative peak of the alpha cycle neural plasticity can be induced while stimulating (using rTMS) at the positive peak of alpha there is no effect. However the phases of oscillations at which TMS pulses are efficient may differ over frequency bands and brain areas. In some embodiments of the disclosed technology, (1) source estimation, (2) frequency transform and (3) a priori information on underlying brain dynamics on which oscillatory phase to target may be used in phase-guided rTMS stimulation. This information is then used to tune rTMS parameters including location, frequency, and phase for more effective rTMS stimulation.

[0057] In some implementations, TMS can be applied as trains of stimuli delivered at a fixed frequency (repetitive TMS or‘rTMS’), typically repetition rates are within 1-20 Hz. An important aspect of TMS is that the effects of each single pulse or train can summate with repeated application leading to long-lasting neuromodulatory effects of rTMS which are of great interest for therapeutic applications in neurological and psychiatric disorders.

[0058] The method implemented based on some embodiments of the disclosed technology includes tuning rTMS pulses to the optimal phase of endogenous oscillations by employing phase amplitude coupling. In one example, phase and amplitude of the brain signal of interest will be predicted using autoregressive modeling. The method estimates in real time which phases in a target oscillatory cycle are related to increased neural population activity in the underlying brain area. This allows to detect and target high excitability phases of a target oscillation and specifically and precisely target those phases by delivering TMS pulses at these instants. This will increase the neural response relative to single TMS pulses and enhance the effect of rTMS in terms of neural plasticity. The method implemented based on some embodiments of the disclosed technology uses spatial, temporal and spectral information on the underlying brain dynamics to maximize the effectiveness and consistency of rTMS interventions, allowing for more precise and effective transcranial magnetic stimulation (TMS) of the brain, than is currently applied in clinical treatments.

[0059] The method implemented based on some embodiments of the disclosed technology can be implemented to better target functional oscillatory networks in the brain with rTMS, by using more flexible data-driven models of spectral dynamics. The method can allow fast estimation of high excitability phases of a target oscillation in a specific brain area both before and in real time during rTMS stimulation. This will be achieved by using methods of phase amplitude coupling and autoregressive models. The method is then able to stimulate the specific brain area or a connected brain area at those high excitability phases estimated in real time from the EEG signal. The method will also allow to determine longer periods (seconds to minutes) of increased spontaneous excitability of a certain brain area, and thus allow to deliver trains of rTMS during these periods.

[0060] The method implemented based on some embodiments of the disclosed technology can quickly estimate (e.g., in about 5 min prior to starting the rTMS treatment or within a few milliseconds during the treatment) which are the high excitability phases of a target oscillation in an individual in a specific brain location. The method can then target those phases during stimulation.

[0061] The method implemented based on some embodiments of the disclosed technology can identify the phase of the underlying brain oscillation at which to deliver the TMS pulse by identifying the high-excitability phases of the target oscillation. It is assumed that the phase of spontaneous low-frequency EEG oscillations controls the excitability of local cortical neuronal ensembles making them more likely to fire. This results in a systematic enhancement of responses to events occurring during high excitability phases (concurrent with gamma frequency bursts in the EEG) and suppression of responses to events that occur during low excitability phases.

[0062] The gamma (30-200Hz) oscillatory power increase in EEG reflects state of high neuronal excitability and is often coupled to the phase of lower frequency oscillations. The amplitude modulation of high frequency oscillation by the phase of slower rhythms is a phenomena known as phase amplitude coupling (PAC). One interpretation of this process is made by linking the high frequency oscillation to the activity of local cortical ensembles while slower phase oscillations are associated to larger brain areas. Given this, PAC can be used to identify high-excitability phases of the target oscillation in the TMS context. [0063] To use PAC to this end, the coupling estimation method is embedded in a recursive- predictive scheme, so time is allowed to determine if a pulse will be delivered or not ahead of time. The method implemented based on some embodiments of the disclosed technology can track phase and amplitude of endogenous oscillations of interest by using a Kalman filter like scheme and then computing a PAC measure using the predicted phase and amplitude evolution. The method implemented based on some embodiments of the disclosed technology can predict an increase in cortical excitability through PAC. Through this estimation process, the target phase may be predicted as well.

[0064] The method implemented based on some embodiments of the disclosed technology allows delivery of more effective and consistent electromagnetic brain stimulation of a number of cortical brain regions and can be used to enhance the clinical effects of existing methods of rTMS -supported depression treatment, or obsessive compulsive disorder treatment, and treatments for addictions and many other conditions. The disclosed technology can also be used for producing more reliable stimulation effects in basic neurocognitive studies. One of the advantages of the disclosed technology is the fast estimation of the optimal phase for stimulation, which makes the approach suitable for integration into clinical therapies. Another advantage is that the disclosed technology can estimate high excitability phases during the treatment in real time and also estimate longer periods of spontaneous fluctuations in excitability which will increase efficacy of rTMS stimulation even further.

[0065] Short-term and long-term neural effects produced by repetitive TMS may be mediated by the interactions with endogenous brain oscillations. The stimulation frequency plays thereby a crucial role with weak electric fields enhancing endogenous oscillations only when stimulation frequencies match the endogenous oscillation. There is thus a constraint on the effect of non-invasive brain stimulation imposed by endogenous network dynamics.

[0066] The method implemented based on some embodiments of the disclosed technology includes tuning rTMS stimulation frequency to endogenous oscillation frequency by using blind source separation to detect target frequency peaks in electrophysiological signals. The method detects independent spectral modulators contributing to cortical oscillatory activity and allows this way to specifically and precisely target distinct underlying oscillatory networks with rTMS.

[0067] The method implemented based on some embodiments of the disclosed technology may use alpha oscillations (8-l2Hz), which allow to characterize distinct, alpha frequency modulations, to determine the inter-relationship of observed alpha modulations in brain source electroencephalography (EEG) activities. The method implemented based on some embodiments of the disclosed technology includes linearly decomposing the EEG data using independent component analysis (ICA) into maximally independent component (IC) processes with a subsequent joint decomposition of IC log spectrograms. Mean spectra of cortical EEG sources are composed of various modes of oscillatory activity. The

decomposition of IC log-spectral activities can reveal contributions of independent spectral modulators to oscillatory activity in these cortical sources. Independent modulator analysis can detect spectral comodulation between various combinations of cortical sources. Such ‘independent modulator’ processes might derive from coordinated actions of modulatory factors, for example thalamocortical feedback loops or brainstem-based neuromodulatory systems. Identifying these independent modulator processes is crucial to identify and efficiently target distinct oscillatory brain networks with rTMS.

[0068] The method implemented based on some embodiments of the disclosed technology may apply to all frequency bands of endogenous oscillations such as theta, alpha, beta and gamma bands.

[0069] Cortical alpha oscillations have a typical power peak between 8 and 12 Hz which has been repeatedly discussed as a heritable and stable neurophysiological“trait” marker reflecting anatomical properties of the brain, and individuals’ cognitive capacity. However, the alpha peak frequency is highly volatile at shorter time scales, dependent on the individuals’ cognitive state. Associations between cognitive performance and endogenous modulations of oscillatory neuronal activity in the individual alpha frequency range suggest that upper and lower alpha band power are separately regulated during certain cognitive processes.

[0070] Some embodiments of the disclosed technology take into account that the mean peak alpha frequency across scalp electrodes represents an average of widely varying frequency peaks of independent EEG processes. This alpha peak variability could be explained by at least two different processes. First, multiple EEG sources projecting to the same channel with comparable strengths may express unique frequency characteristics that differ from one another by 1 Hz or more. Else, each EEG source may produce for example alpha activity at varying frequencies, possibly for specific cognitive or behavioral purposes.

[0071] The method implemented based on some embodiments of the disclosed technology linearly may decompose the EEG data using independent component analysis (ICA) into maximally independent component (IC) processes with a subsequent joint decomposition of IC log spectrograms. When mixed signals at scalp electrodes are decomposed using independent component analysis (ICA), independent sources with variable peak alpha frequencies can be revealed. The method implemented based on some embodiments of the disclosed technology reveals and identifies not only considerable inter-subject variability in peak frequencies, but also a variety of peak frequencies that are exhibited by independent sources in the same brain area.

[0072] As shown in FIG. 2, repetitive TMS delivered at the wrong frequency which results in stimulation at random phases of ongoing oscillations (top), does not result in entrainment while frequency tuned repetitive TMS triggered at the frequency. The mean peak alpha frequency across scalp electrodes represents an average of widely varying frequency peaks of independent EEG processes. Analysis of scalp EEG data is complicated by the fact that activity recorded at each scalp channel sums activities from several cortical source areas separated as widely as occipital and frontal cortex. To accurately model the relation of EEG dynamics to behavior and experience, therefore, activities of distinct brain and non-brain EEG sources must first be isolated.

[0073] The method implemented based on some embodiments of the disclosed technology includes identifying a variety of peak frequencies which are commonly exhibited by independent sources in the same brain area. The method implemented based on some embodiments of the disclosed technology includes linearly decomposing the EEG data using independent component analysis (ICA) into maximally independent component (IC) processes with a subsequent joint decomposition of IC log spectrograms. Independent component analysis (ICA) decomposes scalp-recorded EEG data into a weighted set of maximally temporally independent component (IC) processes by learning spatial filters that maximize the temporal independence of the resulting IC-filtered output time series. Scalp projections of ICs each strongly resemble the projection of a single equivalent dipole source, a result compatible with generation by partially synchronous local field activity across a cortical patch.

[0074] FIG. 3 illustrates a chart demonstrating the effect of frequencies and excitability phases on gamma bursts in at least one implementation described herein.

[0075] Magnetic stimulation pulses are represented by vertical lines in the upper part of the figure. The upper horizontal line of the figure indicates a magnetic stimulation pulse at a positive peak with a vertical tick. The lower horizontal line of the figure indicates a magnetic stimulation pulse at a negative peak with a vertical tick. [0076] On the left panel, the cycle is asymmetric, and the maxima phase is more apt to modulate than the minima phase. In this case of weak signals, gamma bursts can occur at any time during the wave cycle. On the right panel, the signal is symmetrically fluctuating around 0 uV, and stimulation at either the cycle maxima or minima are similarly effective. In this case, gamma bursts occur at minima only when the oscillation reaches a threshold amplitude. The bursts of gamma in each alpha cycle facilitate neuronal processing. If magnetic stimulation pulses arrive during the high-excitability minimum phase of the cycle, gamma bursts result simultaneously with the pulses. On the other hand, gamma bursts do not result if magnetic stimulation pulses instead arrive during low excitability phases at wave peaks.

[0077] As will be discussed in more detail below,

[0078] FIG. 4 illustrates a system of a scalp EEG used for modeling spectral modulations in at least one implementation described herein.

[0079] ICA, applied to EEG data recorded at a large number of scalp electrodes, identifies (A) temporally distinct independent) signals generated by partial synchronization of local field potentials within cortical patches (B), the resulting far-field potentials summed (å), in differing linear combinations, at each electrode depending on the distance and orientation of each cortical patch generator relative to the (A) recording and (C) reference electrodes. On average, power in the cortical IC signals decrease monotonically with frequency, but also exhibit continual, marked, and complex variations across time. Rather than viewing these variations as occurring independently at each frequency, spectral modulations may be modeled as exponentially weighted influences of several distinct but possibly overlapping modulator (IM) processes (D) that independently modulate via multiplicatively scaling (P) the activity spectra of one or more independent component (IC) signals. On converting the IC spectra to log power, combined IM influences on IC spectra are converted to log-linear weighted sums of IM influences, allowing a linear ICA decomposition of the IC log-power spectra to separate the effects of the individual IM processes (D) on power at selected frequencies of IC sources (B).

[0080] As shown in FIG. 4, reference and scalp electrodes may be placed around the periphery of the electrophysiological signal region (e.g., scalp). Under these electrodes, various cortex and independent modulators are situated. The system allows for Independent Component Analysis (ICA) to identify temporally independent signals generated by partial synchronization of local field potentials within cortical patches. The system then sums the resulting far-field potentials, in differing linear combinations, at each electrode depending on the distance and orientation of each cortical patch generator relative to the recording and reference electrodes.

[0081] The system is advantageous because spectral modulations may be modeled as exponentially weighted influences of several distinct rather than viewing these variations as occurring independently at each frequency. The modulator processes independently modulate via multiplicatively scaling the activity spectra of one or more independent component (IC) signals. Upon converting the IC spectra to log power, combined IM influences on IC spectra are converted to log-linear weighted sums of IM influences, allowing a linear ICA decomposition of the IC log-power spectra to separate the effects of the individual IM processes on power at selected frequencies of IC sources.

[0082] In at least one implementation, a system may identify peak frequencies which are commonly exhibited by independent sources in the same brain area. The method linearly decomposes the EEG data using independent component analysis (ICA) into maximally independent component (IC) processes by learning spatial filters that maximize the temporal independent of the resulting IC-filtered output time series. Scalp projections of ICs each strongly resemble the projection of a single equivalent dipole source, a result compatible with generation by partially synchronous local field activity across a cortical patch.

[0083] Mean spectra of independent component process or cortical EEG sources are composed of various modes of oscillatory activity. Log-spectral fluctuations of multiple ICs are decomposed to into a product of distinct spectral modulator processes. The decomposition of IC log-spectral activities can reveal contributions of independent spectral modulators to oscillatory activity in these cortical sources. Such modulator processes might derive from coordinated actions of modulatory factors (e.g., thalamocortical feedback loops or brainstem- based neuromodulatory systems).

[0084] The method implemented based on some embodiments of the disclosed technology includes utilizing external stimulation parameters. Repetitive TMS (rTMS) is characterized by the ability to manipulate a wide array of simulation parameters which function in concert to produce cognitive, behavioral, and emotional changes. The relationships between these parameters and their biologic effects are proving to be quite complex. Simulation parameters are summarized in Table 1.

[0085] Table 1 : External Stimulation parameters and examples of clinical TMS Paradigms

[0086] A very wide range of TMS parameters are possible. The method implemented basec on some embodiments of the disclosed technology focuses on, among others, frequency and intensity, as well as the stimulation parameters that are guided by brain activity such as the frequency and phase of underlying cortical oscillations.

[0087] Excitatory and inhibitory effect of TMS pulses depends on the frequency of stimulation. Low frequency (e.g. 1 Hz) TMS induces a transient suppression of excitability in the affected cortical region(s), while high frequency rTMS (>5 Hz) transiently enhances local cortical excitability.

[0088] As opposed to low frequency rTMS, high frequency rTMS (>5 Hz) enhances cortical activity and has been shown to improve performance in certain cognitive tasks and has longer lasting effects in plasticity inducing protocols. Stimulation in the alpha band (8 Hz 12 Hz) is of special interest here since most rTMS protocols in the clinical arena involve some form of 10 Hz stimulation. For example, 10 Hz rTMS over frontal brain areas has proven therapeutic benefits in persons with treatment resistant depression.

[0089] The high frequency stimulation may be more effective when the frequency of stimulation is matched or aligned to the patient’s native frequency. Individual patients or brain circuits have specific frequencies of underlying brain oscillations at which their neuronal networks will optimally function and that rTMS at these individualized frequencies may produce improved results compared to conventional rTMS where all patients receive the same stimulation frequency. [0090] The method implemented based on some embodiments of the disclosed technology includes guiding stimulation parameters using underlying brain activity. The behavioral enhancement produced by high frequency rTMS is mediated by affecting endogenous oscillations in cortical field potentials. Cortical oscillations have known functional behavioral correlates. The rTMS tuned to a specific frequency band for example the alpha band (8 Hz 12 Hz) enhances cortical oscillations in this frequency band and produces associated changes in behavioral performance. It is assumed that the cortical effect of oscillatory enhancement is mediated by phase coupling (entrainment) between oscillatory brain activity and the external electromagnetic stimulus. This effect forms the theoretical basis for frequency tuning of rTMS to underlying brain oscillations. The tACS (and/or tDCS) stimulation may reliably entrain endogenous neural oscillations in the alpha band, for example, with low stimulation intensities.

[0091] The disclosed method may be implemented to achieve frequency tuning of rTMS to endogenous brain oscillations. The rTMS at subject specific alpha frequencies (8 Hz 12 Hz) increases behavioral performance in visual tasks and increases offline EEG power spectra in the alpha band. The alpha power increase after individually tuned alpha frequency rTMS is related to a significant improvement in performance of a mental rotation task. The rTMS evoked EEG oscillation outlasts stimulation offset, which suggests that indeed an oscillation has been entrained. As such, rTMS interacts selectively with the target oscillations and associated function by synchronization.

[0092] In one example, matching stimulation to endogenous neural activity is crucial for electric fields to have an effect on network dynamics since the depolarization caused by the weak electric field is too small to activate neurons at rest. In essence, neurons need to be close to their firing threshold for the resulting sub millivolt perturbation of the membrane voltage to be an effective modulator of endogenous network activity. A key mechanism of tACS (and/or tDCS) and rTMS may be enhancing, but not overriding, intrinsic network dynamics, adding to a system in resonance.

[0093] In one example, stimulation may be depending on the phase of underlying oscillations. The variable results of TMS may be related to the underlying dynamics of the brain state. Contradicting effects in plasticity evoked rTMS protocols may be due to individual differences in the recruitment of cortical intemeuron networks. The effect of excitatory and inhibitory theta burst stimulation protocols are highly correlated with the latency of motor evoked potentials, induced with TMS pulses. [0094] FIG. 5 shows possible scenarios of frequency and phase tuning of rTMS to underlying brain dynamics. Vertical lines represent single TMS pulses, sine waves represent underlying brain oscillations. The stars indicate time points at which the TMS pulse hits the endogenous oscillation. As shown in FIG. 5, from top to bottom: 1) low intensity rTMS not matching the frequency of endogenous underlying oscillations results in stimulation at random phase of the underlying oscillation and fails to entrain brain oscillations 2) Low intensity frequency tuned rTMS delivered at low excitability phases of the underlying oscillation does not result in entrainment 3) frequency tuned rTMS triggered at high excitability phases of the underlying oscillation results in entrainment and enhancement of the underlying oscillations. 4) High intensity rTMS delivered at non matching frequency to endogenous oscillations overrides underlying brain activity and results in entrainment at the frequency of stimulation.

[0095] In some embodiments of the disclosed technology, specific mu-phase-guided stimulation can open a window to plasticity via the functional mechanisms of coupling between different oscillatory processes in the brain.

[0096] The method implemented based on some embodiments of the disclosed technology can include alpha and mu phase guided stimulation. Alpha oscillations (8 Hz- 12 Hz) play a significant role in modulating input information to the brain by inhibiting visual and other neural processing. Alpha oscillations may act as a pulsed-inhibition of neural processing by constraining the timing of neural firing, leading to periodic suppression of neural processing on every cycle. Specifically, the brain exploits the alpha rhythm to appropriate time microstates of excitation and inhibition so as to be optimally ready to process or inhibit incoming information. The inhibition indexed by alpha thereby varies according to the peaks and troughs in the alpha cycle.

[0097] The phase of pre-stimulus alpha oscillations modulates visual detection. The alpha oscillatory cycle reflects changes in the ongoing excitability of the underlying cortical areas.

[0098] The same relationship seems to hold also for ongoing mu oscillations in the alpha band (8 Hz - 12 Hz) in the sensorimotor cortex. Mu oscillations constitute the most prominent rhythm in the frequency spectrum of sensorimotor cortex at rest. EEG spectral power in the mu band decreases over sensorimotor areas during motor preparation and movement, when compared to a rest (non-movement) condition.

[0099] Sensorimotor mu oscillations exercise a strong inhibitory influence on both neuronal spike timing and firing rate. Furthermore, neuronal firing rates increase with a decrease in alpha-power and predicted better discrimination performance in a discrimination task. [00100] Mu oscillations are recorded over somatosensory cortex with EEG predict successful tactile perception. The ability and failure to perceive an upcoming tactile stimulus may be predicted by the mu phase angle concentration at stimulus onset. Sensorimotor mu and alpha rhythms wield a strong inhibitory control on tactile perception, similar to what has been shown in the visual system. The timing of conscious perception is hereby regulated through brain activity states controlled by these oscillations. Thus, pulsed inhibition by alpha oscillations plays an important functional role in the extended motor, sensory, and visual system.

[00101] The method implemented based on some embodiments of the disclosed technology can utilize the functional mechanisms by which mu or alpha phase is changing the excitability of underlying cortical areas/states. EEG oscillations are hypothesized to reflect cyclical variation in the excitability of neuronal ensembles. It is hereby assumed that the phase of these spontaneous low-frequency oscillations control the excitability of local cortical neuronal ensembles making them more likely to fire.

[00102] This results in a systematic enhancement of responses to events occurring during high excitability phases (concurrent with oscillatory gamma frequency (30 Hz-200 Hz) bursts in the EEG) and suppression of responses to events that occur during low excitability phases.

[00103] In one example, gamma (30 Hz-200 Hz) oscillatory synchrony reflects a state of high neuronal excitability in the brain and is often coupled to the phase of lower frequency oscillations. In one example, gamma amplitude in the EEG is modulated by alpha phase.

[00104] The alpha cycle supposedly acts here as pulsed inhibition - gamma bursts can only occur when the alpha signal is sufficiently weak (i.e. at the troughs). There are two possible ways alpha oscillations may be acting on gating gamma bursts, as shown in FIG. 3:

[00105] 1) The alpha signal is asymmetric (i.e. always fluctuating above zero) and peaks are modulated stronger than are the troughs. In this case, when the amplitude of alpha oscillations is low enough, gamma bursts occur during the whole alpha cycle. This notion is consistent with alpha activity reflecting functional inhibition.

[00106] 2) The alpha signal is symmetrically fluctuating around zero (peaks and troughs are modulated similarly). In this scenario, gamma bursts occur at troughs only when the alpha oscillatory cycle reaches an amplitude below zero. The bursts of gamma at each alpha cycle through represent windows of neuronal processing

[00107] In one example, connecting this neurophysiological model with phase-guided TMS stimulation, the mechanism by which EEG mu phase-state-dependent rTMS stimulation works may be related to mechanisms of phase amplitude coupling (PAC) in the brain. Delivering rTMS pulses at mu negative peaks can increase firing of neurons and increase gamma synchrony during these high excitability phases. Delivery of rTMS at mu negative peaks might enhance the functional mechanism of phase amplitude coupling.

[00108] In one example, PAC reflects the means through which multiple overlapping brain networks can communicate by biasing the extracellular membrane potential in local cortical regions, allowing coordinated firing of neurons over brain areas. It is likely that rTMS stimulation can influence these network properties in a similar way by biasing and enhancing certain nodes in the network

[00109] Similar to the idea of alpha frequency phase-guided stimulation through enhancement of high excitability states during alpha troughs, theta burst stimulation (TBS) has its origins in the neurophysiological findings that rat hippocampal cells fire in bursts of theta frequency. Furthermore, the fact that human EEG theta frequency increases during learning theoretically supports a plasticity-inducing effect of TBS.

[00110] The method implemented based on some embodiments of the disclosed technology may improve TBS protocols by adapting stimulation parameters to participant’s brain states. In one example, the TBS may be guided in a more adaptive manner by delivering triplets of pulses at 50 Hz at specific phases of underlying oscillations and at frequencies other than 5 Hz.

[00111] In one example, variability (LTP versus LTD) in plasticity-inducing protocols is due to the interaction between provided stimulation and intrinsic neural network dynamics. It is crucial to take brain state-dependent effects, such as alpha phase, into account at the time of the TMS pulse. Controlling for such factors may significantly reduce the variability of cognitive and behavioral results of TMS, both in basic research and clinical practice. Another important factor to consider is the variability of oscillatory frequencies in the brain, both between subjects and also within the same subject over the scope of time. The alpha peak frequency is volatile on short time scales, may depend on cognitive or emotional state, and can differ between cortical source areas. The method implemented based on some embodiments of the disclosed technology may include identifying and targeting distinct alpha oscillatory brain networks for stimulation based on the characteristics of alpha band dynamics.

[00112] The mu phase-guided rTMS modulation of corticospinal excitability may be altered as a function of stimulation intensity. The effects of high stimulation intensity are less influenced by mu rhythm phases due to saturation of motor evoked potentials. This means that if stimulation intensity is sufficiently above the motor threshold, any ongoing endogenous fluctuations influence behavioral outcomes to a lesser degree. Thus, both individual alpha frequency and phase are increasingly relevant when stimulation intensity is low. On the other hand, high intensity stimulation is strong enough to override endogenous brain dynamics and TMS alignment with the intrinsic phase or frequency may not be as important during the stimulation with high intensity to evoke visible effect in EEG.

[00113] Another factor to consider is that low-intensity stimulation greatly decreases the risk and discomfort associated with rTMS. Low-intensity rTMS may allow less expensive and more portable TMS devices to be developed, which would make rTMS treatments more readily available for a larger number of persons. Underlying brain activity in the selection of stimulation parameters has to be taken into account, especially with respect to lower stimulation intensities.

[00114] In some implementations, mu-phase-guided rTMS and/or TBS may work by enhancing underlying coupling of mu and high-frequency gamma oscillations. The rTMS and tACS may be enhancing, but not overriding, intrinsic neural network dynamics.

[00115] In some implementations, electroencephalography (EEG) and transcranial magnetic stimulation (TMS) may enhance neuroplastic effects of TMS by adapting stimulation to underlying brain states. A method for electromagnetic stimulation implemented based on some embodiments of the disclosed technology may utilize interactions between endogenous oscillatory brain dynamics and externally induced electromagnetic field activity.

[00116] The method implemented based on some embodiments of the disclosed technology can be based on alpha band (8-12 Hz) activities because of the wide application and therapeutic effectiveness of rhythmic TMS (rTMS) using a stimulus repetition frequency at or near 10 Hz. In one example, alpha oscillatory cycles produce periodic inhibition or excitation of neuronal processing through phase-amplitude coupling (PAC) of low-frequency oscillations with high-frequency broadband (or gamma) bursting. Such alpha-gamma coupling may reflect excitability of neuronal ensembles underlying neuroplasticity effects of TMS. The method implemented based on some embodiments of the disclosed technology may use TMS delivery with simultaneous EEG recording and near real-time estimation of source-resolved alpha-gamma PAC to select the precise timing of TMS pulse deliveries so as to enhance the neuroplastic effects of TMS therapies.

[00117] Understanding how cortical excitability is affected by endogenous local field potentials can be crucial to further development and optimization of TMS stimulation protocols. Oscillations in local cortical field potentials can reflect and induce cyclical variation in the excitability of involved cortical neuronal ensembles, making them more likely to fire in one phase of the cycle than in another. Targeting oscillations in the (8-12 Hz) alpha frequency band is of special interest, as most current clinical TMS protocols involve some form of stimulation in this frequency range. For example, 10-Hz rTMS over frontal brain areas has proven to have therapeutic benefit in treatment-resistant depression.

[00118] In one example, 8-12 Hz posterior alpha and sensorimotor (mu) oscillations play a significant role in modulating brain information processing in humans by providing a periodic inhibitory influence within their generator regions. Mu rhythms exercise strong inhibitory influence on local neuronal spike timing firing rate. A rhythmic relation between mu-rhythm oscillations in monkey sensorimotor cortex and neuronal spiking, with neuronal firing highest at the (surface-negative) trough of the mu-rhythm cycle demonstrates that humans’ ability to perceive a weak tactile stimulus is predicted by the mu phase angle at stimulus onset in the EEG, suggesting that sensorimotor mu rhythms wield a strong inhibitory control on tactile perception.

[00119] In one example, electrocorticography (ECoG) data with respect to alpha oscillations in the visual cortex show reaction times are faster when local auditory and visual cortical theta/low alpha rhythms (5-8 Hz) are both in phase with the onset of an audiovisual stimulus. In one example, both phase and power of pre-stimulus alpha oscillations affect visual detection. Visual discrimination ability decreases with an increase in pre-stimulus alpha power while detection performance for attended stimuli fluctuates in time with the pre stimulus phase of spontaneous alpha oscillations. This phasic modulation of detection performance increases with stronger alpha entrainment to a rhythmic stimulus presentation.

In some embodiments of the disclosed technology, the phase of EEG alpha rhythm over posterior brain regions can reliably predict both stimulus-elicited cortical activation levels and subsequent visual detection. As well, blood oxygenation-level-dependent (BOLD) responses to brief fixation events have also been shown to vary as a function of the alpha phase of EEG independent component effective source processes.

[00120] In one example, alpha oscillations influence the temporal resolution of perception. Two briefly presented visual stimuli may be perceived as a single stimulus or as two separate stimuli depending on whether they fall in one or two separate alpha cycles depending on the frequency of the alpha oscillation. These and related findings have led to the conclusion that the frequency of the alpha cycle indexes the duration of“perceptual windows” (e.g., during the surface-negative phase of the alpha cycle), and controls variation in both the sensitivity and temporal resolution of visual perception. [00121] In some implementations, mu and alpha rhythm cycles constrain neural spikes into occurring during brief time windows, leading to periodic suppression of neural processing with cortical surface negative and positive peaks in the mu/alpha cycle representing high and low excitability states respectively.

[00122] The method implemented based on some embodiments of the disclosed technology can utilize the functional mechanisms by which oscillatory phase changes the excitability of the local cortical area and state. In one example, nested hierarchical cross-frequency PAC of cortical potentials, wherein phase in lower frequency bands modulates amplitude in respectively higher bands, is a general mechanism supporting the encoding, storage, and retrieval of information in neural networks. Slow oscillations consist of alternating states of synchronized depolarization (up-state) and hyperpolarization (down-state) that propagate throughout the cortex, also reaching the thalamus via cortico-thalamic projections. In one example, cortico-thalamic feedback may play a key role in the temporal control of cortical excitability by mediating phase alignment of neuronal firing and slow oscillatory peak depolarization.

[00123] In an example of theta-gamma PAC in the hippocampus and cortex during working memory, information encoding, and retrieval that is linked to theta phase-dependent processes of synaptic potentiation and depotentiation, the phase of these spontaneous low- frequency oscillations may control the excitability of local cortical neuronal ensembles, making them more likely to fire. This results in a systematic enhancement of responses to events occurring during high-excitability phases concurrent with broadband (30-200 Hz) gamma oscillatory bursts in cortical recordings, and suppression of responses to events occurring during low-excitability phases activity has been suggested to reflect and index local neuronal population activity indicating a state of high neuronal excitability.

[00124] FIG. 6 shows possible mechanism of how alpha oscillations act on gating neural excitability. First periods 610 indicate periods of inhibition (alpha positive peak), while second periods 620 indicates periods of activation (alpha trough). Transcranial magnetic stimulation (TMS) pulses are represented by vertical lines in the upper part of the figure. The bursts of gamma at each alpha cycle trough represent windows of neuronal processing. Left panel indicates that TMS pulses are delivered at alpha troughs. They thus coincide with gamma bursts that are coupled to alpha troughs, enhancing them. If TMS pulses arrive during high excitability phases at alpha troughs (blue colored periods), they occur simultaneously with gamma bursts and are able to enhance local brain processing. Right panel indicates that TMS pulses are delivered at alpha positive peaks. They thus occur at periods of relative inhibition when gamma bursts are absent, and no enhancement of neural activity may occur. When alpha oscillations are sufficiently suppressed neurons can fire freely and TMS pulses delivered during this period can enhance gamma bursts irrespective of the phase of alpha oscillations.

[00125] In one example, timing of gamma bursts in the EEG is commonly modulated by alpha phase. The alpha cycle supposedly acts here as periodic inhibition— gamma bursts occur only during the cortical surface-negative troughs of the alpha cycle, and when the amplitude of alpha oscillations is sufficiently low. The strength of this relationship may change with movement or other cortical activation states as shown in FIG. 6. The modulation of stimulus-induced gamma-band oscillations through alpha oscillatory phase may be demonstrated by applying weak alternating currents at subject’s individual alpha frequency 4 Hz to the occipital cortex to mimic the functional effects of periodic inhibition during spontaneous alpha oscillations. The induced currents rhythmically can suppress visual stimulus-induced gamma-band power. The degree of gamma-band suppression predicts the reduction in visual detection performance, suggesting a direct modulation of cortical excitability by rhythmically shifting the neurons’ membrane potential.

[00126] Many EEG and ECoG studies show that a decrease in mu power in motor cortices is related to increased activation of the cortical area. During movement as well as other activation states known to transiently block mu rhythm amplitude, alpha-gamma PAC may be diminished or eliminated, and gamma bursts may occur freely throughout the alpha cycle. Other studies investigating the relationship between corticospinal excitability (as measured with MEPs) and alpha power showed that MEPs are larger when pre-stimulus mu power is lower, and pre-stimulus gamma power is higher. In one example, this effect is specific for local EEG alpha activity at sites overlying the cortical motor areas to which the TMS pulses were applied (as verified using source localization).

[00127] Thus, during a cortical activation state where alpha/mu power is suppressed, TMS pulses delivered at any phase of mu/alpha cycles may increase neuronal firing thus increasing subsequently cortical excitability. Instead during periods of increased mu/alpha power TMS pulses may best be delivered during surface negative alpha troughs to increase cortical activation states to be most effective.

[00128] In some embodiments of the disclosed technology, mechanisms of PAC in local cortical brain field activities, the most prominent of which may dominate scalp EEG signals, may be exploited as a tool for more efficient TMS stimulation by incorporating information on the timing of neuronal excitability states.

[00129] Clinical TMS therapy has not changed much over the last 30 years with similar treatment protocols applied across different patient groups and a variety of disorders. One of the main practical issues in TMS therapy is that TMS after-effects are notoriously inconsistent, the same stimulation protocol inducing neural plasticity effects in opposite directions. The embodiments of the disclosed technology provide various methods to improve stimulation protocols by increasing neuroplasticity through timing TMS pulses to oscillatory high excitability phases. These results can be obtained in the motor cortex, however high and low excitability oscillatory phases may differ over brain areas and frequency bands. The method implemented based on some embodiments of the disclosed technology utilizes estimation of PAC to determine which exact phase of a given oscillation in the target brain area has the highest excitability. Methods for the estimation of event-related and time- resolved PAC may be either implemented before TMS stimulation or integrated into a real time system to adapt timing of TMS pulses. Real-time estimation of PAC during TMS stimulation might be used to index of neuroplasticity and help determine the efficiency of the stimulation or predict the success of TMS therapy.

[00130] FIG. 7 shows an example method for magnetic stimulation based on some embodiments of the disclosed technology. The method includes, at 710, identifying, from endogenous oscillatory data, an excitable phase of a target oscillation, at 720, calculating a repetitive magnetic pulse based on the excitable phase of the target oscillation, and, at 730, sending an instruction for generating the repetitive magnetic pulse with an external electromagnetic stimulus.

[00131] FIG. 8 shows another example method for magnetic stimulation based on some embodiments of the disclosed technology. The method includes, at 810, identifying, from a first set of endogenous oscillatory data of a living organism, a first excitable phase of a target oscillation, at 820, calculating a first repetitive magnetic pulse based on the first excitable phase of the target oscillation, at 830, applying the first repetitive magnetic pulse to the living organism, and at 840, receiving, in response to applying the first repetitive magnetic pulse to the living organism, a second set of endogenous oscillatory data from the living organism, wherein the second set of endogenous oscillatory data includes an amplitude increase in the target oscillation. [00132] The method for magnetic stimulation based on some embodiments of the disclosed technology may further include, at 850, identifying, from the second set of endogenous oscillatory data, a second excitable phase of the target oscillation, at 860, calculating a second repetitive magnetic pulse based on the second excitable phase of the target oscillation, and at 870, applying the second repetitive magnetic pulse to the living organism based on the response of the target oscillation of the first repetitive magnetic pulse to the living organism.

[00133] FIG. 9 shows an example method of a non-invasive brain stimulation based on some embodiments of the disclosed technology. The method includes, at 910, applying, to nerve cells in a brain, pulses to induce changes in oscillatory brain activity, at 920, estimating an excitable phase of a target oscillation of the brain based on a coupling between the applied pulses and the oscillatory brain activity, and, at 930, adjusting stimulation parameters at least based on the estimated excitable phase of the target oscillation.

[00134] While this patent document contains many specifics, these should not be construed as limitations on the scope of any invention or of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular inventions. Certain features that are described in this patent document in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

[00135] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. Moreover, the separation of various system components in the

embodiments described in this patent document should not be understood as requiring such separation in all embodiments.

[00136] It is specifically intended that embodiments not be limited to the embodiments and illustrations contained herein, but include modified forms of those embodiments including portions of the embodiments and combinations of elements of different embodiments as come within the scope of the following claims.

Claims

CLAIMS WE CLAIM:

1. A method for magnetic stimulation, comprising:

identifying, from endogenous oscillatory data, an excitable phase of a target oscillation;

calculating a repetitive magnetic pulse based on the excitable phase of the target oscillation; and

sending an instruction for generating the repetitive magnetic pulse with an external electromagnetic stimulus.

2. The method of claim 1, wherein the identifying of the excitable phase of the target oscillation is performed in real-time with the calculating of the repetitive magnetic pulse based on the excitable phase of the target oscillation.

3. The method of claim 1, wherein the endogenous oscillatory data is oscillatory brainwave data.

4. The method of claim 3, wherein the oscillatory brainwave data includes spatial, temporal, and spectral information from an electrophysiological region.

5. The method of claim 1, wherein the endogenous oscillatory data includes at least one of frequency, phase, amplitude, type of brainwave, emission area of the cranium, neural emitter, or electrophysiological data.

6. The method of claim 1, wherein the repetitive magnetic pulse has a phase corresponding to the excitable phase of the target oscillation.

7. The method of claim 1, wherein the repetitive magnetic pulse has a phase at or about a negative peak of the target oscillation.

8 The method of claim 1, further comprising: measuring a frequency, a phase, or an amplitude of the excitable phase of the target oscillation.

9. The method of claim 8, wherein the measuring of the frequency, the phase, or the amplitude of the excitable phase of the target oscillation is performed through autoregressive modeling.

10. The method of claim 1, further comprising:

providing instructions by which the repetitive magnetic pulse is to be generated.

11. The method of claim 1 , wherein the endogenous oscillatory data includes a frequency sweep across an endogenous oscillation bandwidth.

12. The method of claim 1, wherein the endogenous oscillatory data is a phase sweep across a sinusoidal parameter of an endogenous oscillation bandwidth.

13. The method of claim 1, wherein the endogenous oscillatory data includes data obtained from at least one of functional magnetic resonance or positron emission

tomography.

14. A method for magnetic stimulation, comprising:

identifying, from a first set of endogenous oscillatory data of a living organism, a first excitable phase of a target oscillation;

calculating a first repetitive magnetic pulse based on the first excitable phase of the target oscillation;

applying the first repetitive magnetic pulse to the living organism; and

receiving, in response to applying the first repetitive magnetic pulse to the living organism, a second set of endogenous oscillatory data from the living organism, wherein the second set of endogenous oscillatory data includes an amplitude increase in the target oscillation.

15. The method of claim 14, further comprising: identifying, from the second set of endogenous oscillatory data, a second excitable phase of the target oscillation;

calculating a second repetitive magnetic pulse based on the second excitable phase of the target oscillation; and

applying the second repetitive magnetic pulse to the living organism based on the response of the target oscillation of the first repetitive magnetic pulse to the living organism.

16. The method of claim 15, further comprising:

providing instructions by which the repetitive magnetic pulse is to be generated.

17. The method of claim 15, wherein the identifying of the second excitable phase of the target oscillation is performed in real-time with the calculating of the second repetitive magnetic pulse based on the second excitable phase of the target oscillation.

18. The method of claim 15, further comprising:

measuring a frequency, a phase, or an amplitude of the excitable phase of the target oscillation.

19. The method of claim 18, wherein the measuring of the frequency, the phase, or the amplitude of the excitable phase of the target oscillation is performed through autoregressive modeling.

20. The method of claim 14, wherein the endogenous oscillatory data is oscillatory brainwave data.

21. The method of claim 20, wherein the oscillatory brainwave data includes spatial, temporal, and spectral information from an electrophysiological region.

22. The method of claim 14, wherein the endogenous oscillatory data includes at least one of frequency, phase, amplitude, type of brainwave, emission area of the cranium, neural emitter, or electrophysiological data.

23. The method of claim 14, wherein the endogenous oscillatory data includes a frequency sweep across an endogenous oscillation bandwidth.

24. The method of claim 14, wherein the endogenous oscillatory data is a phase sweep across a sinusoidal parameter of an endogenous oscillation bandwidth.

25. The method of claim 14, wherein the endogenous oscillatory data includes data obtained from at least one of functional magnetic resonance or positron emission

tomography.

26. A method of a non-invasive brain stimulation, comprising:

applying, to nerve cells in a brain, pulses to induce changes in oscillatory brain activity;

estimating an excitable phase of a target oscillation of the brain based on a coupling between the applied pulses and the oscillatory brain activity; and

adjusting stimulation parameters at least based on the estimated excitable phase of the target oscillation.

27. The method of claim 26, the stimulation parameters include at least one of a cortical location, a frequency of intrinsic brain oscillations, or a phase of an intrinsic brain activity.

28. The method of claim 26, the coupling includes a phase amplitude coupling (PAC).

29. The method of claim 26, the applying of the pulses includes repetitive transcranial magnetic stimulation (rTMS) pulses.

30. The method of claim 29, the adjusting of the stimulation parameters includes tuning a frequency of the rTMS stimulation pulses to endogenous oscillation frequency by using blind source separation to detect target frequency peaks in electrophysiological signals

31. A system comprising:

a non-volatile memory;

at least one processor configured to identify, from endogenous oscillatory data, an excitable phase of a target oscillation; and

calculate a repetitive magnetic pulse based on the excitable phase of the target oscillation.

32. The system of claim 31 , wherein the identifying of the excitable phase of the target oscillation is performed in real-time with the calculating of the repetitive magnetic pulse based on the excitable phase of the target oscillation.

33. The system of claim 31, wherein the endogenous oscillatory data is oscillatory brainwave data.

34. The system of claim 33, wherein the oscillatory brainwave data includes spatial, temporal, and spectral information from an electrophysiological region.

35. The system of claim 31, wherein the endogenous oscillatory data includes at least one of frequency, phase, amplitude, type of brainwave, emission area of the cranium, neural emitter, or electrophysiological data.

35. The system of claim 31, wherein the at least one processor is further configured to adjust stimulation parameters at least based on the excitable phase of the target oscillation.

36. The system of claim 35, the stimulation parameters include at least one of a cortical location, a frequency of intrinsic brain oscillations, or a phase of an intrinsic brain activity.

37. The system of claim 31, the identifying of the excitable phase of the target oscillation is performed based on a phase amplitude coupling (PAC) between the repetitive magnetic pulse and an oscillatory brain activity.