Systems and methods for brain neuromodulation

Abstract

There is provided, a system for neuromodulation, comprising: at least one processor executing a code for: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, in a plurality of iterations: generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, monitoring the current activity of the neural network in response to the non-invasive feedback, and terminating the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

Description

[0001] SYSTEMS AND METHODS FOR BRAIN NEUROMODULATION

[0002] RELATED APPLICATION

[0003] This application claims the benefit of priority of U.S. Provisional Patent Application Nos. 63 / 553,700 filed on February 15, 2024 and 63 / 649,488 filed on May 20, 2024, the contents of which are incorporated herein by reference in their entirety.

[0004] FIELD AND BACKGROUND OF THE INVENTION

[0005] The present invention, in some embodiments thereof, relates to valence representations and, more specifically, but not exclusively, to systems and methods based on measuring valence representations of neurons.

[0006] In neuroscience, the valence representation of neurons refers to how individual neurons or groups of neurons encode the emotional valence of stimuli. Emotional valence refers to the intrinsic attractiveness or aversiveness of an event, object, or situation. It is typically represented along a continuum ranging from positive (pleasant) to negative (unpleasant).

[0007] SUMMARY OF THE INVENTION

[0008] According to a first aspect, a system for neuromodulation, comprises: at least one processor executing a code for: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, in a plurality of iterations: generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, monitoring the current activity of the neural network in response to the non-invasive feedback, and terminating the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

[0009] According to a second aspect, a computer implemented method for neuromodulation, comprising: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, in a plurality of iterations: generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, monitoring the current activity of the neural network in response to the non-invasive feedback, and terminating the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

[0010] According to a third aspect, a non-transitory medium storing program instructions for neuromodulation, which when executed by at least one processor, cause the at least one processor to: compute a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, in a plurality of iterations: generate instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, monitor the current activity of the neural network in response to the non-invasive feedback, and terminate the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

[0011] According to a fourth aspect, at least one processor executes a code for: in a plurality of iterations: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence according to signals sensed by at least one electrode when implanted in the at least one region of the brain, generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, monitoring the current activity of the neural network in response to the non-invasive feedback according to the signals sensed by the at least one electrode when implanted in the at least one region of the brain, dynamically adapting the non-invasive feedback during each respective iteration according to a real-time or near-real time current valance state measurement of the neural network, and terminating each respective iteration when the current activity of the neural network meets the predefined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold. In a further implementation form of the first, second, third, and fourth aspects, the predefined state is defined as a state that corresponds to a second external non-invasively applied stimuli with neutral valence, or with valence opposite to the valence used to compute the baseline state. In a further implementation form of the first, second, third, and fourth aspects, the predefined state is conditioned with a neutral valence or with a valence opposite to the valence used to compute the baseline state.

[0012] In a further implementation form of the first, second, third, and fourth aspects, the valence of the applied stimuli used to identify the neural network comprises a negative valence, and the non-invasive feedback is designed to shift the baseline state of activity to be closer to the predefined state identified and / or associated with a neutral valence and / or a less negative valence and / or more positive valence and / or less intense valence than the valence of the applied stimuli.

[0013] In a further implementation form of the first, second, third, and fourth aspects, the valence of the applied stimuli used to identify the neural network comprises a positive valence, and the non-invasive feedback is designed to shift the baseline state of activity to be closer to the predefined state identified and / or associated with a neutral valence and / or a more negative valence and / or less positive valence than the valence of the applied stimuli.

[0014] In a further implementation form of the first, second, third, and fourth aspects, the subject has an addiction associated with a positive valence that is triggered by the external non-invasively applied stimuli, and the subject is being treated for the addiction by generating the non-invasive feedback for changing from the baseline state to the pre-defined state of activity identified and / or associated with the neutral and / or the more negative valence and / or the less positive valence than the valence associated with the addiction.

[0015] In a further implementation form of the first, second, third, and fourth aspects, the external non-invasively applied stimuli and / or the non-invasive feedback applied during the iterations, are generated by at least one of: a visual presentation on a display, audio played over speakers, a tactile sensation generated by a haptic device, a decision-making task, and a motor skill task.

[0016] In a further implementation form of the first, second, third, and fourth aspects, the first predefined threshold and / or the second pre-defined threshold are selected for treatment of the subject for a mental disorder triggered by the external non-invasively applied stimulus.

[0017] In a further implementation form of the first, second, third, and fourth aspects, the mental disorder is selected from: posttraumatic stress disorder (PTSD), anxiety, phobia, obsessive compulsive disorder, and social disorder.

[0018] In a further implementation form of the first, second, third, and fourth aspects, the non- invasive feedback is generated according to a transfer function that maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state. In a further implementation form of the first, second, third, and fourth aspects, the transfer function is implemented as a machine learning model, and the non-invasive feedback is generated by the machine learning model in response being fed the baseline state and the current activity of the neural network.

[0019] In a further implementation form of the first, second, third, and fourth aspects, the machine learning model is trained on a training dataset of multiple records, wherein a record is generated based on a sample subject, the record including a sample baseline state of the neural network, a sample current activity of the neural network, a sample pre-defined state, and a ground truth of a target non-invasive feedback that when applied to the subject shifted the sample current activity from the sample baseline state towards the sample pre-defined state.

[0020] In a further implementation form of the first, second, third, and fourth aspects, further comprising code for dynamically training the machine learning model during an initial phase in which records are created while the non-invasive feedback is varied and resulting current activity is computed, and the trained machine learning model is used during a subsequent phase in which the non-invasive feedback is generated for guiding the current activity toward the pre-defined state.

[0021] In a further implementation form of the first, second, third, and fourth aspects, further comprising code for dynamically selecting the transfer function from a plurality of defined transfer functions according to at least one of: brain region, signal quality, and clinical protocol.

[0022] In a further implementation form of the first, second, third, and fourth aspects, the baseline state of activity and the current activity of the neural network is computed based on at least one of: single-neurons (1-many), local-field-potentials, synchronicity between activity in different cells within a region, synchronicity between different cells in different brain regions, as a multidimensional representation that is computed from the activity of many cells (2-many), an evoked local-field-potential including a spectral power and / or phase across several spectral frequencies, and synchronicity of power and / or phase across electrodes located in one or more brain regions

[0023] In a further implementation form of the first, second, third, and fourth aspects, the current activity is computed from signals obtained from at least one electrode surgically implanted within the at least one region of the brain of the subject and configured for monitoring at least one of: action potentials of individual neurons, and local field potentials of at least one brain region.

[0024] In a further implementation form of the first, second, third, and fourth aspects, the electrode is designed to sense a neuronal population of less than about 1000 neurons, wherein the current activity of the neural network is computed for the neuronal population.

[0025] In a further implementation form of the first, second, third, and fourth aspects, the region of the brain in which the at least one electrode is implanted includes one or more of: an amygdala, insula, Anterior-cingulate-cortex (ACC), medial-prefrontal-cortex (mPFC), nucleus accumbens (NAc), hippocampus, para hippocampus, and temporal pole.

[0026] In a further implementation form of the first, second, third, and fourth aspects, in each iteration of the plurality of iterations the non-invasive feedback is adapted according to the current activity level of the neural network at a current iteration for incrementally advancing towards the pre-defined state.

[0027] In a further implementation form of the first, second, third, and fourth aspects, further comprising code for: measuring EEG signals by EEG electrodes simultaneously with the baseline state and / or current activity and / or the pre-defined state computed from signals obtained from at least one electrode surgically implanted within the at least one region of the brain, computing a mapping data structure that maps between the EEG signals and the baseline state and / or current activity and / or during the pre-defined state, and in a subsequent phase, without using signals obtained from the at least one electrode surgically implanted within the at least one region of the brain: measuring EEG signals, and applying the mapping data structure to the measured EEG signals for estimating the baseline state and / or current activity and / or the pre-defined state, wherein the generating the non-invasive feedback is generated based on the current activity and / or the baseline state and / or the pre-defined state estimated based on the EEG signals.

[0028] In a further implementation form of the first, second, third, and fourth aspects, further comprising the feedback device.

[0029] In a further implementation form of the first, second, third, and fourth aspects, further comprising: measuring local field potentials (LFP) simultaneously with spikes from signals obtained from at least one electrode surgically implanted within the at least one region of the brain, wherein the baseline state and / or current activity and / or pre-defined state are computed from a combination of the measured LFP and measured spikes, computing a mapping data structure that maps between the LFP signals and the spikes, in response to lack of sufficient spike measurements, measuring LFP, applying the mapping data structure to the measured LFP for estimating the spike measurement, and computing the baseline state and / or current activity and / or pre-defined state as a combination of the measured LFP and / or estimated spike measurements.

[0030] In a further implementation form of the first, second, third, and fourth aspects, further comprising code for dynamically adjusting a transfer function according to a state of available signals selected from: LFP without significant spike measurement, spike measurement without significant LFP, and a combination of LFP and spike measurements, wherein the transfer function maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state. In a further implementation form of the first, second, third, and fourth aspects, a single iteration is performed in about one minute or less.

[0031] In a further implementation form of the first, second, third, and fourth aspects, at least 10 iterations are performed during a single session.

[0032] In a further implementation form of the first, second, third, and fourth aspects, in a single iteration, the external non-invasively applied stimuli with valence is active for less than about 5 seconds, and then de-activated until a next iteration.

[0033] In a further implementation form of the first, second, third, and fourth aspects, the non- invasive feedback is continuously dynamically adapted throughout the iteration.

[0034] In a further implementation form of the first, second, third, and fourth aspects, the computing the baseline state and the monitoring the current activity are performed at a sampling rate higher than 2000 Hz.

[0035] In a further implementation form of the first, second, third, and fourth aspects, the computing the baseline state and the monitoring the current activity is performed by measuring single spikes of neurons.

[0036] In a further implementation form of the first, second, third, and fourth aspects, monitoring the current activity of the neural network in response to the non-invasive feedback comprises monitoring the real-time or near-real time current valance state measurement of the neural network.

[0037] Unless otherwise defined, all technical and / or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention pertains. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of embodiments of the invention, exemplary methods and / or materials are described below. In case of conflict, the patent specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not intended to be necessarily limiting.

[0038] BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0039] Some embodiments of the invention are herein described, by way of example only, with reference to the accompanying drawings. With specific reference now to the drawings in detail, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of embodiments of the invention. In this regard, the description taken with the drawings makes apparent to those skilled in the art how embodiments of the invention may be practiced.

[0040] In the drawings: FIG. 1 is a block diagram of components of a system for modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity, in accordance with some embodiments of the present invention;

[0041] FIG. 2 is a flowchart of a method of modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity, in accordance with some embodiments of the present invention;

[0042] FIG. 3 is a schematic depicting an exemplary dataflow of a brain-computer interface for neuromodulation of a neural network of a brain of a subject via a closed loop, in accordance with some embodiments of the present invention;

[0043] FIG. 4 includes graphs depicting an example of how valence is represented in the activity of single-neurons in the amygdala, in accordance with some embodiments of the present invention;

[0044] FIG. 5 includes graphs of a baseline neural activity, a transfer function, and activity in a second state after generating feedback, in accordance with some embodiments of the present invention;

[0045] FIG. 6 includes graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention;

[0046] FIG. 7 includes additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention;

[0047] FIG. 8 includes yet additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention;

[0048] FIG. 9 includes yet additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention; and

[0049] FIG. 10 includes panels depicting graphs and / or schematics indicating examples of more complex behaviors that are exhibited in anxiety and / or PTSD which may be treated using neuromodulation, in accordance with some embodiments of the present invention.

[0050] DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION

[0051] The present invention, in some embodiments thereof, relates to valence representations and, more specifically, but not exclusively, to systems and methods based on measuring valence representations of neurons.

[0052] An aspect of some embodiments of the present invention relates to systems, methods, computing devices, and / or code instructions (stored on a data storage device and executable by one or more processors) for modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity, for example, for treatment of mental disorders such as post-traumatic stress disorder (PTSD) and / or treatment of addictions. A baseline state of activity of a neural network in one or more regions of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, is computed. The baseline state may represent the state of the subject that to be treated, for example, a disturbing memory and / or a distressing emotion and / or invasive thoughts, such as in a patient suffering from PTSD, that is triggered by the external non-invasively applied stimuli with valence. In one or multiple iterations, a non-invasive feedback is generated, for example, an audio tone, a visual presentation, and / or a tactile sensation. The non-invasive feedback is designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity. The non-invasive feedback may be dynamically adapted for incentivizing the brain, for example, a task is increased in difficulty, and / or a tone where different frequencies of the tone are associated with varying appetitive / aversive reward levels is changed. The pre-defined state of activity may represent the state at which the patient experiences a certain way, for example, relief from the disturbing memory and / or the distressing emotion. Other examples of the pre-defined state of activity are described herein. The current activity of the neural network in response to the non-invasive feedback is monitored. The monitoring is performed while the non-invasive feedback is dynamically adapted. The iterations are terminated when the current activity of the neural network meets the pre-defined state within a first predefined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

[0053] Optionally, the features including computing baseline state of activity in response to an external non-invasively applied stimuli with valence and / or including the monitoring the current activity, are dynamically iterated in multiple iterations as part of a single session (i.e., single treatment session). In each iteration, the non-invasive feedback may be dynamically adapted, optionally continuously, optionally in real-time (or near real-time), according to a real-time current valence state measurement of the neural network of the subject. The features of computing the baseline state of activity and / or the monitoring the current activity are performed according to signals sensed by one or more electrodes when implanted in the region of the brain. Using the implanted electrode, each iteration described herein, that include the features of computing the baseline state of activity, generating instructions for operating a feedback device for generating a non-invasive feedback, and monitoring the current activity of the neural network in response to the non-invasive feedback, may be performed in about 1 minute or less, or in about 5, 4, 3, or 2 minutes or less, or about 45, 30, or 15 seconds or less. Alternatively or additionally, the number of iterations performed in a single session is at least about 5, or 10, or 15, or 20, or greater. Each respective iteration may be terminated when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

[0054] An example of the goal of treatment, which is not necessarily part of the criteria for success in a specific session of treatment, is now described: When the subject has reached the pre-defined state, the initial external non-invasively applied stimuli with valence that triggered the disturbing memory and / or distressing emotion and / or invasive thoughts has shifted to triggering a more positive and / or less negative emotional and / or less intense response (e.g., positive valence) and / or has shifted to triggering a tolerable memory and / or tolerable emotion and / or tolerable thoughts (e.g., less negative and / or less intense valence) and / or has shifted to triggering a non-consequential memory and / or emotion and / or thought (e.g., neutral valence).

[0055] As used herein, the phrase stimulus (or stimuli) with valence may refer to a stimulus with negative valence, or a stimulus with a positive valence. A stimulus with negative valence may be a stimulus that a subject would prefer to avoid and / or a stimulus causing negative subjective sensation, for example, a picture that elicits a bad memory, a picture that elicits anxiety, and / or a sound that elicits fear. A stimulus with positive valence may be a stimulus that the subject would want / desire, and / or a stimulus causing positive subjective sensation for example, a picture of a tasty food or of a drug. It is noted that the stimulus may be combination of positive and negative valence, in which case the positive or negative valence may refer to the overall and / or higher effect. For example, the stimulus may result in a greater positive sensation but also associated with a smaller effect of an undesirable sensation such as stress. As used herein, the term positive or negative refers to the dominant effect, without necessarily excluding a smaller opposite effect.

[0056] As used herein, the change from the baseline state to the pre-defined state may be based on change on valence and / or change in intensity. A change from a baseline state that is negative may be to a state that is less-negative or more-positive or negative with less intensity. A change from a baseline state that is positive may be to a state that is less-positive or more-negative or positive with less intensity.

[0057] Other examples of definitions of the pre-defined state of activity are now described. Where the pre-defined state of activity was identified earlier (i.e., while presenting the stimuli to identify the baseline state) stimuli with a neutral / less-negative / positive value are also presented for defining the pre-defined state). This pre-defined state is then determined as being reached, optionally within a pre-defined threshold. In another example, a new state that is significantly different (e.g., statistically significantly different) than the baseline state is selected. Every time the neural state gets closer enough to this pre-defined state (e.g., within a threshold), the new state is reinforced with a positive outcome, hence enforcing this new state to be positive / neutral, and / or vice versa.

[0058] As used herein, the activity of the neural network may refer to a neural representation of the neural network.

[0059] At least one embodiment described herein addresses the technical problem and / or the medical problem of providing a system and / or method for neuromodulation, such as for treatment of mental disorders. At least one embodiment described herein improve the technology of neuromodulation, such as improving treatment of mental disorders. At least one embodiment described improves upon prior approaches for neuromodulation, such as improving over prior approaches for treating mental disorders.

[0060] Mental disorders take a huge and significant toll on individuals and societies. Some current treatment approaches are based on systemic pharmacology, for example, drugs such as SSRI / SNRI. These drugs are designed to target the availability of neurotransmitters across the brain and are therefore not specific and not precise in addressing the source of the disorder. This in turn leads to very limited success in patients and major side-effects. Other approaches include behavioral therapy (e.g., CBT and CPT, Cognitive-Behavioral / Proces sing-Therapy), which commonly relies on exposure and re-appraisal of a value with an attempt to behaviorally manipulate the patient to assign a new value to the original experience / memory / stimuli. These approaches lack specificity and sensitivity to the neural representations that give rise to the symptoms, lack specificity to the individual brain, and do not target directly the pathophysiological source.

[0061] A network of neurons from different brain regions processes emotional / affective information and / or communicates with other brain regions to assign value to experiences and to impact behavior in different scenarios. These brain networks are mainly involved in intense experiences and / or memories, and show malfunctioned activity in PTSD, anxiety, and other mental disorders. The brain network may involve among others the Amygdala, the anterior-cingulate and / or ventromedial pre-frontal cortices, the Insula, the Hippocampus, the nucleus accumbens and / or ventral striatum.

[0062] Recent approaches to manipulating activity in the amygdala and related structures that process affective / emotional information have not yielded promising results and do not address the core problem. This could be because these approaches target a brain area as a whole and measure an overall sum of the activity in these regions indirectly (e.g., measuring electrical / magnetic waves from the brain surface as in EEG, or BOLD activity in fMRI). Therefore, they do not measure and manipulate the specific components of the neural representations that give rise to the malfunction.

[0063] Instead, representations of valence in the amygdala and related regions reside in the activity of populations of neurons that are spread across the whole region and not in any specific anatomical part within it; they do not necessarily have specific molecular properties; and they change from person-to-person and from memory / stimuli / experience to another.

[0064] Targeting a whole brain region with methods that lack spatial and temporal resolution such as Electro Encephalograms (EEG) and / or fMRI, cannot be specific to the neural network that underlies the original stimulus / experience / memory. Therefore, if used for providing feedback, it might lead to side effects in addition to harming other behaviours represented in this network. In this respect, these approaches are similar to the pharmacological approach and lack specificity and sensitivity and do not target the specific malfunction.

[0065] At least one embodiment described herein solves the aforementioned technical problem(s), and / or improves the aforementioned technical field(s), and / or improves upon the aforementioned prior approach(es), by modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity. A baseline state of activity of a neural network in one or more regions of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence, is computed. The external non-invasively applied stimuli with valence may, for example, trigger a traumatic memory and / or disturbing emotion such as in a patient suffering from PTSD. In one or multiple iterations, a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity, is generated. The current activity of the neural network in response to the non-invasive feedback is monitored. The iterations are terminated when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold. The pre-defined state of activity may represent a treatment target, such as where the patient feels less distress and / or positive emotions in response to the non-invasive feedback.

[0066] In at least one embodiment described herein, a brain-computer-interface may be applied using a real-time closed-loop approach to manipulate valence of stimuli and / or memories. The manipulation is designed to generate a shift in the activity of neurons associated with the stimuli and / or memory to be more-positive and / or less-negative (or vice versa, less positive and / or morenegative). This modulation of activity is designed to lead to a modulation in the behavioural response to the stimuli (e.g., memories). At least one embodiment described herein may be used to alleviate symptoms that are attributed to mental disorders associated with a highly emotional / affective memory / experience / stimuli. For example, in PTSD and anxiety, shifting the activity of neurons associated with an original aversive event, may lead to less negative behavioral outcomes when the subject is confronted (or spontaneously remembers) with a similar experience. Other mental disorders where a maladaptive behavior can be triggered by a specific stimuli (e.g., memory and / or experience), may be treated. For example, drug-abuse (addiction), phobias, obsessive-compulsive disorders, and the like.

[0067] At least one embodiment described herein addresses the technical problem and / or the medical problem of providing a system and / or method for improving treatment of a mental disorder based on providing an external stimulus. At least one embodiment described herein improves the technical field of neuromodulation by providing a system and / or method for improving treatment of a mental disorder based on providing an external stimulus. At least one embodiment described herein improves upon prior approaches of providing a system and / or method for improving treatment of a mental disorder based on providing an external stimulus that is measurable by the implanted electrode. Standard neuromodulation treatment approaches, for example, neurofeedback based on EEG and / or fMRI, are slow, requiring multiple sessions over a long time interval to achieve results. For example, to obtain noticeable results using standard neurofeedback, about 20-40 minutes may be required, where sessions are usually conducted about 1-3 times per week, where each session typically lasts about 30-60 minutes.

[0068] At least one embodiment described herein improves upon the aforementioned technical problem, and / or improves upon the aforementioned technical field, and / or improves upon standard neurofeedback approaches, and / or provides the practical application of, using signals sensed by electrode(s) implanted inside the brain for rapid iterations of computing a baseline state of activity in response to an external non-invasively applied stimuli with valence, monitoring current activity of the neural network in response to a non-invasive feedback, and dynamically adapting the non- invasive feedback accordingly. Performing each iteration rapidly enables performing multiple iterations during a single session. The rapid iterations may be based on a fast sampling rate of the implanted electrode(s). The external non-invasively applied stimuli with valence may be selected to trigger rapidly a baseline state of activity, and / or the non-invasive feedback may be selected to trigger a rapid change in current activity of the neural network. The change in the current activity of the neural network of the subject may be an aggregate of multiple relatively small incremental changes in response to the dynamic adaption of the non-invasive feedback over each single iterations. Each iteration may terminate when the current activity of the neural network meets the pre-defined state within the first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by the second pre-defined threshold. Both the non-invasive stimuli with valence and the non-invasive feedback are selected to be sufficiently fast to allow the shift from one state to the other (e.g., the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold) within each iteration, optionally in less than a minute. The external non-invasively applied stimuli with valence is presented at the beginning of each iteration. The presentation of the external non-invasively applied stimuli and the expected response should be fast, for example, within less than 5 seconds. For example, the non-invasive stimuli with valence may be presented for about 2 seconds. It may take about 2 seconds for the user to apprehend and / or understand the non-invasive stimuli. The non-invasive feedback may be for example, an analogue signal allowing immediate detection of change, so as to allow fast allow convergence. The change in the current activity of the neural network in response to the following non-invasive feedback may terminate in less than about 1 minute in each iteration. Inventors discovered that multiple rapid iterations increase effectiveness of the induced neuromodulation, for example, based on the experiment conducted by Inventors described in the “Examples” section below.

[0069] Before explaining at least one embodiment of the invention in detail, it is to be understood that the invention is not necessarily limited in its application to the details of construction and the arrangement of the components and / or methods set forth in the following description and / or illustrated in the drawings and / or the Examples. The invention is capable of other embodiments or of being practiced or carried out in various ways.

[0070] The present invention may be a system, a method, and / or a computer program product. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present invention.

[0071] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0072] Computer readable program instructions described herein can be downloaded to respective computing / processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and / or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing / processing device.

[0073] Computer readable program instructions for carrying out operations of the present invention may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present invention.

[0074] Aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer readable program instructions.

[0075] These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks. These computer readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function / act specified in the flowchart and / or block diagram block or blocks.

[0076] The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0077] The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustration, and combinations of blocks in the block diagrams and / or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.

[0078] Reference is now made to FIG. 1, which is a block diagram of components of a system 100 for modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity, in accordance with some embodiments of the present invention. Reference is also made to FIG. 2, which is a flowchart of a method of modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity, in accordance with some embodiments of the present invention. Reference is also made to FIG. 3, which is a schematic depicting an exemplary dataflow of a brain-computer interface for neuromodulation of a neural network of a brain of a subject via a closed loop, in accordance with some embodiments of the present invention. Reference is also made to FIG. 4, which includes graphs depicting an example of how valence is represented in the activity of single-neurons in the amygdala, in accordance with some embodiments of the present invention. Reference is also made to FIG. 5, which includes graphs of a baseline neural activity, a transfer function, and activity in a second state after generating feedback, in accordance with some embodiments of the present invention. Reference is also made to FIG. 6, which includes graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention. Reference is also made to FIG. 7, which includes additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention. Reference is also made to FIG. 8, which includes yet additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention. Reference is also made to FIG. 9, which includes yet additional graphs depicting results of experiments for neuromodulation of a neural network of a brain, in accordance with some embodiments of the present invention. Reference is also made to FIG. 10, which includes panels depicting graphs and / or schematics indicating examples of more complex behaviors that are exhibited in anxiety and / or PTSD which may be treated using neuromodulation, in accordance with some embodiments of the present invention.

[0079] System 100 described with reference to FIG. 1 may implement the features of the method described with reference to FIGs 2-3, by one or more hardware processors 102 of a computing device 104 executing code instructions stored in a memory (also referred to as a program store) 106.

[0080] System 100 may include one or more sensors 112 for sensing signals used for computing the valence representation described herein. Optionally, sensors 112 are implemented as one or more electrodes designed for surgical implantation in one or more regions of the brain, optionally the amygdala.

[0081] Sensors 112 may be used for stimulation. Optionally, the electrodes designed for surgical implantation in the brain may be used for application of electrical stimulation. Sensors 112 may be in communication with a controller 108 designed to receive signals from sensors 112 and / or designed to operator sensors 112 to apply electrical stimulation. Controller 108 may be implemented as, for example, a separate controller designed for surgical implantation in the brain and / or to be worn on the body of the subject external to the brain. In other embodiments, controller 108 may be implemented by computing device 104 and / or integrated within computing device 104. Controller 108 may be implemented in hardware and / or firmware and / or software.

[0082] System 100 may include one or more feedback devices 150 designed to generate a non- invasive feedback and / or applied stimulus, as described herein. For example, triggering a memory for which the subject is being treated for (e.g., PTSD, depression) and / or triggering an addition for which the subject is being treated for (e.g., drugs, alcohol, smoking). Examples of feedback devices 150 include a display for presenting images and / or video, speakers for playing audio, a haptic device such as for generating a sensation, a decision-making task, and / or a motor skill task.

[0083] System 100 may include a computing device 104 that obtains measurements made by sensor(s) 112 and / or obtains the valence representation computed from the measurements (e.g., the valence representation may be computed by controller 108). Computing device 104 may compute the value of the valence representation (e.g., baseline and / or current). Computing device 104 may generate instructions for operating feedback device 150 according to the value of the valence representation, as described herein.

[0084] Computing device 104 may be implemented as, for example, a client terminal, a server, a virtual machine, a virtual server, a computing cloud, a mobile device, a desktop computer, a thin client, a Smartphone, a Tablet computer, a laptop computer, a wearable computer, glasses computer, and a watch computer.

[0085] Multiple architectures of system 100 based on computing device 104 may be implemented:

[0086] In an exemplary implementation of a localized architecture, computing device 104 may be a local device executing code 106 A, for example, a standalone computer, code running on a surgical workstation, code running on a mobile device, and / or integrated with controller 108. Computing device 104 is in communication with controller 108 and / or sensor(s) 112 and / or feedback device 150. Computing device 104 obtains measurements made by sensors 112, and / or operates feedback device 150, as described herein. When computing device 104 is physically separate from controller 108 and / or sensor(s) 112, computing device 104 may communicate with controller 108 and / or sensor(s) 112 via a network 110 and / or other data interface.

[0087] In an exemplary implementation of a centralized architecture, computing device 104 storing code 106A may be implemented as one or more servers (e.g., network server, web server, a computing cloud, a virtual server, surgical workstation) that provides centralized services for generating a feedback to one or more controllers 108 and / or feedback devices 150 and / or sensor(s) 112 over network 110, for example, providing software as a service (SaaS), providing software services accessible using a software interface (e.g., application programming interface (API), software development kit (SDK)), providing an application for local download, and / or providing functions using a remote access session. For example, each controller 108 sends measurements made by its associated sensor(s) 112 and / or sends the computed valence representation to computing device 104. Computing device 104 generates instructions for operating respective feedback devices 150 associated with respective controllers 108 and / or sensor(s) 112.

[0088] Sensor (s) 112 may transmit sensed signals to controller 108 and / or computing device 104, for example, via a direct connected (e.g., local bus and / or cable connection and / or short range wireless connection), and / or via network 110 and a network interface 122 of computing device 104 and / or controller 108 (e.g., where sensor(s) 112 and / or controller(s) 108 are connected via short range wired and / or wireless communication channel and / or are located remotely from the computing device).

[0089] One or more components described with reference to computing device 104 may be implemented with respect to controller 108.

[0090] Network interface 122 may be implemented as, for example, a wire connection (e.g., physical port), a wireless connection (e.g., antenna), a network interface card, a wireless interface to connect to a wireless network, a physical interface for connecting to a cable for network connectivity, and / or virtual interfaces (e.g., software interface, application programming interface (API), software development kit (SDK), virtual network connection, a virtual interface implemented in software, network communication software providing higher layers of network connectivity).

[0091] Memory 106 stores code instructions executable by hardware processor(s) 102. Exemplary memories 106 include a random access memory (RAM), read-only memory (ROM), a storage device, non-volatile memory, magnetic media, semiconductor memory devices, hard drive, removable storage, and optical media (e.g., DVD, CD-ROM). For example, memory 106 may code 106A that execute one or more acts of the method described with reference to FIGs. 2-3.

[0092] Computing device 104 may include data storage device 120 for storing data, for example, a repository of baseline valence values 120A designed to store baseline valence values for different subjects or the same subject, a threshold repository 120B designed to store different thresholds such as for different subjects and / or for different treatments, and / or a transfer function repository 120C designed to store different transfer functions such as for different treatment targets. Data storage device 120 may be implemented as, for example, a memory, a local hard-drive, a removable storage unit, an optical disk, a storage device, a virtual memory and / or as a remote server 118 and / or computing cloud (e.g., accessed over network 110).

[0093] Computing device 104 and / or controller 108 may include and / or may be in communication with one or more physical user interfaces 124 that include a mechanism for inputting data (e.g., enter name of subject, select which disorder is being diagnosed) and / or for viewing data, for example, an indication of whether the desired value representation was achieved by a treatment session. Exemplary user interfaces 124 include, for example, one or more of, a touchscreen, a display, a keyboard, a mouse, and voice activated software using speakers and microphone.

[0094] Reference is also made to FIG. 2, which is a flowchart of a method for neuromodulation of a subject for changing from a baseline state of activity of a neural network to a pre-defined state, in accordance with some embodiments of the present invention.

[0095] Referring now back to FIG. 2, at 202, a subject may be selected for treatment using approaches described herein. The treatment may be based on modulating a current activity of a neural network of a brain of a subject to change from a baseline state towards a pre-defined state of activity.

[0096] Subjects suffering from a mental disorder may be selected for treatment. The mental disorder may be triggered by one or more external stimuli, which may generate disturbing emotions and / or memories in the subject.

[0097] The subject may be suffering from a mental disorder in which the subject feels a negative valence, such as negative affect, distress, and / or otherwise disturbing emotions and / or thoughts. Such subjects may be treated by shifting from the negative valence to a neural, positive valence, or less negative valence, and / or less intense valence. Examples of such mental disorders include: PTSD, anxiety, phobia, obsessive compulsive disorder, and social disorder.

[0098] Alternatively, the subject may be suffering from a mental disorder in which the subject feels a positive valence, such as positive affect, euphoria, and / or otherwise pleasant emotions and / or thoughts. The positive valence may be associated with harmful behavior. Such subjects may be treated by shifting from the positive valence to a neural, negative valence, or less positive valence, and / or less intense valence. Examples of such mental disorders include: additions such as to alcohol and / or drugs, and obsessive compulsive disorder (OCD).

[0099] Specific memories and / or stimuli that evoke substantial distress to the subject may be identified.

[0100] A 204, a neural network is identified in one or more regions of a brain of the subject. Examples of one or more regions of the brain in which the electrode(s) is implanted includes: amygdala, insula, Anterior-cingulate-cortex (ACC), medial-prefrontal-cortex (mPFC), nucleus accumbens (NAc), hippocampus, para hippocampus, and temporal pole.

[0101] Activity of the neural network is impacted by an external non-invasively applied stimuli with valence.

[0102] One or more electrodes may be surgically implanted within the region of the brain.

[0103] The electrodes may be designed for monitoring action potentials of individual neurons, and / or monitoring local field potentials of one or more brain regions that include multiple neurons.

[0104] In some embodiments, the electrodes may include one or multiple contacts. Each contact may be associated with about 1-4 nerve cells, or other number. The number of contacts on an electrode may range, for example, from 1 (or 2) to about 1000.

[0105] The electrode(s) may be designed to sense a neuronal population of less than about 5000, or about 1000, or about 500, or about 10 neurons, or other values, or a single neuron. The current activity of the neural network may be computed for the neuronal population as a whole.

[0106] At 206, a baseline state of activity of the neural network that corresponds to the external non-invasively applied stimuli with valence is computed.

[0107] The baseline state of activity may represent the state of the subject while the subject is experiencing (e.g., suffering from) a situation that is to be treated, for example, the state of the subject remembering a traumatic event, the state of the subject feeling a disturbing emotion, the state of the subject experiencing a disturbing thought, and / or unpleasant physiological phenomena (e.g., chest pain, rapid breathing, increased heart rate).

[0108] The baseline state of activity may represent a pre-existing memory, thought, experience, and / or emotion, which is re-triggered by the external non-invasively applied stimuli with valence. Alternatively, the baseline state of activity may represent a new association of the external non- invasively applied stimuli with valence with an appetitive and / or aversive reinforcement.

[0109] Examples of baseline states of activity that correspond to the external non-invasively applied stimuli with valence include:

[0110] • A disturbing memory, for example, associated with PTSD.

[0111] • A distressing emotion, for example, associated with a phobia.

[0112] • Symptoms associated with the valence representation may include physiological changes in the body such as increased heart rate, chest pain, and rapid breathing, for example, associated with anxiety.

[0113] • Urge to perform repetitive tasks, for example, associated with OCD.

[0114] • Feeling of euphoria, for example, associated with an addiction to drugs and / or alcohol. Examples of the generated external non-invasively applied stimuli with valence include:

[0115] • As a visual presentation on a display, such as a still image and / or video. The image and / or video may be, for example, ocean waves crashing against a coast, rain falling into a puddle, written text, a non-realistic pattern such as a mix of colors, a single color, and the like.

[0116] • Audio played over speakers, for example, a tone at a selected frequency, music, speech, and the like.

[0117] • A tactile sensation generated by a haptic device, for example, a surface with a soft texture, a complex surface with elevations and depressions, and non-painful sharp ends.

[0118] • A decision-making task, for example, presenting four differently randomly selected images, and asking the user to pick their favorite. In another example, having the user play a quest video game.

[0119] • A motor skill task, for example, tracing a finger over curves, building a tower of blocks, and threading a thread through an eye of a needle.

[0120] At 208, a non-invasive feedback is generated. The non-invasive feedback is designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity. The non-invasive feedback may be dynamically adapted, for example, according to a current state of activity for reaching the predefined state of activity. For example, adapting a difficulty of a game being played, adapting a frequency of a tone associated with aversive / appetitive reward, and the like.

[0121] The pre-defined state may be defined as a state that corresponds to a second external non- invasively applied stimuli with neutral valence. The pre-defined state is a state that is significantly different from the baseline state, for example, a Euclidean distance above a threshold away from the baseline state. Alternatively or additionally, the pre-defined state may be a state that is defined independently of the baseline state, for example, the activity level falling within a range and / or above another threshold. Alternatively or additionally, the pre-defined state may be defined as a state having valence opposite to the valence used to compute the baseline state.

[0122] The pre-defined state may be conditioned with a neutral valence or with a valence opposite to the valence used to compute the baseline state.

[0123] The non-invasive feedback may be, for example, a visual presentation on a display, audio played over speakers, a tactile sensation generated by a haptic device, a decision-making task, and / or a motor skill task. Examples of the non-invasive feedback may be as described with reference to the examples of the external non-invasively applied stimuli. The non-invasive feedback may be an adaptation of the external non-invasively applied stimuli. For example, the external non-invasively applied stimuli may be a sound, and the non- invasive feedback may be change in the frequency of one component of the sound. In another example, the external non-invasively applied stimuli may be a movie of a war at a certain field, while the non-invasive feedback may be another movie at the same field without the war, instead showing flowers, cows grazing, and children playing.

[0124] Alternatively, the non-invasive feedback may be different than the external non-invasively applied stimuli. For example, the external non-invasively applied stimuli is a movie showing a war scene that triggers traumatic thoughts of war by a patient, while the non-invasive feedback is a tone that changes in frequency, or classical music playing over speakers which reduces the intensity and / or frequency of the traumatic thoughts of war by the patient.

[0125] The neurons may be unable to “switch” their activity quickly and / or in a single session from the baseline towards the pre-defined state of activity. Rather, the non-invasive feedback may be designed to incrementally modulate the current activity from the baseline towards the predefined state of activity, for example, in multiple sessions, and / or over time during a single session.

[0126] The pre-defined state of activity may represent a treatment target, such as a state in which the subject does not experience what the subject experienced in response to the external non- invasively applied stimuli (e.g., neutral valence). Alternatively, the pre-defined state of activity may represent the state in which the subject experiences a significant dampening of what the subject experienced in response to the external non-invasively applied stimuli, such that the quality of life and / or ability of the subject to function is not significantly impacted as it was in response to the external non-invasively applied stimuli (e.g. less negative valence, less positive valence, and / or less intense valence). Alternatively, the pre-defined state of activity may represent an “opposite” of what the subject experienced in response to the external non-invasively applied stimuli (e.g., change from negative valence to positive valence, change from positive valence to negative valence, and / or change in intensity of valence).

[0127] Different exemplary scenarios of the valence and corresponding non-invasive feedback are now described:

[0128] • When the valence of the applied stimuli used to identify the neural network represents a negative valence, the non-invasive feedback may be designed to shift the baseline state of activity to be closer to the pre-defined state identified and / or associated with a neutral valence, and / or to shift to a less negative valence and / or to shift to a more positive valence and / or less intense valence than the valence of the applied stimuli. • When the valence of the applied stimuli used to identify the neural network represents a positive valence, the non-invasive feedback may be designed to shift the baseline state of activity to be closer to the pre-defined state identified and / or associated with a neutral valence, and / or shift to a more negative valence, and / or shift to a less positive valence than the valence of the applied stimuli.

[0129] • When the subject has a mental condition (e.g., addiction) associated with a positive valence that is triggered by the external non-invasively applied stimuli, the subject may be treated for the addiction by generating the non-invasive feedback for changing from the baseline state to the pre-defined state of activity identified, and / or to the pre-defined state associated with the neutral and / or the more negative valence, and / or the predefined state associated with the valence that is less positive than the valence associated with the addiction.

[0130] The non-invasive feedback may be determined according to a transfer function. Additional exemplary details of the transfer function are described with reference to 212 of FIG. 2.

[0131] In some cases, electrical microstimulation might be delivered via one or more electrodes implanted in the brain, to facilitate the modulation of the current activity of the neural network to change from the baseline state towards the pre-defined state of activity. The micro stimulation may be applied, for example, during the delivery of the non-invasive feedback, and / or before and / or after the neuromodulation session, for creating the effect of facilitating modulation. The electrical micro stimulation is different than DBS (deep-brain- stimulation). The electrical microstimulation may be a very mild stimulation for promoting neuro-plasticity, which may help the brain to move from different states, as described herein.

[0132] At 210, the current activity of the neural network in response to the non-invasive feedback is monitored. The current activity of the neural network may be computed while the non-invasive feedback is applied. The current activity of the neural network may be computed from signals outputted by the electrode(s) implanted in the brain of the subject.

[0133] The current activity of the neural network may include the current valance state measurement of the neural network, which may represent a real-time or near-real time current valance state measurement of the neural network.

[0134] The current activity of the neural network (and / or the activity defining the baseline state) may be computed based on one or more of: single-neurons (1-many), local-field-potentials, synchronicity between activity in different cells within a region, synchronicity between different cells in different brain regions, as a multi-dimensional representation that is computed from the activity of many cells (2-many), an evoked local-field-potential including a spectral power and / or phase across several spectral frequencies, and synchronicity of power and / or phase across electrodes located in one or more brain regions

[0135] At 212, a transfer function may be computed (e.g., trained) and / or updated (e.g., adapted).

[0136] The transfer function may be dynamically computed and / or updated during treatment sessions, in response to application of the non-invasive feedback and monitoring of the current activity generated in response to the application of the non-invasive feedback. Alternatively or additionally, the transfer function may be a mathematical operator defined in advance of the treatment session, and dynamically used on data collected during the treatment session for generation of the non-invasive feedback, for example, as described with reference to 208 of FIG. 2.

[0137] The transfer function may by dynamically adapted, for example, over time, over different treatment sessions, and the like. The transfer function may be dynamically adapted, for example, due to changing conditions of the brain, when another transfer function may provide better results after a few sessions in comparison to an initial transfer function, and / or in response to changing conditions of the signal acquisition (e.g., quality of signals).

[0138] Multiple transfer functions may be used for treatment combining multiple neuromodulation effects. For example, one transfer function may be selected based on its predicted ability to lead to neuromodulation of one brain region (e.g., amigdala), while another transfer function may be selected based on a prediction of leading to correlated neuromodulation between different regions in the brain (e.g., amygdala and substanitia-innomata). The selection and / or use of the different transfer functions may follow an assessment of the treatment effect achieved with each transfer function and / or proper selection of the most advantageous transfer function with regard to a desired effect. Additionally, it may be aligned with the treatment progress, for example, starting with evaluation of different transfer functions aiming to maximize the therapeutic effect and further along the treatment shifting / adding transfer functions which promote neuromodulation related to long term neuro-plasticity.

[0139] The transfer function maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state.

[0140] Examples of transfer functions:

[0141] Where the non-invasive stimuli provided as feedback is represented as a linear or nonlinear monotonic function of the number of emitted spikes (e.g., firing rate) of 1.. ,N neurons in a pre-defined time window. • Where the non-invasive stimuli provided as feedback is represented as a monotonic function of the synchrony between emitted spikes of 1..N neurons and 1..N other neurons.

[0142] • Where the non-invasive stimuli provided as feedback is represented as a monotonic function of the distance between the current network state and a pre-defined network state, and the metrics is computed on the low-dimensional representation (manifold) of the combined activity in 2.. ,N neurons.

[0143] The non-invasive stimuli provided as feedback may either have a metric(s) imposed by its physical properties (examples: tone frequency, volume, picture contrast, color) and / or is a set of stimuli with pre-defined order.

[0144] The transfer function may be implemented as a machine learning model. The non-invasive feedback may be generated by the machine learning model in response being fed the current activity of the neural network, and optionally the baseline state and / or the pre-defined state.

[0145] The machine learning model may be trained on a training dataset of multiple records. A record may include at least a sample current activity of the neural network, and a target non- invasive feedback that when applied to a subject shifted the sample current activity towards the sample pre-defined state. The record may further include a sample baseline state of the subject. The record may further include a sample pre-defined state. The ground truth may indicate the target non-invasive feedback that when applied to the subject shifted the sample current activity from the sample baseline state towards the sample pre-defined state.

[0146] Exemplary architectures of the ML model include, for example, statistical classifiers and / or other statistical models, neural networks of various architectures (e.g., convolutional, fully connected, deep, encoder-decoder, recurrent, transformer, graph), support vector machines (SVM), logistic regression, k-nearest neighbor, decision trees, boosting, random forest, a regressor, and / or any other commercial or open source package allowing regression, classification, dimensional reduction, supervised, unsupervised, semi-supervised, and / or reinforcement learning. Machine learning models may be trained using supervised approaches and / or unsupervised approaches.

[0147] The records may be personalized records, which may be dynamically created for the subject during treatment sessions. For example, the non-invasive feedback is dynamically varied, and the corresponding current activity is computed. Records including the non-invasive feedback and corresponding current activity may be dynamically generated, and used to train the customized ML model. The records may include the activity of the baseline state measured for the subject. The customized ML model may serve as the transfer function for generating and / or selecting the non- invasive feedback. Alternatively or additionally, the records may be obtained from different subjects. The ML model may be a generic ML model trained on data from different subjects, and / or used for guiding the non-invasive feedback for different subjects. The generic ML model may be customized for the subject, by using a transfer learning approach to further train the generic ML model on the personalized records of the subject.

[0148] At 214, the current activity is evaluated relative to the pre-defined state.

[0149] The iterations may be terminated when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold. Alternatively or additionally, the iterations may be terminated when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold. The pre-defined state may be implicitly defined as being statistically significantly different from the baseline by the second pre-defined threshold.

[0150] The first pre-defined threshold and / or the second pre-defined threshold may be selected for treatment of the subject for a mental disorder triggered by the external non-invasively applied stimulus. For example, when the subject is suffering from symptoms of PTSD experienced by an external stimulus, reaching and / or exceeding the first pre-defined threshold and / or the second predefined threshold may indicate that the subject is no longer experiencing PTSD symptoms triggered by an external stimulus.

[0151] Alternatively, when the iterations are not terminated, at 216, features described with reference to 208-214 may be iterated over a single treatment session, and / or over multiple treatment sessions.

[0152] Alternatively, at 216, features described with reference to 206-214 are iterated in multiple iterations, optionally rapidly. For example, a single iteration of features 206-214 is performed in in about one minute or less, or about 10, 9, 8, 7, 6, 5, 4, 3, 2 minutes or less, or about 45, 30, or 15 seconds, or less.

[0153] Performing each iteration rapidly enables performing multiple iterations during a single session, for example, at least about 10, or 5, or 15, or 20, or 25, or greater number of iterations. A single session includes multiple iterations performed sequentially over a relatively short amount of time, for example, a total of about 20 - 60 minutes, or about 30 - 90 minutes, and the like. The single session may include rest intervals between subsequent iterations.

[0154] As described herein, when the electrode(s) are implanted within the region(s) of the brain, rapid and / or multiple iterations may be implemented during a single session. The baseline state of activity may be computed according to signals sensed by the electrode(s) when implanted in the region(s) of the brain. The current activity of the neural network in response to the non-invasive stimuli with valence may be monitored according to the signals sensed by the electrode(s) when implanted in the region(s) of the brain. The current activity of the neural network in response to the non-invasive feedback may be monitored according to the signals sensed by the electrode(s) when implanted in the region(s) of the brain.

[0155] Optionally, in each iteration, the external non-invasively applied stimuli with valence may be active for a short amount of time, for example, less than about 5 seconds, or 2 seconds, or 10 seconds, or 15 seconds, or 30 seconds, or other values. After activation, the non-invasive stimuli with valence may be deactivated until the next iteration.

[0156] Optionally, the non-invasive feedback is dynamically adapted during each iteration according to measurements made by the electrodes, optionally a real-time (or near real-time) current valence state measurement of the neural network. The non-invasive feedback may be dynamically adapted during each iteration, optionally in real-time (or near real-time), optionally continuously throughout each iteration.

[0157] The change in the current activity of the neural network of the subject may be an aggregate of multiple relatively small incremental changes in response to the dynamic adaption of the non- invasive feedback over each single iterations.

[0158] Both the non-invasive stimuli with valence and the non-invasive feedback are selected to be sufficiently fast to allow the shift from one state to the other (e.g., the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second predefined threshold) within each iteration, optionally in less than a minute. The external non- invasively applied stimuli with valence is presented at the beginning of each iteration. One or more external non-invasively applied stimuli may be presented, for example, a single image, a sequent of images, a combination of an image and audio, and the like. The presentation of the external non-invasively applied stimuli and the expected response should be fast, for example, within less than 5 seconds. For example, the non-invasive stimuli with valence may be presented for about 2 seconds. It may take about 2 seconds for the user to apprehend and / or understand the non-invasive stimuli. The non-invasive feedback may be for example, an analogue signal allowing immediate detection of change, so as to allow fast allow convergence.

[0159] For example, an image selected to trigger a strong emotional response, for example, trauma is presented on a display for about 2 seconds, during and after which the baseline state of activity is measured. A feedback device, for example, a task for the user to perform such as to align two crosses, is presented on a display, for example, designed to take the user less than one minute to perform. The current real-time or near real-time valence state of the neural network of the subject is monitored while the user is performing the task using measurements made by the electrodes. The non-invasive feedback, i.e., the task, is dynamically adapted during the iteration, optionally in real-time or near real-time according to the current real-time or near real-time valence state of the neural network, for example, the difficulty level is increased or decreased by changing the distance between the two crosses corresponding to a distance from the target state. The iteration may terminate when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold. In a next iteration, the same image or another image is selected to trigger the strong emotional response, and another baseline state of activity is measured. The feedback device is operated again, with the user instructed to perform the task. The non-invasive feedback, i.e., the task, is dynamically adapted during the next iteration, optionally in real-time or near real-time according to the current real-time or near real-time valence state of the neural network, in an attempt to incrementally shift the current activity of the neural network until the pre-defined state is met within the first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by the second predefined threshold, in less than about 1 minute. Multiple iterations may be performed. The change in the current activity of the neural network of the subject may be an aggregate of multiple relatively small incremental changes in response to the dynamic adaption of the non-invasive feedback over each single iterations.

[0160] The computing the baseline state and / or the monitoring of the current activity may be performed at a fast sampling rate of at least about 2000 Hertz (Hz), or at least about 1000 Hz, or at least about 5000 Hz, or other values. The use of implanted electrode(s) and fast sampling rate of neuronal activity facilitate fast response to the non-invasive feedback. The computing the baseline state and / or the monitoring the current activity may performed by measuring single spikes of neurons, which may enable more accurate measurement. The use of implanted electrode(s) enables the measuring of single spikes of neurons in deep brain structures.

[0161] Optionally, the iterations may begin with the presentation of the external non-invasive stimuli with valence (e.g., as described herein) and then followed by non-invasive feedback (e.g., as described herein) to help the brain make the shift from one state to the other, in an attempt to incrementally shift the current activity of the neural network until the pre-defined state is met within the first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by the second pre-defined threshold, in less than for example, about 1 minute or other value described herein. In order to meet the fast cycle both the external non-invasively applied stimuli and non-invasive feedback are selected and / or designed to be provided fast, as well as provoke a fast response in the brain. Optionally, both the external non-invasively applied stimuli with valence and the non- invasive feedback are designed to be sufficiently fast to allow the shift from one state to the other (e.g., when the current activity of the neural network meets the pre-defined state within a first predefined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold) within less than about two minutes, or about a minute, or about 30 seconds, or other values. The external non-invasively applied stimuli may be presented at the beginning of each iteration. The presentation of the external non-invasively applied stimuli is selected to be sufficiently fast and the expected response is predicted to be sufficiently fast, for example, within less than about 10 seconds, or about 5 seconds, or about 3 seconds, or other values. The non-invasive feedback may be implemented as an analogue signal, which may allow fast detection of change, which may enable fast convergence.

[0162] Two examples of the external non-invasively applied stimuli may be distinct in the time each provokes the effect: a horror picture provoking aversive feeling (fast) vs. listening to a provoking story (slow). The fast stimuli may be selected.

[0163] Two examples of the non-invasive feedback may be distinct in the time each provokes the effect: analogue change of tone of an auditory feedback or focus of a video clip (fast detection of change) vs. changing the difficulty of a complex assignment, for example, taking care of sitting down patients in a waiting room of a clinic and changing the difficulty level by increasing the pace of newcomers and / or adding noise (cries, complaints). The feedback for fast detection of change may be selected.

[0164] It is noted that iterations may be continued when the current activity of the neural network meets the pre-defined state, for enhancing and / or strengthening a new association between the predefined state and the feedback, and / or to make the association more stable over time after the treatment sessions are over.

[0165] Iterations may be terminated prior to reaching the pre-defined state, for example, if progress is not proceeding as expected, the patient becomes too tired, movement of the state is away (e.g., direction).

[0166] The iterations may be designed to incrementally modulate the current activity from the baseline towards the pre-defined state of activity. For example, during each iteration, the non- invasive feedback may be adapted to the current activity level of the neural network, for incrementally advancing towards the pre-defined state.

[0167] Upon termination of the iterations, when the subject that underwent treatment experiences the original, or similar to the original stimuli that invoke the memory, thought, emotion, and / or experience, an activity of the neural network that is associated with a dampened / opposite value may be generated. In turn, this may lead to a more adaptive moderate behavioral response, and / or alleviation of symptoms.

[0168] Treatment sessions may be repeated in order to achieve higher success and / or induce plasticity and / or re-organization of the neural network. The time intervals between sessions may be on the scale of hours or days. Later sessions may follow in intervals in the scale of days and / or following of weeks.

[0169] Optionally, one or more iterations are performed without using electrodes surgically implanted in the brain. For example, the surgically implanted electrodes are removed. The iterations may be performed using EEG signals obtained from EEG electrodes placed on the head of the subject. For example, during a maintenance phase. The following is an exemplary process for using EEG signals. EEG signals are measured by EEG electrodes simultaneously (or substantially simultaneously) with the baseline state and / or current activity and / or the pre-defined state computed from signals obtained from the electrode(s) surgically implanted within the region(s) of the brain. A mapping data structure that maps between the EEG signals and the baseline state and / or current activity and / or during the pre-defined state, is computed. The mapping data structure may be computed, for example, as a function, a machine learning model, and the like. In a subsequent phase, which may be the maintenance phase, the following features are implemented without using signals obtained from the electrode(s) surgically implanted within the region(s) of the brain. The surgically implanted electrodes may be removed. EEG signals are measured from EEG electrodes placed on the head of the subject. The mapping data structure is applied to the measured EEG signals for estimating the baseline state and / or current activity and / or the predefined state. One or more features described with reference to 208-214 are implemented based on the current activity and / or the baseline state and / or the pre-defined state estimated based on the EEG signals (rather than based on signals obtained from the surgically implanted electrodes).

[0170] Alternatively or additionally, one or more iterations are performed where measurements vary, such as measurements of different signals by electrodes implanted in the brain vary. The electrodes implanted in the brain may measure local field potentials (LFP) and / or spike measurements. The LFP may represent the summed electrical activity of a population of neurons, which may include their synaptic inputs and local processing. The spike measurements may represent action potential of one or more neurons. The baseline state and / or current activity and / or predefined state may be computed as a combination of LFP and spike measurements. Alternatively or additionally, the baseline state and / or current activity and / or predefined state may be computed based on LFP excluding spikes (e.g., only on LFP), or based on spikes excluding LFP (e.g., only spikes). The following is an exemplary process for enabling computing of the baseline state and / or current activity and / or predefined state, when spike measurements are lost and / or insufficient for use. The exemplary process may enable for a longer “effective period” in which neurofeedback may be applied, during times when spike measurements are not available and / or insufficient. Unlike spike measurements, LFP is expected to remain in good quality over longer periods of weeks and more.

[0171] The exemplary process may be as follows. LFP is simultaneously measured with spikes from signals obtained from electrode(s) surgically implanted within region(s) of the brain. The baseline state and / or current activity and / or pre-defined state are computed from a combination of the measured LFP and measured spikes. A mapping data structure that maps between the LFP signals and the spikes is computed. The mapping data structure may be computed, for example, as a function, a machine learning model, and the like. In response to lack of sufficient spike measurements (e.g., lost, weak, noisy), LFP is measured. The mapping data structure is applied to the measured LFP for estimating the spike measurement. The baseline state and / or current activity and / or pre-defined state may be computed based on the measured LFP excluding the spike measurement. The measured LFP may be used without the spike measurement based on the assumption that since it is known that the LFP was initially measured together with the spikes, the LFP provides a good indication of the baseline state and / or current activity and / or pre-defined state. Alternatively, the baseline state and / or current activity and / or pre-defined state is / are computed as a combination of the measured LFP and estimated spike measurements. One or more features described with reference to 208-214 may be implemented based on the current activity and / or the baseline state and / or the pre-defined state computed based on the combination include the estimated spike measurements (rather than based on measured spikes).

[0172] It is noted that the spike measurements may be by the implanted electrodes may refer to individual spikes of individual neurons. When spike measurements are not obtainable (e.g., lack of spikes, noisy spikes, weak spikes), the spike measurement estimated from the LFP measurements using the mapping data structure may refer to overall spiking activity of multiple neurons (e.g., in a region in proximity to the implanted electrodes) rather than an estimate of individual spikes of individual neurons. When spike measurements are not obtainable (e.g., cannot be measured reliably), estimates of individual spikes of individual neurons may be inaccurate to be useful for accurate computation of the baseline state and / or current activity and / or pre-defined state, whereas the estimates of the overall spiking activity of multiple neurons may be accurate enough to enable computation of the baseline state and / or current activity and / or pre-defined state.

[0173] Optionally, a transfer function is dynamically according to a state of available signals: LFP without significant spike measurement, spike measurement without significant LFP, and a combination of LFP and spike measurements. The transfer function maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state, as described herein.

[0174] The selection of the baseline and / or pre-defined states for neurofeedback and / or the respective transfer function may change according to the signal quality. For example, to include both spike and LFP measurements when the spike measurements are of good quality, and only LFP when spike measurements are of poor quality or vanished.

[0175] Referring now back to FIG. 3, an exemplary dataflow 302 of a brain-computer interface for neuromodulation of a neural network of a brain of a subject via a closed-loop is depicted.

[0176] At 304, a brain of a subject is depicted. Penetrating electrodes are positioned in one or multiple bran regions. In the example of 304, electrodes are shown in the Amygdala, the anterior- cingulate-cortex (ACC), and the substantia-innominata.

[0177] At 306, neural recordings are obtained from the penetrating electrodes.

[0178] At 308, the neural signals may be amplified into an acquisition system that collects the signals and extracts action potentials (spikes) from single neurons. Shown are plots of signals for different single neurons as a function of time, for each brain region.

[0179] At 310, a transfer-function may be computed based on the neural signals (e.g., as shown in 308).

[0180] At 312, external non-invasive feedback for the subject is determined, according to the transfer-function. In the depicted example, the feedback is an auditory tone whose frequency is determined by a transformation of the neural signals.

[0181] The transfer-function computes the current state of the neural activity, and determines the target goal of the neural state. The external feedback is correlated with the movement of the neural state from the original initial state to its target goal.

[0182] The simplest example of a transfer-function correlation of the tone frequency with the amount of activity in one single-neuron (measured as the number of spikes in a pre-determined time window). More complex forms of transfer-functions include, for example: the synchronous activity among several neurons, the synchronous activity between several brain regions, and the high-dimensional activity of all (or part of) the neurons recorded simultaneously.

[0183] It is noted that the external non-invasive feedback may be selected to be of different modalities.

[0184] At 314, the auditory tone feedback with selected frequency is played to the subject.

[0185] Features described with reference to 304-314 are iterated, optionally for dynamically adapting the external non-invasive feedback for obtaining a change in state, as described herein. Referring now back to FIG. 4, graphs 402 and 404 depict an example of how valence is represented in the activity of single-neurons in the amygdala. Graphs 406, 408, and 410 depict spikes elicited in response to a tone that was paired with an appetitive valence 412 or a different tone that was paired with an aversive valence 414. The two neurons show differential responses to both stimuli.

[0186] Referring now back to FIG. 5, graphs 502-512 are based on experiments performed by the Inventors. The first block including graphs 502 and 504 is used to characterize the neural representation to two different stimuli: one aversive 502 and one appetitive 504. The second block including graphs 506 and 508 denotes the brain-computer interface (BCI) loop, where the transferfunction 508 is defined according to the activity of one single-neuron, and the feedback (e.g., tone frequency) is correlated with the activity. In this block each trial starts at some frequency and the activity of this neuron is used to provide the auditory feedback until an upper threshold is reached and a positive outcome is delivered. Hence, if the neuron was responsive to aversive stimuli to being with in the first block, it is now used to bring a positive outcome, and hence changing its valence. In the final block including graphs 510 and 512, the neural and behavioral responses to the original aversive / appetitive stimuli after the BCI were measured.

[0187] Referring now back to FIG. 6, graphs 602-610 present results of experiments for neuromodulation of a neural network of a brain performed by the Inventors. Graph 602 indicates the activity of the neuron (averaged over all neurons recorded) during one trial of the BCI. Only in successful trials, the activity of the neuron was increased and reached the upper threshold that brings a positive outcome. Graph 604 indicates the activity of the neural when the transfer- function is defined so that the tone frequency is negatively correlated with the neuron activity. Graphs 606, 608, and 610 indicate the activity of the neuron (averaged over all neurons) over the whole session (60 BCI trials).

[0188] Referring now back to FIG. 7, graphs 702-706 present additional results of experiments for neuromodulation of a neural network of a brain performed by the Inventors. Graph 702 indicates the number of neurons that had a significant change in their activity after the BCI block (compared to before). Graphs 704 and 706 indicate the cumulative distribution of activity change in neurons (after minus before), showing that neurons increased their activity, and that responses to aversive stimuli was reduced.

[0189] Referring now back to FIG. 8, graphs 802-810 present yet additional results of experiments for neuromodulation of a neural network of a brain performed by the Inventors. Graphs 802-810 indicate behavioral changes after the BCI. Graphs 802-810 indicate that after the BCI, there were more preparatory (safety) responses to the stimuli, and less responses to the aversive stimulus itself. These behavioral responses were correlated with the amount of activity change in the neurons.

[0190] Referring now back to FIG. 9, graphs 902 present yet additional results of experiments for neuromodulation of a neural network of a brain performed by the Inventors. The transfer-function is defined based on the synchronous activity between two brain regions, the amygdala and the substantia-innominata. As a result, a successful BCI session result in the increase in synchrony and directionality of SI to amygdala activity. This is used to facilitate plasticity in representations of valence in amygdala neurons.

[0191] Referring now back to FIG. 10, which includes panels 1002-1006 depicting graphs and / or schematics indicating examples of more complex behaviors that are exhibited in anxiety / PTSD which may be treated using neuromodulation are described. The examples are of a generalization around an aversive stimulus, and exploration in dangerous environments. Panel 1002 depict neural activity prior to neuromodulation (e.g., baseline). Panel 1004 indicates examples of different transfer-functions which may be used: activity of one single-neuron 1004A, the synchronous activity among several neurons 1004B, the synchronous activity between several brain regions 1004C, the high-dimensional activity of all (or part of) the neurons recorded simultaneously 10004D. Panel 1006 indicates neural activity after neuromodulation as described herein.

[0192] The descriptions of the various embodiments of the present invention have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

[0193] It is expected that during the life of a patent maturing from this application many relevant approaches for computing valence will be developed and the scope of the term valence is intended to include all such new technologies a priori.

[0194] As used herein the term “about” refers to ± 10 %.

[0195] The terms "comprises", "comprising", "includes", "including", “having” and their conjugates mean "including but not limited to". This term encompasses the terms "consisting of" and "consisting essentially of".

[0196] The phrase "consisting essentially of" means that the composition or method may include additional ingredients and / or steps, but only if the additional ingredients and / or steps do not materially alter the basic and novel characteristics of the claimed composition or method. As used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a compound" or "at least one compound" may include a plurality of compounds, including mixtures thereof.

[0197] The word “exemplary” is used herein to mean “serving as an example, instance or illustration”. Any embodiment described as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments and / or to exclude the incorporation of features from other embodiments.

[0198] The word “optionally” is used herein to mean “is provided in some embodiments and not provided in other embodiments”. Any particular embodiment of the invention may include a plurality of “optional” features unless such features conflict.

[0199] Throughout this application, various embodiments of this invention may be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[0200] Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range. The phrases “ranging / ranges between” a first indicate number and a second indicate number and “ranging / ranges from” a first indicate number “to” a second indicate number are used herein interchangeably and are meant to include the first and second indicated numbers and all the fractional and integral numerals therebetween.

[0201] It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.

[0202] Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims.

[0203] It is the intent of the applicant(s) that all publications, patents and patent applications referred to in this specification are to be incorporated in their entirety by reference into the specification, as if each individual publication, patent or patent application was specifically and individually noted when referenced that it is to be incorporated herein by reference. In addition, citation or identification of any reference in this application shall not be construed as an admission that such reference is available as prior art to the present invention. To the extent that section headings are used, they should not be construed as necessarily limiting. In addition, any priority document(s) of this application is / are hereby incorporated herein by reference in its / their entirety.

Claims

WHAT IS CLAIMED IS:

1. A system for neuromodulation, comprising: at least one processor executing a code for: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence; in a plurality of iterations: generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity; monitoring the current activity of the neural network in response to the non- invasive feedback; and terminating the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

2. The system of claim 1, wherein the pre-defined state is defined as a state that corresponds to a second external non-invasively applied stimuli with neutral valence, or with valence opposite to the valence used to compute the baseline state.

3. The system of claim 1, wherein the pre-defined state is conditioned with a neutral valence or with a valence opposite to the valence used to compute the baseline state.

4. The system of claim 1, wherein the valence of the applied stimuli used to identify the neural network comprises a negative valence, and the non-invasive feedback is designed to shift the baseline state of activity to be closer to the pre-defined state identified and / or associated with a neutral valence and / or a less negative valence and / or more positive valence and / or less intense valence than the valence of the applied stimuli.

5. The system of claim 1, wherein the valence of the applied stimuli used to identify the neural network comprises a positive valence, and the non-invasive feedback is designed to shift the baseline state of activity to be closer to the pre-defined state identified and / or associated with a neutral valence and / or a more negative valence and / or less positive valence than the valence of the applied stimuli.

6. The system of claim 5, wherein the subject has an addiction associated with a positive valence that is triggered by the external non-invasively applied stimuli, and the subject is being treated for the addiction by generating the non-invasive feedback for changing from the baseline state to the pre-defined state of activity identified and / or associated with the neutral and / or the more negative valence and / or the less positive valence than the valence associated with the addiction.

7. The system of claim 1, wherein the external non-invasively applied stimuli and / or the non-invasive feedback applied during the iterations, are generated by at least one of: a visual presentation on a display, audio played over speakers, a tactile sensation generated by a haptic device, a decision-making task, and a motor skill task.

8. The system of claim 1, wherein the first pre-defined threshold and / or the second pre-defined threshold are selected for treatment of the subject for a mental disorder triggered by the external non-invasively applied stimulus.

9. The system of claim 8, wherein the mental disorder is selected from: posttraumatic stress disorder (PTSD), anxiety, phobia, obsessive compulsive disorder, and social disorder.

10. The system of claim 1, wherein the non-invasive feedback is generated according to a transfer function that maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state.

11. The system of claim 10, wherein the transfer function is implemented as a machine learning model, and the non-invasive feedback is generated by the machine learning model in response being fed the baseline state and the current activity of the neural network.

12. The system of claim 11 , wherein the machine learning model is trained on a training dataset of multiple records, wherein a record is generated based on a sample subject, the record including a sample baseline state of the neural network, a sample current activity of the neural network, a sample pre-defined state, and a ground truth of a target non-invasive feedback that when applied to the subject shifted the sample current activity from the sample baseline state towards the sample pre-defined state.

13. The system of claim 10, further comprising code for dynamically training the machine learning model during an initial phase in which records are created while the non-invasive feedback is varied and resulting current activity is computed, and the trained machine learning model is used during a subsequent phase in which the non-invasive feedback is generated for guiding the current activity toward the pre-defined state.

14. The system of claim 10, further comprising code for dynamically selecting the transfer function from a plurality of defined transfer functions according to at least one of: brain region, signal quality, and clinical protocol.

15. The system of claim 1, wherein the baseline state of activity and the current activity of the neural network is computed based on at least one of: single-neurons (1-many), local-field- potentials, synchronicity between activity in different cells within a region, synchronicity between different cells in different brain regions, as a multi-dimensional representation that is computed from the activity of many cells (2-many), an evoked local-field-potential including a spectral power and / or phase across several spectral frequencies, and synchronicity of power and / or phase across electrodes located in one or more brain regions16. The system of claim 1, wherein the current activity is computed from signals obtained from at least one electrode surgically implanted within the at least one region of the brain of the subject and configured for monitoring at least one of: action potentials of individual neurons, and local field potentials of at least one brain region.

17. The system of claim 16, wherein the electrode is designed to sense a neuronal population of less than about 1000 neurons, wherein the current activity of the neural network is computed for the neuronal population.

18. The system of claim 16, wherein the region of the brain in which the at least one electrode is implanted includes one or more of: an amygdala, insula, Anterior-cingulate-cortex (ACC), medial-prefrontal-cortex (mPFC), nucleus accumbens (NAc), hippocampus, para hippocampus, and temporal pole.

19. The system of claim 1, wherein in each iteration of the plurality of iterations the non-invasive feedback is adapted according to the current activity level of the neural network at a current iteration for incrementally advancing towards the pre-defined state.

20. The system of claim 1, further comprising code for: measuring EEG signals by EEG electrodes simultaneously with the baseline state and / or current activity and / or the pre-defined state computed from signals obtained from at least one electrode surgically implanted within the at least one region of the brain; computing a mapping data structure that maps between the EEG signals and the baseline state and / or current activity and / or during the pre-defined state; and in a subsequent phase, without using signals obtained from the at least one electrode surgically implanted within the at least one region of the brain: measuring EEG signals; and applying the mapping data structure to the measured EEG signals for estimating the baseline state and / or current activity and / or the pre-defined state, wherein the generating the non-invasive feedback is generated based on the current activity and / or the baseline state and / or the pre-defined state estimated based on the EEG signals.

21. The system of claim 1, further comprising the feedback device.

22. The system of claim 1, further comprising code for: measuring local field potentials (LFP) simultaneously with spikes from signals obtained from at least one electrode surgically implanted within the at least one region of the brain, wherein the baseline state and / or current activity and / or pre-defined state are computed from a combination of the measured LFP and measured spikes; computing a mapping data structure that maps between the LFP signals and the spikes; in response to lack of sufficient spike measurements, measuring LFP;applying the mapping data structure to the measured LFP for estimating the spike measurement; and computing the baseline state and / or current activity and / or pre-defined state as a combination of the measured LFP and / or estimated spike measurements.

23. The system of claim 22, further comprising code for dynamically adjusting a transfer function according to a state of available signals selected from: LFP without significant spike measurement, spike measurement without significant LFP, and a combination of LFP and spike measurements, wherein the transfer function maps between activity of the neural network and the non-invasive feedback predicted to shift the current activity of the neural network from the baseline state to the pre-defined state.

24. The system of claim 1, further comprising code for delivering electrical micro stimulation via one or more electrodes implanted in the brain, to facilitate the modulation of the current activity of the neural network to change from the baseline state towards the pre-defined state of activity.

25. A system for neuromodulation, comprising: at least one processor executing a code for: in a plurality of iterations: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence according to signals sensed by at least one electrode when implanted in the at least one region of the brain; generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity; monitoring the current activity of the neural network in response to the non- invasive feedback according to the signals sensed by the at least one electrode when implanted in the at least one region of the brain; dynamically adapting the non-invasive feedback during each respective iteration according to a real-time or near-real time current valance state measurement of the neural network; andterminating each respective iteration when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

26. The system of claim 25, wherein a single iteration is performed in about one minute or less.

27. The system of claim 25, wherein at least 10 iterations are performed during a single session.

28. The system of claim 25, wherein in a single iteration, the external non-invasively applied stimuli with valence is active for less than about 5 seconds, and then de-activated until a next iteration.

29. The system of claim 25, wherein the non-invasive feedback is continuously dynamically adapted throughout the iteration.

30. The system of claim 25, wherein the computing the baseline state and the monitoring the current activity are performed at a sampling rate higher than 2000 Hz.

31. The system of claim 25, wherein the computing the baseline state and the monitoring the current activity is performed by measuring single spikes of neurons.

32. The system of claim 25, wherein monitoring the current activity of the neural network in response to the non-invasive feedback comprises monitoring the real-time or near-real time current valance state measurement of the neural network.

33. A computer implemented method for neuromodulation, comprising: computing a baseline state of activity of the neural network in at least one region of a brain of a subject that corresponds to the external non-invasively applied stimuli with valence; in a plurality of iterations:generating instructions for operating a feedback device for generating a non-invasive feedback designed to incentivize the brain of the subject to modulate a current activity of the neural network to change from the baseline state towards a pre-defined state of activity; monitoring the current activity of the neural network in response to the non- invasive feedback; and terminating the iterations when the current activity of the neural network meets the pre-defined state within a first pre-defined threshold and / or when the current activity is statistically significantly different from the baseline state by a second pre-defined threshold.

Images