Method of smart feedback for stimulating the brain

Abstract

The present solution relates to a method of smart feedback for stimulating the morphofunctional structures of the brain. The method includes assessing the state of regions of the brain to identify regions exhibiting degenerative changes in the case of diseased subjects, or selecting regions which permit an improvement in the level of intellectual functions of the brain. A user is trained to use a brain-computer interface as an input device for performing a mental exercise in a virtual environment with the aid of feedback. The brain-computer interface additionally comprises a classifier that is trained on brain activity data pertaining to the user in order to classify mental patterns in the user's brain activity as commands for mental control. The parameters of the classifier of the brain-computer interface are adaptively tuned to alter the patterns of mental control and engage the morphofunctional centres of the brain in the control process, and the user and the classifier with the adaptively tuned parameters are adaptively trained. The classifier with the adaptively tuned parameters processes information from the signals of the brain according to the tuned parameters to involve the identified regions of the brain exhibiting degenerative changes or the selected regions of the brain for improvement of the desired intellectual functions of the brain. The result is an increase in the accuracy with which a brain-computer interface signal is classified.

Description

[0001] INTELLIGENT FEEDBACK METHOD FOR BRAIN STIMULATION

[0002] AREA OF TECHNOLOGY

[0003] This technical solution relates to the field of digital medicine, namely to methods of adaptive learning using intelligent feedback, associated with morpho-functional changes in the brain matter, allowing to increase the number of synapses, neurons, service cells and vessels in the brain matter, used for the purpose of enhancing cognitive abilities, rehabilitation and prevention of neurodegeneration, and / or allowing to increase the level of intellectual function of the brain, and / or develop and / or acquire intellectual abilities.

[0004] LEVEL OF TECHNOLOGY

[0005] The prior art includes information source AU 2018202658 B2, published on 25.06.2020, and disclosing a computer system and method for providing a virtual reality environment for improving memory. The solution disclosed in the information source includes the following steps. Using at least one virtual memory training module, including one or more memory training tasks, the user performs these tasks in a three-dimensional (3D) navigation environment. The navigation 3D environment is displayed by outputting to a display. Input data is received from an interactive navigation controller. In a particular implementation case, one or more scans of the user's brain activity are performed during the task, whereby the effectiveness of at least one virtual reality memory training module for a brain region can be measured.Based on the measured effectiveness of the previous training session, the virtual memory training module determines which virtual memory training modules need to be performed to improve memory.

[0006] The prior art includes information source WO 2020139108 A1, published on July 2, 2020, and disclosing a method for conducting cognitive research using a neuroimaging system and a feedback mechanism. The method comprises a system training stage and a testing stage. At the system training stage, the subject is shown mental state triggers, while simultaneously recording a multichannel EEG signal in real time, determining brain activity zones. Then, the neuroimaging system trigger generator is trained by converting signal features into new image samples. Then, at the testing stage, the subject independently imagines the mental state triggers he or she has seen before, seeing the target trigger forming on the screen.The technical result of the research is the definition of a set of mental states of the subject available for recognition based on the conduct of a cognitive test with the presentation of external triggers of the mental state. In addition, the result is the training of the subject to independently generate mental states based on the imagination of target triggers without their direct presentation and the optimization of the number and location of points for picking up electroencephalogram (EEG) signals from the surface of the subject's skull.

[0007] The difference between the proposed solution and the solutions known from the state of the art is that in the proposed solution, areas of the brain with degenerative changes of sick people are identified and as a result of adaptive adjustment of the parameters of the neurointerface classifier, the identified areas of the brain with degenerative changes or selected areas of the brain are activated in mental control to improve the desired intellectual functions when implementing mental control of movement commands in a virtual environment using a brain-computer interface, which leads to an increase in the number of synapses, neurons, service cells and vessels in the brain matter.

[0008] In addition, the difference between the proposed solution and the solutions known from the state of the art is that the proposed solution uses a simulation model of the brain with centers of interest (morpho-functional structures of the brain) to model their activation (singly and / or in combination) in the simulation and solves the inverse problem of forming the weights of the classifier of brain activity, according to the selected centers of interest of the brain, to teach the subject to activate certain areas of the brain and / or their combination without limitation on the morphology and function of the centers.

[0009] ESSENCE OF THE INVENTION

[0010] The technical problem, which the claimed technical solution is aimed at solving, is the morphofunctional change of the central nervous system of humans and animals with the purpose of increasing intellectual (cognitive, emotional) functions and / or acquiring new intellectual abilities and skills of sick and healthy people, neurorehabilitation and prevention of neurodegeneration, by increasing the number of synapses, increasing the size of neurons, service cells, improving blood supply (by increasing the size of vessels and the appearance of new vessels (for example, capillaries) in the brain tissue when implementing the proposed method.

[0011] The technical result achieved by solving the above technical problem is an increase in intellectual (cognitive, emotional) functions and / or the acquisition of new intellectual abilities and skills of sick and healthy people, neurorehabilitation and prevention of neurodegeneration.

[0012] The claimed technical result is achieved by implementing a method of intelligent feedback for the purpose of activating morphofunctional structures of the brain for the implementation of rehabilitation and / or prevention of neurodegeneration and / or improvement of intellectual functions and / or acquisition of new intellectual abilities and skills of sick or healthy people, comprising the following stages: assessing the morphofunctional state of areas of the brain to identify areas of the brain with degenerative changes in sick people, by comparing the morphofunctional characteristics of the user's brain in medical images and / or morphofunctional tests obtained as a result of examining the user, with normal values ​​​​of the corresponding social group or selecting areas of the brain to improve the desired intellectual functions of healthy people;training a user and a classifier as part of an invasive and / or non-invasive brain-computer interface, wherein the user is trained to use the invasive and / or non-invasive brain-computer interface as an input device for performing mental control in a virtual environment using feedback based on a signal to the user indicating the effectiveness of control using the brain-computer interface, and the classifier is trained based on brain signals to classify patterns of brain activity received from the user, with the aim of engaging areas of the brain when performing mental control in a virtual environment using the brain-computer interface;adaptively adjust the parameters of the brain-computer interface classifier in order to change the patterns of mental control so that, with further use by the user of the invasive and / or non-invasive brain-computer interface, the identified areas of the brain with degenerative changes or selected areas of the brain are involved in mental control to improve the desired intellectual functions when implementing mental control of commands in a virtual environment using the brain-computer interface;adaptive training of the user and the classifier with adaptively configured parameters is carried out as part of an invasive and / or non-invasive brain-computer interface, wherein the classifier with adaptively configured parameters processes information from brain signals according to the configured parameters at the previous stage to activate the identified areas of the brain with degenerative changes and / or selected areas of the brain to improve the desired intellectual functions when implementing mental control of movement commands in a virtual environment using the brain-computer interface in order to activate neuroplasticity in these areas of the brain.

[0013] In a particular embodiment of the proposed method, the morphofunctional characteristics of the brain are the morphological size of the brain regions and / or the intensity of signal absorption in a medical image and / or the location of the regions and / or the change in the electrode-magnetic activity of the brain, measured by functional tests and / or measurements of the saturation of the brain region with oxygenated blood, and / or the change in the metabolism of the brain region.

[0014] In another particular embodiment of the proposed method, the invasive brain-computer interface is a neuroimplant or vascular stent with the function of an electrode and / or sensor of a fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristic of brain activity.

[0015] In another particular embodiment of the proposed method, the non-invasive brain-computer interface is based on an electroencephalogram and / or a method for measuring fluorescent and / or luminescent and / or chemical and / or magnetic and / or electromagnetic characteristics of brain activity.

[0016] In another particular embodiment of the proposed method, activation of neuroplasticity in areas of the brain activates the processes of synaptogenesis and / or cytogenesis and / or regeneration and / or neuroplasticity of the central nervous system and peripheral nervous system.

[0017] DESCRIPTION OF DRAWINGS

[0018] The implementation of the invention will be described further in accordance with the attached drawings, which are presented to explain the essence of the invention and in no way limit the scope of the invention. The following drawings are attached to the application:

[0019] Fig. 1 - illustrates a system for implementing the proposed method of intelligent feedback for the purpose of stimulating morphofunctional structures of the brain for implementing rehabilitation and / or preventing neurodegeneration and / or improving intellectual functions of sick or healthy people. Fig. 2 - illustrates a sequence of operations of the method of intelligent feedback for the purpose of stimulating morphofunctional structures of the brain for implementing rehabilitation and / or preventing neurodegeneration and / or improving intellectual (cognitive, emotional) functions and / or acquiring new intellectual abilities and skills of sick or healthy people.

[0020] DETAILED DESCRIPTION OF THE INVENTION

[0021] In the following detailed description of the embodiment of the invention, numerous implementation details are provided in order to provide a clear understanding of the present invention. However, it will be obvious to one skilled in the art how the present invention may be used with or without these implementation details. In other instances, well-known methods, procedures, and components have not been described in detail in order not to obscure the features of the present invention.

[0022] In addition, it will be clear from the above description that the invention is not limited to the embodiment shown. Numerous possible modifications, changes, variations and substitutions, preserving the essence and form of the present invention, will be obvious to those skilled in the art.

[0023] The proposed method is implemented by means of a software and hardware complex for implementing intelligent feedback in order to stimulate the morphofunctional structures of the human or animal brain for the implementation of neurorehabilitation, prevention of neurodegeneration and / or improvement of intellectual (cognitive, emotional) functions and / or acquisition of new intellectual abilities and skills of sick or healthy people.

[0024] The hardware part of the complex includes the following devices and is shown in Fig. 1:

[0025] • A noninvasive brain-computer interface or invasive brain-computer interface (1), located on the head or introduced endovascularly or by other means - a specialized input device for obtaining electromagnetic or other information about the brain. The brain-computer interface includes an analog-to-digital converter, microcontrollers for pre-processing information (1.1), may contain software and hardware information converters, for example, in the form of modules for obtaining the frequency spectrum, spatial and temporal domain in the brain signal, GPU / TPU / CPU / neuromorphic processor (1.2) for training and operating a classifier, which can be built into the brain-computer interface or located separately, including location on the training server, a hardware module for storing and operating the trained / trained / configured classifier (1.3) (for example, a neural network, a mathematical algorithm without using a neural network, etc.), a hardware module for storing and operating the processor, which is a command converter for feedback (1.4), implemented on the microcontrollers of the brain-computer interface and / or located on the training server.

[0026] • sensors for receiving brain activity signals (2), for example, electrodes, accelerometers, temperature sensors, photodetectors, etc.),

[0027] • training server for training the classifier in training mode (not shown in Fig. 1);

[0028] • a device for implementing biological feedback, which can be represented by a screen / display (3.1) (for visual feedback) and / or an electrical stimulator of the sense organs (for example, it can be represented by a translingual electrical stimulator) (3.2), and / or a skin electrical stimulator (3.3)) and / or a device reproducing an audio signal (3.4) (used in the proposed method for creating interaction in a virtual environment). Biological feedback in Fig. 1 is indicated under the number 4;

[0029] • translingual electrode stimulator (3.2) - a specialized output device for implementing biofeedback and / or augmented reality (AR), which is a matrix of electrodes located on the tongue that receives a feedback signal.

[0030] Additional devices of the hardware part of the complex are:

[0031] • polymer glove - a specialized input-output device for receiving information from the user's muscles (myography) and the dynamic position of the hand in space (not shown in Fig. 1) and for receiving a feedback signal;

[0032] • transcranial stimulator or invasive endovascular stimulator - specialized devices for electromagnetic stimulation of the brain, used when it is necessary to enhance the effect of activation of a certain area of ​​the brain (not shown in Fig. 1);

[0033] The software part of the complex includes the following:

[0034] • software for assessing the morphofunctional state of brain regions for the purpose of segmenting brain structures and automatically detecting areas of brain regions with degenerative changes using medical images (CT, MRI, etc.);

[0035] • software for pre-processing the brain signal, including statistical methods, machine learning methods, spectral methods, frequency, time and spatial domain analysis methods. This software can be executed on brain-computer interface processors (including microcontrollers, single-board solutions, neuromorphic processors, etc.), desktop PCs and / or on a training server;

[0036] • software implementing a simulation model of the brain with centers of interest (morphofunctional structures of the brain) for modeling their activation (singly and / or in combination) in the simulation. This software builds a model of the brain with centers of interest and solves the problem of forming weights of the classifier of brain activity, in the selected centers of interest of the brain, for activation of the user's brain areas of interest, without restrictions on the morphology and function of the centers. This software is deployed on the training server;

[0037] • a classifier of brain activity patterns, as part of a brain-computer interface, with the aim of engaging areas of the brain in the implementation of mental control in a virtual environment using a brain-computer interface;

[0038] • software for selecting the mode (frequency, duration, strength) and mapping (location of the source relative to brain structures) of electromagnetic stimulation of the brain through endovascular electrodes and / or a transcranial stimulator, or other.

[0039] The non-invasive brain-computer interface contains a classifier for classifying patterns of brain activity and may be based on an electroencephalogram and / or a technique for measuring fluorescence or luminescence or chemical or magnetic or electromagnetic characteristics of brain activity.

[0040] The invasive brain-computer interface contains a classifier for classifying patterns of brain activity and is a neuroimplant or vascular stent with the function of an electrode and / or a sensor of fluorescent and / or luminescent and / or chemical and / or magnetic and / or electromagnetic characteristics of brain activity.

[0041] Mental control (input of commands, quasi-movements, etc.) is understood as input of commands through the brain-computer interface by means of the conscious or unconscious formation by the user of brain activity, which is recognized by the brain-computer interface classifier as a command and / or other service information (information about the user's emotional state, for example, but not limited to, a deprived state, drowsiness, agitation, stupor, stupor, alcohol or drug intoxication, etc.).

[0042] The virtual environment can be presented in the form of an interactive system, or in the form of a video game, or in the form of a metaverse, or in the form of another system.

[0043] The proposed method of intelligent feedback for the purpose of stimulating the morphofunctional structures of the brain for the implementation of rehabilitation and / or prevention of neurodegeneration and / or improvement of intellectual (cognitive, emotional) functions and / or acquisition of new intellectual abilities and skills of sick or healthy people includes the following stages.

[0044] Stage 1. Using morphofunctional recording methods (e.g. fMRI, EEG, MRI, CT, magnetometry, etc.), spatial, temporal and frequency patterns of the user's brain activity are recorded when performing certain cognitive tasks (speech tests, counting, listening to music, etc.); The obtained patterns of brain activity are used to identify the centers and areas of the brain that are activated when performing the above-mentioned cognitive tasks. The corresponding areas of the brain are then used for their active involvement in targeted neuroplasticity and for comparison with the patterns of brain activity when performing the same tasks and / or at rest after the neuroplasticity activation procedure (steps 200-201 of Fig. 2).

[0045] Next (step 202) a three-dimensional morphofunctional map of the brain is constructed using mapping (representation of the brain in the form of, for example, a graph or connectome), analysis of brain signal data, analysis of the architecture of neural connections of the brain and densely connected centers of the brain, analysis of morphofunctional structures. The main morphofunctional centers of the brain (centers of interest) are identified. At this stage, the brain centers and other structures of the user are identified that have morphofunctional differences with the average population values ​​of a given social group.

[0046] For example, to construct a three-dimensional morphofunctional map of the brain, methods are used for constructing a connectome (brain graph) and solving the inverse problem of obtaining a morphofunctional map of the brain by simulating the activity of morphofunctional centers of the brain and / or simulating the activity of modeled morphofunctional centers of the brain.

[0047] First, an adjacency matrix is ​​constructed, which can be two-dimensional (for static analysis), three-dimensional (for the analysis of functional connections, for example, in time). Additional data can increase the matrix dimension. For calculating adjacency matrices or functional connectivity matrices, for example, mutual correlation, average phase coherence, mutual information, phase synchronization value, and others can be used.

[0048] Secondly, the resulting adjacency matrix is ​​thinned, for example, using simulation analysis of the data and preserving connections with statistically significant correlation values. As a result of this step, a matrix containing only significant connections is obtained. To assess the significance of the obtained connections, one can additionally use synthetic data analysis and set the significance level, allowing users to preserve only statistically significant connections. Thirdly, a graph is constructed based on the resulting matrix, where the vertices are areas of interest (signal sources in the brain, represented in three-dimensional space), and the edges correspond to functional connectivity. Analysis of structural MRI data is carried out, for example, using FreeSurfer, using the corresponding anatomical atlases of brain segmentation (for example, Desikan 1 and Destriux 2) and allows to solve the problem of reconstructing the source (cortical or subcortical) of electromagnetic activity of the brain from the projective signal on the surface of the head, in accordance with the location of the channels for collecting information about electromagnetic and / or chemical activity of the brain (for example, EEG channels). This problem, for example, can be solved by various nonparametric methods for estimating the minimum norm and their generalization, such as: the minimum norm (MN), weighted minimum norm (WMN), low resolution electromagnetic tomography (LORETA, and variations: VARETA, bc-VARETA, sLORETA, eLORETA, swLORETA, Shrinking LORETA FOCUSS (SLF) and others) 3, Backus-Gilbert, Weighted Resolution Optimization (WROP), S-MAP, ST-MAP, Backus-Gilbert, LAURA, SSLOFO and ALF. This problem, for example, can be solved by various parametric methods, such as beamforming methods, Krylov-type methods, projection methods for solving linear equations, source localization methods (BESA), other subspace methods, MUSIC and its derivatives 4 , 5 , 6 , FINES, simulated annealing algorithms, etc. In addition, this problem can be solved by artificial intelligence methods (for example, neural networks).

[0049] Next, a simulation of the activity of the main morphofunctional centers of the brain, obtained as a result of constructing a three-dimensional morphofunctional map of the brain, is carried out at step 202. Using the reconstructed sources of electromagnetic activity of the brain, a simulation of the activation of these sources is performed and projective signal data is obtained on the channels for collecting information on electromagnetic and / or other (e.g., chemical, photoactivity) activity of the brain (e.g., EEG channels).

[0050] 1 Desikan RS, Segonne F, Fischl B, Quinn VT, Dickerson BC, Blacker D, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;

[0051] 31(3):968-80. PMID: 16530430

[0052] 2Destrieux C, Fischl B, Dale A, Halgren E. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. NeuroImage. 2010; 53(1 ):1-15. doi: 10.1016 / j. neuroimage.2010. 06.010 PMID: 20547229

[0053] 3 Baillet S: Toward Functional Brain Imaging of Cortical Elec- trophysiology Markovian Models for Magneto and Electroen- cephalogram Source Estimation and Experimental Assessments. In Ph.D thesis University of Paris- ParisXI, Orsay, France; 1998.

[0054] 4 Mosher JC, Leahy RM: Recursive MUSIC: A Framework for EEG and MEG Source Localization. IEEE Transactions on Biomed- ical Engineering 1998, 45(11 ): 1342-1354.

[0055] 5 Mosher JC, Leahy RM: Source Localization Using Recursively Applied and Projected (RAP) MUSIC. IEEE Transactions on Sig- nal Processing 1999, 47(2):332-340.

[0056] 6Ermer JJ, Mosher JC, Huang M, Leahy RM: Paired MEG Data Set Source Localization Using Recursively Applied and Projected (RAP) MUSIC. IEEE Transactions on Biomedical Engineering 2000, 47(9): 1248–1260. The resulting data are stored as a set of brain activity values ​​corresponding to channels of electromagnetic and / or other (e.g., chemical, photoactivity) brain activity collection (e.g., EEG channels).

[0057] At stages 300-303, an assessment of the morphofunctional state of brain regions is carried out to identify brain regions with degenerative changes in sick people, by comparing the morphofunctional characteristics of the user's brain in medical images (fMRI, EEG, MEG, etc.) and morphofunctional texts obtained as a result of examining the user, with normal values ​​of the corresponding social group using, for example, FreeSurfer, using the corresponding anatomical atlases of brain segmentation (for example, Desikan 2006) and Destriux 2010).

[0058] To assess the morphofunctional state of brain regions, medical images and morphofunctional test data (MRI, CT, fMRI, EEG, X-ray, etc.) are used. In a resting state, the morphofunctional state of the user's brain is recorded, for example, using MRI and / or CT. The obtained brain image data is segmented automatically into structures and parts: white matter, gray matter, vessels, brain nuclei, pathological structures, connective and bone tissue structures, with subsequent measurement of these parts and structures using specialized software (for example, using FreeSurfer (https: / / surfer.nmr.mgh.harvard.edu / ), using the appropriate anatomical atlases of brain segmentation (for example, Desikan 2006) and Destriux 2010).

[0059] Segmentation is performed automatically, for example, using artificial intelligence, and the following algorithms can be applied: clustering algorithms, neural networks of various architectures for solving segmentation and automatic labeling problems, mathematical segmentation algorithms without using artificial intelligence, etc. Segmentation of brain structures can be adapted to modern methods when new capabilities of science and technology appear. In training artificial intelligence, both open CT and / or MRI data and / or other brain images, as well as newly collected data, can be used, while segmentation of brain structures can be performed automatically for the subject to whom the technique is applied, based on configured (trained and / or calibrated) automatic segmentation methods (for example, the technique described at the link can be used: https: / / www.frontiersin.org / articles / 10.3389 / fnana.2022.894606 / full).

[0060] The state of the brain structures is assessed by comparing such parameters as: the morphological size of brain regions and / or the intensity of signal absorption in a medical image and / or the location of regions and / or changes in the electromagnetic activity of the brain measured by morphofunctional tests and / or measurements of the saturation of a brain region with oxygenated blood and / or changes in the metabolism of a brain region with normal population values ​​of the corresponding social group. In this case, normal population values ​​are calculated based on open data from medical images and their analysis (e.g., Desikan 7 ) and Destriux 8 ). Detection of statistically significant deviations from the norm, indicating degenerative and / or other (for example, hypo- and hyperplasia) changes, is noted as the main stimulation targets (step 302 in Fig. 2).

[0061] Next, a list of brain regions that have degenerative and / or pathological changes is formed (step 303 in Fig. 2).

[0062] Next, a simulation is performed (as in step 203) of the activity of the brain areas that have degenerative and / or pathological changes (obtained in step 303), and / or areas for improving the desired intellectual functions of healthy people. In this case, these areas may not be detected in step 202. For example, a simulation is performed of the activity of the speech center, which has signs of neurodegeneration due to a stroke and is not activated when performing speech tasks, for a given user.

[0063] In another variant of the proposed method, when a healthy user seeks to improve the desired intellectual functions, and for this purpose, areas of the brain are selected. In this case, these areas may not be detected at step 202 as pathological. For example, a simulation of the activity of the speech center is carried out, which does not have signs of neurodegeneration in this user. In this case, patterns of brain activity obtained during the performance of the corresponding tasks (from the cognitive sphere selected for improvement) and / or at rest at the preparatory stage are taken into account.

[0064] As a result, maps are constructed and patterns of brain activity are obtained (step 205 in Fig. 2), characteristic of activation in the simulation. For each morphofunctional method of brain research, these maps and patterns have their own characteristics. For example, for EEG, the result of this step is obtaining in the simulation a projective signal of brain activity on the electrodes corresponding to the activated (in the simulation) areas. The obtained data are saved as a set of values ​​of electromagnetic activity of the brain, corresponding to the channels for collecting information on electromagnetic activity of the brain (for example, EEG channels) after simulating the activity of those centers that are planned for training / activation by the methods described in the patent using the brain-computer interface.

[0065] 7Desikan RS, Segonne F, Fischl В, Quinn ВТ, Dickerson ВС, Blacker D, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;

[0066] 31 (3):968-80. PMID: 16530430

[0067] 8Destrieux C, Fischl B, Dale A, Halgren E. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. NeuroImage. 2010; 53(1 ):1-15. doi: 10.1016 / j. neuroimage.2010. 06.010 PMID: 20547229 The second stage is convergent training of the user and the classifier as part of an invasive and / or non-invasive brain-computer interface. The user is trained to use the invasive and / or non-invasive brain-computer interface as an input device to perform mental control in a virtual environment using feedback based on a signal to the user indicating the effectiveness of control using the brain-computer interface. Also, during the training and use of the brain-computer interface, the user can also use headphones and / or a polymer glove to receive information from the user's muscles and the dynamic position of the hand in space.The virtual environment of use can be selected individually for each user, taking into account the daily routine, fatigue of the individual and personal characteristics. For example, a video game environment can be used, in which the user controls (plays) through a brain-computer interface. For example, sensors are attached to the user to obtain an electroencephalogram or a fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristic of brain activity and the user must perform mental control in the game, which is projected on the display.In another embodiment, a neuroimplant or vascular stent with the function of an electrode or sensor of fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristics of brain activity is inserted into the user, and a specialized device for immersion in a virtual environment (for example, VR glasses) is put on the user, and the user must perform mental control in the metaverse projected in the VR glasses. In another embodiment, sensors for obtaining an electroencephalogram or fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristics of brain activity are attached to the user, and a neuroimplant or vascular stent with the function of an electrode or sensor of fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristics of brain activity is inserted, and the user must perform mental control in the interactive system that is projected onto the display.

[0068] The classifier of brain activity patterns is trained based on brain signals received from the user, with the purpose of activating brain areas during mental control via the brain-computer interface in a virtual and / or auditory and / or other sensory (translingual electrode stimulator) environment. The proposed solution may use a classifier based on supervised or unsupervised classification, a classifier based on deep learning, and others. The training of the classifier (e.g., a k-means classifier or a neural network) continues until the accuracy of classifying the mental control commands entered by the user is not lower than the specified accuracy threshold. The accuracy threshold specified for the classifier should exceed the accuracy level of classification (e.g., k commands) of a random classifier.

[0069] Example 1. Training of a k-means classifier based on magnetoencephalography (MEG) data is performed in real time by feeding the k-means brain data represented as an n-dimensional vector to the algorithm. The number of classes for the classifier is specified in advance based on the set of commands selected for control using the brain-computer interface. Receiving new data from the user continues until the accuracy of classification of the mental control commands entered by the user is not lower than the specified accuracy threshold (e.g., 95%). The accuracy threshold specified for the classifier must exceed the classification accuracy level (e.g., k commands) of a random classifier.

[0070] Example 2. Training of a neural network-based classifier on EEG data from a brain-computer interface occurs as follows. First, the data is preprocessed so that an n-dimensional vector is formed from the multichannel EEG signal, consisting of: a set of frequencies; parameters characterizing the location of the electrodes on the head; time parameters characterizing the beginning and end of the quasi-movements; statistical data and other patterns. Second, a quasi-movement label is added that corresponds to this vector. Third, the obtained vectors are fed to the input of the neural network and the neural network is trained on the formed labeled sample using the back propagation method. The result of the trained neural network is a classifier that classifies the signal from the brain-computer interface with a given accuracy in relation to the correspondence of this signal to one of the quasi-movements.

[0071] The moment of user learning to use a new input device and / or a new environment (e.g., a gaming environment) stimulates synaptogenesis, neurogenesis, and other neuroplasticity processes. After achieving a given classification accuracy, the processes of neuroplasticity, synaptogenesis, and neurogenesis slow down.

[0072] At the third stage, adaptive adjustment of the classifier is carried out by adjusting the parameters of the classifier and / or the weights of the neural network, with the aim of changing the patterns of mental control so that with further use by the user of the invasive and / or non-invasive brain-computer interface, the identified areas of the brain with degenerative changes or selected areas of the brain are involved in mental control to improve the desired intellectual functions when implementing mental control of movement commands in a virtual environment using the brain-computer interface.

[0073] Adaptive adjustment of classifier parameters or redistribution of neural network weights consists in increasing the contribution of the brain activity signal received from certain brain areas (with degenerative changes or brain areas for improving the desired intellectual functions) for classifying brain activity patterns (frequency, temporal, spatial and other features of electrode-magnetic or other brain activity) coming from the user to the input of the classifier as part of the brain-computer interface for mental control in a virtual environment using the brain-computer interface. For example, the neural network weights are adaptively adjusted in such a way that the neural network increases the contribution of information coming to the k-classifier from electrodes that collect information from brain areas selected for activating neuroplasticity in these areas (with degenerative changes or brain areas for improving the desired intellectual functions).

[0074] The adjustment of the classifier parameters can be done not all at once, but gradually - after each session, the classifier redistributes the weights so that the contribution of the selected brain areas (with degenerative changes or brain areas for improving the desired intellectual functions) increases gradually.

[0075] At the fourth stage, adaptive convergent training of the user and the brain-computer interface classifier with adaptively configured parameters is carried out.

[0076] The user is already partially trained to control the brain-computer interface using mental control. The classifier with adaptively configured parameters processes information from brain signals, according to the configured parameters at the previous stage, to engage the identified areas of the brain with degenerative changes and / or selected areas of the brain to improve the desired intellectual functions when implementing mental control of movement commands in a virtual environment using the brain-computer interface in order to activate neuroplasticity in these areas of the brain. That is, the classifier begins to "require" the user to involve (activate) in the user's usual control, using the brain-computer interface, areas of the brain with degenerative changes and / or areas of the brain to improve intellectual (cognitive, emotional) functions and / or acquire new intellectual abilities and skills.

[0077] The result of this stage is the activation of neuroplasticity in the brain areas with degenerative changes or in the brain areas for improving intellectual (cognitive, emotional) functions and / or acquiring new intellectual abilities and skills, thereby activating the processes of synaptogenesis, cytogenesis, regeneration, neuroplasticity of the brain and peripheral nervous system. Activation of neuroplasticity is adaptive and targeted. Before the training session (using the brain-computer interface for mental control of movement commands in virtual reality), it is possible to conduct electromagnetic stimulation of the brain areas with degenerative changes or brain areas selected to improve intellectual (cognitive, emotional) functions and / or acquire new intellectual abilities and skills, using a transcranial stimulator or an invasive endovascular stimulator.

[0078] Transcranial stimulation performed before the session changes the metabolism and activates the activity of neurons and other brain cells in these areas. Such a change can stimulate the effect of learning to work with an adapted brain-computer interface and can be subjectively perceived by the user as the activity of this brain area. Brain areas, after stimulation, can be more quickly involved in learning to control the brain-computer interface with the already configured classifier parameters (adapted brain-computer interface). In addition, electromagnetic stimulation can enhance the effect of activating brain areas with degenerative changes or brain areas for improving intellectual (cognitive, emotional) functions and / or acquiring new intellectual abilities and skills. Informally, such stimulation “prompts” the individual which areas need to be involved to control the brain-computer interface with the already configured classifier parameters.

[0079] At the fifth stage, after completing all the previous stages of the proposed method, verification of neuroplasticity activation is carried out by obtaining medical images of the subject's brain after completing all the stages of the method.

[0080] Verification is the process of comparing MRI and / or CT and / or fMRI and / or PET and / or other methods of visualizing the morphofunctional properties of the brain, images of the user's brain and the results of psycho-functional tests obtained before and after 3 months of use (other time limits for using the proposed method may also be applied), the described method. When verifying MRI and / or CT and other images, the thickness of the cerebral cortex, the volume of brain structures, the vascular pattern of the brain (provided, for example, by a 3D MRI tomograph) are compared, the appearance of new neurons and service cells at high resolution is recorded (provided, for example, by a >6.9T tomograph), the number of neurons in the cerebral cortex and in other morphofunctional structures is compared (provided, for example, by a >6.9T MRI tomograph).

[0081] The method is confirmed by the following examples of application of the technique.

[0082] Example 1. A healthy 33-year-old man complained of weakness and drowsiness after having had COVID-19. A therapist, neurologist, and psychiatrist did not reveal any pathology. The man is healthy and does not take any medications or other therapy. Biochemical blood tests, general urine and blood tests are normal. Upon examination: the condition is satisfactory. Consciousness is clear. MRI of the brain - no pathological formations were revealed. The results of the questionnaire and cognitive testing are normal.

[0083] MRI (CT) was performed by conducting a T1- and T2-weighted gradient echo pulse sequence using the MPRAGE, FLASH, T2-SPACE protocols 9. During the study, the man was at rest, in a supine position with his eyes closed. Brain structures were segmented using Advanced Normalization Tools (ANTs, https: / / github.com / ANTsX / ANTs), FreeSurfer (https: / / surfer.nmr.mgh.harvard.edu / ) and using automatic labeling of our own multilayer recurrent neural network. The cortex thickness obtained using Freesurfer for each area of ​​interest was normalized to the average thickness of the hemisphere of the subject to account for individual differences in the overall average cortex thickness, which could affect cross-comparisons of participants. Analysis of morphological brain structures did not reveal any pathology. The cortex thickness was recorded over the entire surface of the brain. For example, in the speech center on the left, it averaged 3.1 mm.

[0084] fMRI was performed: the obtained images were combined with the anatomical structures of the brain using a T1-weighted gradient echo pulse sequence with magnetization preparation and rapid data collection. During the study, the man was at rest, in a supine position with his eyes closed. fMRI was performed with cognitive load (speech task, counting task). Analysis of functional connection maps showed a decrease in connectivity between brain structures, lateralization of the hemispheres when performing cognitive tasks, and an expansion of the functional center when performing tasks. EEG was performed with cognitive load (speech task, counting task). Analysis of electrode-magnetic activity revealed an increase in the proportion of high frequencies (> 50 Hz), lateralization of the hemispheres under cognitive load, a decrease in the contribution of the alpha rhythm to the frequency pattern of the brain, and hyperactivity in the structures of the speech center.Psycho-functional tests were conducted: cognitive tests, spatial orientation tests, short-term memory tests, counting tests, and psycho-emotional state.

[0085] After using the proposed method, psycho-functional tests and MRI, fMRI and fMRI and EEG studies with cognitive load were repeated for 3 months (3 days a week, session duration 3 hours) with an emphasis on targeted activation of the neuroplasticity of the speech center.

[0086] 9 https: / / surfer.nmr.mgh. harvard. edu / fswiki / FreeSurferWiki?action=AttachFile&do=get&target=FreeSurfer_Suggested

[0087] _Morphometry_Protocols.pdf According to MRI data, a statistically significant increase in the thickness of the cortex in the speech center (on average 0.4 mm), the appearance of new capillary-type vessels in the Wernicke area and the auditory cortex were revealed. According to fMRI data, a statistically significant decrease in the lateralization of the hemispheres and an increase in connectivity between brain structures were revealed. According to EEG analysis, a decrease in high frequencies in the frequency pattern of the brain, an increase in the contribution of the alpha and beta rhythms in the frequency pattern of the brain were revealed. Cognitive tests showed a statistically significant increase in the results of the speech test, an increase in short-term memory and a decrease in stress levels.

[0088] Example 2. A healthy 37-year-old woman complained of rapid fatigue in a stressful situation at work. A therapist, neurologist, gynecologist, endocrinologist and psychiatrist did not reveal any pathology. The woman is healthy and does not take any medication or other therapy. Biochemical blood tests, general urine and blood tests are normal. Upon examination: condition is satisfactory. Consciousness is clear. MRI of the brain - no pathological formations were revealed. The results of the questionnaire and cognitive testing are normal.

[0089] MRI was performed by conducting a T1- and T2-weighted gradient echo pulse sequence with magnetization preparation and fast data collection, using the MPRAGE, FLASH, T2-SPACE protocols. During the study, the woman was at rest, in a supine position with her eyes closed. Brain structures were segmented using Advanced Normalization Tools (ANTs, https: / / github.com / ANTsX / ANTs), FreeSurfer (https: / / surfer.nmr.mgh.harvard.edu / ) and using automatic labeling of our own multilayer neural network of the transformer class. The cortex thickness obtained with Freesurfer for each area of ​​interest was normalized to the average thickness of the hemisphere of a given subject to account for individual differences in the overall average cortex thickness, which could affect cross-comparisons of participants. Analysis of morphological brain structures did not reveal any pathology. The cortex thickness is fixed over the entire surface of the brain.For example, in the speech center area on the left, it averaged 2.8 mm. Analysis of the morphological structures of the brain did not reveal any pathology. The thickness of the cortex was recorded over the entire surface of the brain.

[0090] fMRI was performed: the obtained images were combined with the anatomical structures of the brain using a T1-weighted gradient echo pulse sequence with magnetization preparation and rapid data collection. During the study, the woman was at rest, in a supine position with her eyes closed. fMRI with cognitive load was performed (speech task, counting task). Analysis of functional connection maps showed a decrease in connectivity between brain structures, expansion of the functional center when performing speech tasks. EEG with cognitive load was performed (speech task, counting task). Analysis of electrode-magnetic activity revealed a significant increase in the proportion of high frequencies (> 50 Hz), a decrease in the contribution of the alpha rhythm to the frequency pattern of the brain. Psychofunctional tests were performed: cognitive tests, spatial orientation tests, short-term memory tests, counting test, psychoemotional state.

[0091] After using the proposed method, for 3 months (3 days a week, session duration 3 hours) with an emphasis on targeted activation of neuroplasticity of the speech center and areas of the brain activated when performing counting tasks, psycho-functional tests and MRI, fMRI, fMRI and EEG studies with cognitive load were repeated.

[0092] MRI data revealed a statistically significant increase in the thickness of the cortex in the speech center (by an average of 0.5 mm) and areas of the brain activated when performing counting tasks (by an average of 0.2 mm). fMRI data revealed a statistically significant increase in connectivity between brain structures. EEG analysis revealed a decrease in high frequencies in the frequency pattern of the brain, an increase in the contribution of alpha and beta rhythms in the frequency pattern of the brain. Cognitive tests showed a statistically significant increase in results in the speech test, the counting test, and a decrease in stress levels.

[0093] Example 3. A 30-year-old man with acquired total blindness (damage to the outer eyes as a result of an explosion) sought cognitive training to improve the functionality of the occipital lobes of the brain with subsequent use of a translingual matrix for alternative orientation in space. A therapist, neurologist and psychiatrist did not reveal any pathology other than visual. The man is healthy, does not take medication or other therapy. Biochemical blood tests, general urine and blood tests are normal. On examination: the condition is satisfactory. Consciousness is clear. MRI of the brain - pathological formations were not revealed. The results of the questionnaire and cognitive testing are normal.

[0094] MRI (CT) was performed by conducting a pulse sequence of T1- and T2-weighted gradient echo with magnetization preparation and rapid data collection, using the MPRAGE, FLASH, T2-SPACE protocols. During the examination, the man was at rest, in a lying position.

[0095] Brain structure segmentation was performed using Advanced Normalization Tools (ANTs, https: / / github.com / ANTsX / ANTs), FreeSurfer (https: / / surfer.nmr.mgh.harvard.edu / ) and automatic labeling of our own multilayer recurrent neural network. Cortical thickness obtained by Freesurfer for each ROI was normalized to the average thickness of the hemisphere of the subject to account for individual differences in overall average cortical thickness, which could affect cross-comparisons of participants. Analysis of morphological brain structures showed a decrease in gray matter thickness in the occipital regions of the brain, bilaterally. Cortical thickness in the occipital lobes did not exceed 1.1 mm, which is two times lower than the age norm.

[0096] fMRI was performed: the obtained images were combined with the anatomical structures of the brain obtained during segmentation of the structures. During the study, the man was at rest, in a lying position. fMRI with cognitive load was performed (navigation in a two-dimensional labyrinth using a translingual matrix as an input-output device). Analysis of functional connection maps showed a decrease in connectivity in the occipital areas and between brain structures. EEG with cognitive load was performed (speech task, counting task). Analysis of electrode-magnetic activity revealed a slight increase in the proportion of high frequencies (> 50 Hz), a decrease in the contribution of the alpha rhythm to the frequency pattern of the brain. Psychofunctional tests were performed: cognitive tests, spatial orientation tests, short-term memory tests, counting test, psychoemotional state.

[0097] After using the proposed method, psychofunctional tests and MRI, fMRI, fMRI and EEG studies with cognitive load were repeated over the course of 3 months (3 days a week, session duration 3 hours) with an emphasis on targeted activation of neuroplasticity of the visual centers of the occipital lobes.

[0098] MRI data revealed a statistically significant increase in the thickness of the cortex in the occipital lobes (by 0.7 mm on average). fMRI data revealed a statistically significant increase in connectivity between brain structures. EEG analysis revealed a decrease in high frequencies in the frequency pattern of the brain, an increase in the contribution of the alpha and beta rhythms in the frequency pattern of the brain. Cognitive tests showed a statistically significant increase in speech test results, an increase in short-term memory, a decrease in stress levels, and an increase in the accuracy of control of the translingual matrix as an input-output device for navigation in a two-dimensional labyrinth.

[0099] Example 4: A 27-year-old woman sought cognitive training to develop her ability to synesthesia, a property of the nervous system in which stimulation in one sensory system leads to a response in another sensory system. The woman wanted to learn to experience visual flashes when listening to music. A therapist, neurologist, gynecologist, endocrinologist, and psychiatrist found no pathology. The woman is healthy and does not take medication or other therapy. Biochemical blood tests, general urine and blood tests are normal. Upon examination: condition is satisfactory. Consciousness is clear. MRI of the brain - pathological formations were not revealed. The results of the questionnaire and cognitive testing are normal.

[0100] MRI (ST) was performed using a T1- and T2-weighted gradient echo pulse sequence with magnetization preparation and fast data collection, using the MPRAGE, FLASH, T2-SPACE protocols. During the study, the woman was at rest, in a supine position with her eyes closed. Brain structures were segmented using Advanced Normalization Tools (ANTs, https: / / github.com / ANTsX / ANTs), FreeSurfer (https: / / surfer.nmr.mgh.harvard.edu / ) and using automatic labeling of our own multilayer neural network of the transformer class. The cortex thickness obtained using Freesurfer for each area of ​​interest was normalized to the average thickness of the hemisphere of the given subject to account for individual differences in the overall average cortex thickness, which could affect cross-comparisons of participants. Analysis of morphological brain structures did not reveal any pathology. The cortex thickness was fixed over the entire surface of the brain.In the visual center area it averaged 2.8 mm, in the auditory center area it averaged 3 mm. Analysis of the morphological structures of the brain did not reveal any pathology. The thickness of the cortex was recorded over the entire surface of the brain.

[0101] An fMRI was performed: the obtained images were combined with the anatomical structures of the brain by conducting a pulse sequence. During the study, the woman was at rest, lying down with her eyes closed.

[0102] fMRI with cognitive load (speech task, counting task, color card analysis task, musical ear test) was performed. Analysis of functional connection maps showed no deviation from the norm. Morphological and functional patterns of brain activity during sound and visual stimulation were revealed.

[0103] An EEG with cognitive load was performed (speech task, counting task, color card analysis task, musical ear test). Analysis of electrode-magnetic activity did not reveal any deviations from the norm. Psycho-functional tests were performed: cognitive tests, spatial orientation tests, short-term memory capacity tests, counting test, musical ear test, psycho-emotional state.

[0104] After using the proposed method, for 3 months (3 days a week, session duration 3 hours) with an emphasis on the simultaneous activation of neuroplasticity of the auditory center and visual cortex when listening to music. Psycho-functional tests and MRI, fMRI, fMRI and EEG studies with cognitive load were repeated.

[0105] According to the survey data, a statistically significant occurrence of visual synesthesia was revealed when listening to loud music. Morphofunctional tests showed an increase in the thickness of the cortex in the auditory center (by an average of 0.5 mm) and occipital lobes of the brain (by an average of 0.6 mm). According to fMRI data, a statistically significant increase in connectivity between brain structures was revealed. According to EEG analysis, an increase in the contribution of the alpha and beta rhythms to the frequency pattern of the brain was revealed. Cognitive tests showed a statistically significant increase in the results of the musical ear test and the involvement of the visual cortex activity when listening to loud music.

[0106] The computing system, providing the data processing necessary for the implementation of the claimed solution, generally contains the following components: one or more processors, at least one memory, a data storage facility, input / output interfaces, an input facility, and network interaction facilities. When executing machine-readable commands contained in the RAM, the device processor is configured to perform the basic computing operations necessary for the operation of the device or the functionality of one or more of its components. The memory, as a rule, is implemented in the form of RAM, where the necessary program logic is loaded, providing the required functionality. When implementing the work of the proposed solution, the memory volume necessary for the implementation of the proposed solution is allocated. The data storage facility can be implemented in the form of HDD, SSD disks, RAID array, network storage, flash memory, etc.The tool allows long-term storage of various types of information. The interfaces are standard means for connecting and operating peripheral and other devices, such as USB, RS232, RJ45, COM, HDMI, PS / 2, Lightning, etc. A keyboard, joystick, display (touch display), projector, touchpad, mouse, trackball, light pen, speakers, microphone, etc. can be used as data input tools in any embodiment of the system implementing the described method. Network interaction tools are selected from a device that provides network data reception and transmission, such as an Ethernet card, WLAN / Wi-Fi module, Bluetooth module, BLE module, NFC module, IrDa, RFID module, GSM modem, etc. The tools are used to organize data exchange via a wired or wireless data transmission channel, such as WAN, PAN, LAN, Intranet, Internet, WLAN, WMAN or GSM. The components of the device are connected via a common data transmission bus.

[0107] The present application materials present a preferred disclosure of the implementation of the claimed technical solution, which should not be used as limiting other, particular embodiments of its implementation that do not go beyond the scope of the requested scope of legal protection and are obvious to specialists in the relevant field of technology.

Claims

Formula 1. A method of intelligent feedback for the purpose of stimulating the morphofunctional structures of the brain for the implementation of rehabilitation and / or prevention of neurodegeneration to improve intellectual functions and / or acquire new intellectual abilities and skills in sick or healthy people, comprising the following steps: assessing the morphofunctional state of areas of the brain to identify areas of the brain with degenerative changes in sick people, by comparing the morphofunctional characteristics of the user's brain in medical images and / or morphofunctional tests obtained as a result of examining the user, with normal values ​​​​of the corresponding social group or selecting areas of the brain to improve the desired intellectual functions of healthy people;training a user and a classifier as part of an invasive and / or non-invasive brain-computer interface, wherein the user is trained to use the invasive and / or non-invasive brain-computer interface as an input device for performing mental control in a virtual environment using feedback based on a signal to the user indicating the effectiveness of control using the brain-computer interface; and the classifier is trained based on the user's brain signals to classify patterns of brain activity coming from the user when performing mental control in a virtual environment using the brain-computer interface;adaptively adjust the parameters of the brain-computer interface classifier in order to change the patterns of mental control so that, with further use by the user of the invasive and / or non-invasive brain-computer interface, the identified areas of the brain with degenerative changes or selected areas of the brain are involved in mental control to improve intellectual functions and / or acquire new intellectual abilities and skills when implementing mental control in a virtual environment using the brain-computer interface;carry out adaptive training of the user and the classifier with adaptively configured parameters as part of an invasive and / or non-invasive brain-computer interface, wherein the classifier with adaptively configured parameters processes information from brain signals in accordance with the configured parameters at the previous stage to activate the identified areas of the brain with degenerative changes or selected areas of the brain to improve intellectual functions and / or acquire new intellectual abilities and skills when implementing mental control of commands in a virtual environment with the help of; 22 brain-computer interface to activate neuroplasticity in these areas of the brain.

2. The method according to paragraph 1, characterized in that the morphofunctional characteristics of the brain are the morphological size of the regions of the brain and / or the intensity of signal absorption in a medical image and / or the location of the regions and / or the change in the electromagnetic activity of the brain, measured by morphofunctional tests.

3. The method according to claim 1, characterized in that the invasive brain-computer interface is a neuroimplant or vascular stent with the function of an electrode and / or sensor of a fluorescent or luminescent, or chemical, or magnetic, or electromagnetic characteristic of brain activity.

4. The method according to claim 2, characterized in that the non-invasive brain-computer interface is based on an electroencephalogram and / or a method for measuring fluorescent and / or luminescent and / or chemical and / or magnetic and / or electromagnetic characteristics of brain activity.

5. The method according to paragraph 1, characterized in that the activation of neuroplasticity in areas of the brain activates the processes of synaptogenesis and / or cytogenesis and / or regeneration and / or neuroplasticity of the brain and peripheral nervous system.

Images