Device for assessing rhythmic brain activity

Abstract

The claimed technical solution relates generally to the field of computing, and more particularly to a device for assessing rhythmic brain activity parameters in real time on a stationary multi-channel electroencephalograph. The technical result consists in faster and more accurate real-time processing of an electroencephalogram (EEG) signal to assess rhythmic brain activity parameters. The claimed technical result is achieved by means of a device for assessing rhythmic brain activity parameters in real time on a stationary multi-channel electroencephalograph, said device comprising: a unit for initially setting and calibrating a brain parameter assessment algorithm, which is designed to be capable of calculating parameters of a rhythmic signal and noise signal model and adjusting the weighting coefficients of a spatial filter on the basis of a fragment of an EEG recording; and a digital multi-channel electroencephalograph containing: a unit for preamplifying and digitizing a calibrated EEG signal, which is designed to be capable of recording low-amplitude brain potentials and also preamplifying and digitizing same; and a signal microprocessor designed to be capable of assessing rhythmic brain activity parameters in real time.

Description

[0001] A DEVICE FOR ASSESSING THE PARAMETERS OF RHYTHMIC ACTIVITY OF THE BRAIN IN REAL TIME ON BOARD A MULTICHANNEL STATIONARY ELECGROENCEPHALOGRAPH FIELD OF TECHNOLOGY

[0002] The claimed technical solution generally relates to the field of computing technology, and in particular to a device for assessing the parameters of rhythmic brain activity in real time on board a multi-channel stationary electroencephalograph.

[0003] LEVEL OF TECHNOLOGY

[0004] The prior art includes patent for utility model RU207767U1 "LOW-LATENT NEUROFEEDBACK DEVICE", FEDERAL STATE AUTONOMOUS EDUCATIONAL INSTITUTION OF HIGHER EDUCATION "NATIONAL RESEARCH UNIVERSITY "HIGHER SCHOOL OF ECONOMICS", published on 15.11.202 E

[0005] This patent describes a low-latency neurofeedback device comprising an elastic carrier in which electrodes are placed that record electroencephalography signals, characterized in that the elastic carrier is connected to a rim on which a multi-channel analog-to-digital converter unit is placed for digitizing a multi-channel electroencephalogram recorded using electrodes, and a compact high-speed monitor for presenting neurofeedback, equipped with a system for monitoring the total delay in presenting a neurofeedback signal, the multi-channel analog-to-digital converter unit for digitizing a multi-channel electroencephalogram contains an on-board computer implemented on a microprocessor, control buttons for setting neurofeedback parameters, such as electroencephalogram recording, electroencephalogram frequency band, status indicators,a power button and a micro-USB connector for communication with a personal computer for reprogramming the computer, a power battery and is connected to headphones for presenting neurofeedback via an acoustic channel.

[0006] The prior art includes patent RU2818466C1 "DEVICE FOR TRANSCRANIAL ELECTROSTIMULATION OF THE BRAIN", SIDORUK NIKOL, published 02.05.2024.

[0007] This patent describes a device for transcranial electrical stimulation of the brain, comprising a pulsed monopolar current signal generator, the output of which is connected to a first electrode, and a second electrode, characterized in that an electrode block for recording brain biopotentials, an electroencephalogram analyzer, a Bluetooth transmitting unit, a smartphone, and a Bluetooth receiving unit, the output of which is connected to the control input of the introduced digital potentiometer, the output of which is connected to the second electrode, and the input is connected to the output of the pulsed monopolar current signal generator, which is configured to supply the first electrode with a pulsed monopolar current with a frequency of 77 Hz and an amplitude value of current of 3 mA, wherein the electrode block for recording brain biopotentials contains three electrodes for installation in the ear leads Al,A2 and frontal lead Fz according to the "10-20" system and is configured to use both a monopolar and a bipolar circuit for recording the alpha rhythm of the electroencephalogram, the electrodes are connected by a shielded cable to an electroencephalogram analyzer, which is configured to pre-filter and amplify the biopotentials of the brain and convert the analog signal into a digital one for further processing, the smartphone is configured to analyze the alpha rhythm by amplitude and transmit a control feedback signal wirelessly to the Bluetooth receiving unit for controlling the digital potentiometer, which is configured to supply a pulsed stimulating current to the second electrode, the value of which depends on the amplitude value of the alpha rhythm of the electroencephalogram and varies from 3 to 30 mA.

[0008] Also known from the prior art is the invention patent US11013449B2 “METHODS AND SYSTEMS FOR DECODING, INDUCING, AND TRAINING PEAK MIND / BODY STATES VIA MULTI-MODAL TECHNOLOGIES”, SRIRAM ROSHAN NARAYAN, published 05 / 25 / 2021.

[0009] This patent describes a system and method for monitoring, analyzing, and stimulating peak mental states. Brainwave sensors are placed on the user's scalp to measure the user's mental state. A machine learning model can interpret the measured brainwave signals as a mental state for the user, such as relaxed, focused, stressed, happy, sad, etc. The system can derive the user's desired mental state, such as relaxed, etc. The system can measure the difference between the desired mental state and the measured mental state of the user and provide the user with audible and visual feedback indicating how far the user is from the desired mental state. Furthermore, the system can provide audible and visual feedback to assist or encourage the user in achieving the desired mental state.Visual and audio feedback can be provided using a virtual and / or augmented reality product.

[0010] The disadvantages of the above-mentioned known solutions are the use of signal processing without the application of a random process model, which does not allow for a reduction in delay, which in these systems directly depends on the phase-frequency characteristics of the stationary analog filter and typically exceeds 400 ms.

[0011] The proposed technical solution is aimed at eliminating the shortcomings of the current state of the art through the use of dynamic filtering methods to reduce feedback delay.

[0012] ESSENCE OF THE INVENTION

[0013] The technical challenge addressed by the proposed technical solution is the creation of a new device for assessing the parameters of rhythmic brain activity, operating in true real time on board a multi-channel stationary electroencephalograph.

[0014] This device is designed to assess brain rhythm parameters in real time with minimal and consistent latency. This results in accurate and rapid assessment of brain parameters such as brain rhythms, instantaneous power, instantaneous frequency, and instantaneous rhythm phase using innovative dynamic filtering algorithms and optimized hardware and software implementation.

[0015] In this case, the delay in determining the rhythm power does not exceed 150 ms, and the rhythm phase does not exceed 5 ms, and the accuracy of estimating the instantaneous rhythm amplitude in accordance with the cross-correlation metric is not less than 80% in relation to the true value.

[0016] The technical result consists in increasing the accuracy and speed of processing the electroencephalogram (EEG) signal in real time for assessing the parameters of rhythmic brain activity.

[0017] The claimed technical result is achieved by a device for assessing the parameters of rhythmic brain activity in real time on board a multi-channel stationary electroencephalograph, containing:

[0018] • a block for the primary setup and calibration of the algorithm for assessing the parameters of the brain (GM), designed with the ability to calculate the parameters of the rhythmic signal model and the interference signal and adjust the weighting coefficients of the spatial filter based on a segment of the EEG recording;

[0019] a digital multi-channel electroencephalograph, inside of which are located:

[0020] • a pre-amplification and digitalization unit for the calibrated EEG signal, designed with the ability to record low-amplitude GM potentials, while pre-amplifying and digitalizing them;

[0021] • a signal microprocessor designed with the ability to evaluate the parameters of the rhythmic activity of the brain in real time, while the evaluation of the parameters is reduced to the following:

[0022] carry out frequency processing of the signal using a band-stop filter and low- and high-pass filters,

[0023] - perform spatial filtering using spatial filter coefficients calculated by the preliminary parameter estimation unit and transmitted to the electroencephalograph,

[0024] - a cycle is launched in which an iterative modeling of the rhythm process is performed in real time using a dynamic oscillatory model and an evaluation of the analytical signal based on the dynamic Kalman filtering algorithm, using the calculated parameters of the GM rhythms in the previous stage,

[0025] - carry out real-time calculation of the values ​​of the envelope, phase and instantaneous frequency of the GM rhythms with minimal delay, using the parameters of rhythmic activity and interference calculated by the preliminary parameter assessment unit and transmitted to the electroencephalograph.

[0026] In a particular embodiment, the proposed device evaluates in real time such parameters of brain rhythms as: instantaneous power, instantaneous frequency and instantaneous phase of the rhythm.

[0027] In a particular embodiment, the proposed device implements on board the electroencephalograph a dynamic filtering algorithm using the Kalman method with parameters individually selected using the primary configuration unit.

[0028] DESCRIPTION OF DRAWINGS

[0029] The invention will be further described in accordance with the accompanying drawings, which are provided to illustrate the essence of the invention and in no way limit its scope. The following drawings are attached to the application: Fig. 1 illustrates a block diagram for assessing rhythmic brain activity parameters in real time onboard a multichannel stationary electroencephalograph.

[0030] DETAILED DESCRIPTION OF THE INVENTION

[0031] The following detailed description of the invention includes numerous implementation details to provide a clear understanding of the present invention. However, it will be apparent 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 to avoid obscuring the features of the present invention.

[0032] Furthermore, it will be clear from the foregoing description that the invention is not limited to the embodiment described. Numerous possible modifications, changes, variations, and substitutions, while preserving the spirit and form of the present invention, will be apparent to those skilled in the art.

[0033] The claimed technical solution proposes a device for assessing the parameters of rhythmic brain activity in real time on board a multi-channel stationary electroencephalograph.

[0034] Fig. 1 shows a block diagram of a device for assessing the parameters of rhythmic brain activity in real time on board a multi-channel electroencephalograph, which includes the following components:

[0035] A block for the primary setup and calibration of the brain parameter estimation algorithm (BPA), designed with the ability to calculate the parameters of the rhythmic signal and interference signal model and adjust the weighting coefficients of the spatial filter based on a segment of the EEG recording, namely, determining the parameters of the rhythmic signal and interference signal model that are used in the dynamic filtering algorithm, as well as the weights of the spatial filter.

[0036] A digital multichannel electroencephalograph, which contains: - a pre-amplification and digitalization unit for the calibrated EEG signal, designed with the ability to record low-amplitude GM potentials, while pre-amplifying and digitalizing them;

[0037] - and a signal microprocessor, designed with the ability to evaluate the parameters of the rhythmic activity of the GM in real time, wherein the evaluation of the parameters is reduced to the following: frequency processing of the signal is carried out using a band-stop filter and low- and high-pass filters,

[0038] perform spatial filtering using spatial filter coefficients calculated by the preliminary parameter estimation unit and transmitted to the electroencephalograph,

[0039] a cycle is launched in which the rhythm process is iteratively modeled in real time using a dynamic oscillatory model and the analytical signal is evaluated based on the dynamic Kalman filtering algorithm, using the calculated parameters of the GM rhythms in the previous stage,

[0040] They calculate the values ​​of the envelope, phase and instantaneous frequency of GM rhythms in real time with minimal delay, using the parameters of rhythmic activity and interference calculated by the preliminary parameter assessment unit and transmitted to the electroencephalograph.

[0041] There are certain requirements for a digital multichannel electroencephalograph, including at least two channels for EEG recording and the absence of built-in analog filters that introduce a group signal delay greater than 2 ms. Depending on the intended use, the device must also have trigger or digital outputs for external devices and monitors.

[0042] Additionally, there are certain requirements for a signal microprocessor, specifically a signal processor optimized for mathematical operations. Memory of at least 250 KB for storing intermediate and final results of algorithm operations, a clock frequency of at least 200 MHz, the ability to implement a digital interface with a computer, and preferably DMA (direct memory access) to the microprocessor to eliminate internal delays when transferring data between processes.

[0043] The claimed device can operate in a fully autonomous mode with fixed signal processing parameters and without the ability to record data, for example, to generate triggers on external devices when the rhythm power exceeds the threshold of 50 μV, or it can include an optional unit - a computer, for setting up and transmitting parameters via a digital interface, as well as a computer driver that provides connection via a digital interface.

[0044] An algorithm is implemented on a signal microprocessor for estimating such parameters of brain activity as instantaneous power, instantaneous frequency, and instantaneous phase of brain rhythms based on data obtained with a multichannel digital electroencephalograph with minimal latency. This goal is achieved by dynamic filtering applied to the electroencephalogram signal using a state-space model of rhythmic brain activity (Matsuda, T., & Komaki, F. (2017). Time series decomposition into oscillation components and phase estimation. Neural Computation, 29(2), 332–367.).

[0045] In this case, to reduce the amount of calculations and decrease the load on the signal microprocessor, a special stationary mode of operation of the dynamic filter with a smaller number of calculations is implemented (Grimble, M. J. (1979). Solution of the Kalman filtering problem for stationary noise and finite data records. International Journal Of Systems Science, 10(2), 177-196.). In addition to reducing the algorithmic delay, acceleration of the algorithm and operation in true real time are also achieved due to the implementation of all calculations on the signal microprocessor of the multichannel electroencephalograph, without transmitting data to the computer with the accompanying unstable delays, which are characteristic of the overwhelming majority of existing systems.

[0046] The algorithm for calculating the parameters of GM rhythms is described below:

[0047] The first step involves defining a spatial filter that allows the target brain rhythm to be isolated for analysis in the primary setup and calibration block of the brain parameter (BP) estimation algorithm (hereinafter referred to as Block 1). At this stage, calibration is performed by recording a short EEG segment (from 30 to 120 seconds) and adjusting the spatial filter weighting coefficients based on this recording. These coefficients are then used in the signal microprocessor during the brain rhythm parameter calculation algorithm cycle for multiplication by the corresponding channels to obtain pure cerebral cortex signal sources. At this stage, implementation variations are possible: either the filter is tuned away from artifacts and selective BP signal sources, or it is completely focused on a specific BP signal source.

[0048] In this case, spatial filtering algorithms such as ICA (Independent Component Analysis) or CSP (Common Spatial Patterns) are used, which allow, without marking EEG recording areas (ICA) or with marking contrast conditions (CSP), to isolate pure signal sources such as occipital alpha rhythm, sensorimotor rhythm, and others.

[0049] The next stage involves primary frequency processing of the signal, specifically the application of a band-stop filter, low-pass filters, and high-pass filters. The filters are selected to introduce minimal group delay while maintaining a frequency response as close as possible to the ideal. Standard filters are first-order Butterworth IIR filters with cutoff frequencies of 50, 100, 150, and 200 Hz (for the band-stop) and 2.30 Hz for the high-pass and low-pass filters, respectively.

[0050] The recorded EEG segment is used to determine the optimal parameters of the brain rhythm parameter detection algorithm. Parameter optimality affects the accuracy and speed of obtaining brain rhythm parameters. When using a dynamic model, the parameters include q (process noise), r (observation noise), GO (center frequency of the rhythm), and A (process attenuation parameter). When calculating complex rhythm characteristics requiring multiple components, the parameters are optimized for each rhythm. The GO frequency is found as the argument of the peak value of the spectral power density within the acceptable range. Parameters q and r are found by searching for the global minimum in the root-mean-square loss function between the dynamically filtered signal and the true signal obtained offline by processing with a third-order Butterworth bandpass filter.

[0051] The calculated parameters are transmitted to the EEG device, and a cycle is initiated in which the rhythm process is iteratively modeled in real time using a dynamic oscillatory model and the analytical signal is evaluated using a maximum a posteriori probability algorithm. Rhythm modeling x t = A • G( о ) • x t-1 + q, s t = Hx t , where x t - vector corresponding to the analytical signal, F(.) - process dynamics matrix (rotation matrix), H - observation matrix (1, 0), s t- the registered EEG signal after preprocessing. The analytical signal is evaluated using standard Kalman filter formulas (Zarchan, R., and H. Musoff. "Fundamentals of Kalman filtering: a practical approach, vol. 190." Progress in Astronautics and Aeronautics, American Institute of Aeronautics and Astronautics (AIAA) (2000). The algorithm is optimized to a stationary Kalman filter that does not require extensive matrix multiplications. Depending on the number of desired spectral components in the signal and on the assumptions about the process noise (white, pink, etc.), the algorithm can take more complex forms, while preserving the basic elements listed.

[0052] The parameters of the GM rhythms are calculated in the following ways.

[0053] Envelope value - P t = abs(X t ) , where X t - the value of the analytical signal at the current moment in time.

[0054] The phase value is 0t= angle(Xt).

[0055] Next, more complex indicators are calculated that characterize the relationship between the characteristics of rhythmic activity indicators. Concentration index - 1 с = — , the ratio of rhythmic activity in the ?-range р A

[0056] in the frontal frequency of the GM to rhythmic activity in the a-range in the occipital region.

[0057] Normalized relaxation index - 1 Г = — - — - where FFT stands for

[0058]

[0059] Fourier transform, x win - EEG signal accumulated in the window for several seconds.

[0060] The proposed device enables real-time assessment of rhythmic brain activity with a delay of no more than 150 ms for power assessment and 5 ms for rhythm phase assessment. The proposed solution can be implemented on a wide range of EEG equipment. The device is highly ergonomic due to the reduced number of elements in the signal recording and processing chain compared to existing systems.

Claims

CLAUSES OF THE INVENTION 1. A device for assessing the parameters of rhythmic brain activity in real time on board a multi-channel stationary electroencephalograph, containing: • a block for the primary setup and calibration of the algorithm for assessing the parameters of the brain (GM), designed with the ability to calculate the parameters of the rhythmic signal model and the interference signal and adjust the weighting coefficients of the spatial filter based on a segment of the EEG recording; • digital multichannel electroencephalograph, inside which are located: • a pre-amplification and digitalization unit for the calibrated EEG signal, designed with the ability to record low-amplitude GM potentials, while pre-amplifying and digitalizing them; • a signal microprocessor designed with the ability to evaluate the parameters of the rhythmic activity of the brain in real time, while the evaluation of the parameters is reduced to the following: carry out frequency processing of the signal using a band-stop filter and low- and high-pass filters, - perform spatial filtering using spatial filter coefficients calculated by the preliminary parameter estimation unit and transmitted to the electroencephalograph, - a cycle is launched in which an iterative modeling of the rhythm process is performed in real time using a dynamic oscillatory model and an evaluation of the analytical signal based on the dynamic Kalman filtering algorithm, using the calculated parameters of the GM rhythms in the previous stage, - carry out real-time calculation of the values ​​of the envelope, phase and instantaneous frequency of the GM rhythms with minimal delay, using the parameters of rhythmic activity and interference calculated by the preliminary parameter assessment unit and transmitted to the electroencephalograph.

2. The device according to paragraph 1, characterized in that it evaluates in real time such parameters of brain rhythms as: instantaneous power, instantaneous frequency and instantaneous phase of the rhythm.

3. The device according to paragraph 1, characterized in that it implements on board the electroencephalograph a dynamic filtering algorithm using the Kalman method with parameters individually selected using the primary tuning unit.

Images