A method and system for simulating the acquisition of electroencephalogram (EEG) signals

Abstract

This specification provides a method and system for generating simulated electroencephalogram (EEG) signals. The method includes: obtaining a noise-free simulated EEG signal based on a base signal and a first parameter set; determining a first sub-noise signal based on a region-of-interest (ROI) transfer matrix and simulated brain noise signals; determining a second sub-noise signal based on an electrode point transfer matrix and electrode point noise signals; obtaining a target noise signal based on the first and second sub-noise signals; acquiring a first weight of the simulated EEG signal and a second weight of the target noise signal; and obtaining a target simulated EEG signal based on the simulated EEG signal, the target noise signal, the first weight, and the second weight. The simulated EEG signal, the target noise signal, and the target simulated EEG signal can be used as noise-free EEG signals in the human body, noise signals in acquired EEG signals, and acquired EEG signals, respectively, for analysis and research of EEG signals.

Description

[0001] Case Analysis

[0002] This application is a divisional application of Chinese Patent Application No. 202210659471.8, entitled "A method and system for generating analog acquisition of electroencephalogram (EEG) signals", filed on June 13, 2022. Technical Field

[0003] This specification relates to the field of analog electroencephalogram (EEG) signals, and in particular to a method and system for generating analog EEG signals. Background Technology

[0004] With the advancement of science and the rapid development of biomedical technology, biosignals are being applied to research in various fields. For example, electroencephalogram (EEG) signals can be used for sleep classification, and as an auxiliary diagnostic tool for epilepsy and depression. However, EEG signals have small amplitudes, typically only tens to hundreds of microvolts, and are susceptible to interference from various noise signals during acquisition, most commonly electrooculogram (EOG) and electromyogram (EMG) artifacts. Furthermore, when using EEG acquisition devices to collect human EEG signals, the propagation of both EEG signals and noise signals from their origin to the electrode points in the device is affected by various factors, including brain functional areas and electrode points. This results in the EEG signals and noise signals ultimately acquired by the electrode points differing from the signals emitted from their origin, making it difficult to distinguish between the EEG signals and noise signals within the acquired EEG signal.

[0005] Currently, machine learning models can be used to denoise the collected EEG signals. However, it is difficult to determine the EEG signals and noise signals after propagation in the collected EEG signals, making it difficult to obtain the large amount of training data required for model training.

[0006] Therefore, it is necessary to propose a method for generating simulated EEG signals, which can generate the propagated simulated EEG signals, the propagated target noise signals, and the simulated acquired EEG signals. This can obtain a large amount of training data required for model training, thereby enabling the trained denoised model to have better generalization performance and reducing the cost of collecting human EEG signals when studying EEG signals. Summary of the Invention

[0007] This specification provides one or more embodiments of a method for generating simulated electroencephalogram (EEG) signals. The method includes: obtaining a noise-free simulated EEG signal based on a base signal and a first parameter set, wherein the simulated EEG signal is used as a noise-free EEG signal in the human body for analysis and research; obtaining a target noise signal based on an initial noise signal and a second parameter set, wherein the target noise signal is used as a noise signal in the collected EEG signals of the human body for analysis and research, wherein the second parameter set includes a region of interest (ROI) transfer matrix and an electrode point transfer matrix, the initial noise signal includes simulated brain noise signals and electrode point noise signals, and obtaining the target noise signal based on the initial noise signal and the second parameter set includes: determining a first sub-region based on the ROI transfer matrix and the simulated brain noise signal. The noise signal, wherein the elements in the region of interest transfer matrix represent the transfer relationship of the signal when propagating between different regions of interest; based on the electrode point transfer matrix and the electrode point noise signal, a second sub-noise signal is determined, wherein the elements in the electrode point transfer matrix represent the transfer relationship of the signal when propagating between different electrode points; based on the first sub-noise signal and the second sub-noise signal, the target noise signal is obtained; a first weight of the simulated EEG signal and a second weight of the target noise signal are obtained; based on the simulated EEG signal, the target noise signal, the first weight, and the second weight, a target simulated acquired EEG signal is obtained, which is used as the acquired EEG signal to participate in the analysis and research of the EEG signal and to reduce the acquisition of real EEG signals.

[0008] This specification provides one or more embodiments of a system for generating simulated electroencephalogram (EEG) signals. The system includes: a first acquisition module for obtaining a noise-free simulated EEG signal based on a base signal and a first parameter set, wherein the simulated EEG signal is used as a noise-free EEG signal in the human body for analysis and research of EEG signals; and a second acquisition module for obtaining a target noise signal based on an initial noise signal and a second parameter set, wherein the target noise signal is used as a noise signal in the collected EEG signals of the human body for analysis and research of the EEG signals, wherein the second parameter set includes a region of interest (ROI) transfer matrix and an electrode point transfer matrix, and the initial noise signal includes simulated brain noise signals and electrode point noise signals. The second acquisition module is further configured to: determine a first sub-noise signal based on the ROI transfer matrix and the simulated brain noise signal, wherein... The elements in the region of interest transfer matrix represent the transfer relationship of the signal when propagating between different regions of interest; based on the electrode point transfer matrix and the electrode point noise signal, a second sub-noise signal is determined, wherein the elements in the electrode point transfer matrix represent the transfer relationship of the signal when propagating between different electrode points; based on the first sub-noise signal and the second sub-noise signal, the target noise signal is obtained; a third acquisition module is used to acquire the first weight of the simulated EEG signal and the second weight of the target noise signal; a fourth acquisition module is used to acquire a target simulated acquired EEG signal based on the simulated EEG signal, the target noise signal, the first weight, and the second weight, wherein the target simulated acquired EEG signal is used as the acquired EEG signal to participate in the analysis and research of the EEG signal and to reduce the acquisition of real EEG signals.

[0009] This specification provides one or more embodiments of a device for generating simulated electroencephalogram (EEG) signals, including a processor for executing a method for generating simulated EEG signals as described in any of the above embodiments.

[0010] This specification provides one or more embodiments of a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes them to implement the method for generating simulated electroencephalogram (EEG) signals as described in any of the above embodiments. Attached Figure Description

[0011] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0012] Figure 1This is a schematic diagram illustrating an application scenario of a simulated EEG signal generation system according to some embodiments of this specification;

[0013] Figure 2 This is an exemplary block diagram of a system for generating analog acquisition of electroencephalogram (EEG) signals according to some embodiments of this specification;

[0014] Figure 3 This is a schematic diagram illustrating the generation of simulated electroencephalogram (EEG) signals according to some embodiments of this specification;

[0015] Figure 4 This is an exemplary flowchart illustrating the acquisition of noise-free simulated electroencephalogram signals according to some embodiments of this specification;

[0016] Figure 5 This is an exemplary flowchart illustrating the acquisition of a target noise signal according to some embodiments of this specification;

[0017] Figure 6 This is an exemplary flowchart illustrating the acquisition of the first parameter group according to some embodiments of this specification;

[0018] Figure 7 This is an exemplary flowchart of a method for denoising acquired electroencephalogram signals according to some embodiments of this specification;

[0019] Figure 8 This is an exemplary flowchart illustrating the acquisition of signal transfer relationships according to some embodiments of this specification;

[0020] Figure 9 This is a schematic diagram of a signal transfer model according to some embodiments of this specification;

[0021] Figure 10 This is a schematic diagram of yet another signal transfer model shown according to some embodiments of this specification;

[0022] Figure 11 This is an exemplary flowchart illustrating the determination of a target noise signal according to some embodiments of this specification. Detailed Implementation

[0023] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0024] It should be understood that the terms “system,” “device,” “unit,” and / or “module” used herein are one way to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0025] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0026] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0027] Figure 1 This is a schematic diagram illustrating an application scenario of a system for generating simulated electroencephalogram (EEG) signals, based on some embodiments of this specification.

[0028] like Figure 1 As shown, the application scenario 100 of the simulated EEG signal generation system may include a server 110, a base signal 120, an initial noise signal 130, a network 140, a storage device 150, and a terminal device 160. The simulated EEG signal generation system can acquire simulated EEG signals by implementing the methods and / or processes disclosed in this specification.

[0029] Server 110 can communicate with base signal 120, initial noise signal 130, network 140, storage device 150, and terminal device 160 to realize various functions of the system for generating simulated EEG signals. In some embodiments, server 110 can receive and process base signal 120 and initial noise signal 130 via, for example, network 140. In some embodiments, server 110 can output relevant data to storage device 150 and terminal device 160 via, for example, network 140. In some embodiments, server 110 can be a single server or a group of servers. In some embodiments, server 110 can be locally connected to network 140 or remotely connected to network 140. In some embodiments, server 110 can be implemented on a cloud platform.

[0030] The basic signal 120 can be a simple signal. For example, the basic signal 120 can be a sine wave. The basic signal 120 can be generated by any one or more of a waveform generator, analog signal generator, waveform generator, etc. In some embodiments, the basic signal 120 can be sent to the server 110, storage device 150, and terminal device 160 via network 140.

[0031] The initial noise signal 130 can be a noise signal that has not been propagated during the acquisition of EEG signals. The initial noise signal 130 can be generated by any one or more of the following: a noise generator, an analog signal generator, or a noise generator. In some embodiments, the initial noise signal 130 can be sent to a server 110, a storage device 150, and a terminal device 160 via a network 140.

[0032] Network 140 can be used for the transmission of information and / or data. In some embodiments, one or more components of application scenario 100 (e.g., server 110 and / or storage device 150, etc.) can send information and / or data to another component in application scenario 100 via network 140.

[0033] Storage device 150 may store data and / or instructions. Data may include information relating to server 110, terminal device 160, base signal 120, and initial noise signal 130. For example, storage device 150 may store base signal 120 and initial noise signal 130. In some embodiments, storage device 150 may store data and / or instructions used by server 110 to perform or use in order to complete the exemplary methods described herein. In some embodiments, storage device 150 may be connected to network 140 to communicate with one or more components of application scenario 100 (e.g., server 110 and / or terminal device 160). In some embodiments, storage device 150 may be part of server 110. In some embodiments, storage device 150 may include mass storage, removable storage, volatile read-write memory, read-only memory (ROM), and any combination thereof. In some embodiments, storage device 150 may be implemented on a cloud platform.

[0034] Terminal device 160 can refer to one or more terminals or software used by a user. Here, "user" refers to a researcher. The user can view relevant data for various components in application scenario 100 through terminal device 160. In some embodiments, terminal device 160 may include mobile device 160-1, tablet computer 160-2, laptop computer 160-3, etc., or any combination thereof. In some embodiments, terminal device 160 can be fixed and / or mobile. For example, terminal device 160 can be directly installed on server 110, becoming part of server 110. As another example, terminal device 160 can be a portable device, allowing the user to carry terminal device 160 to a location relatively far from server 110, base signal 120, and initial noise signal 130. Terminal device 160 can connect to and / or communicate with server 110 and storage device 150 via network 140.

[0035] It should be noted that application scenario 100 is provided for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can make various modifications or variations based on the description in this specification. For example, application scenario 100 may also include a database. However, these changes and modifications will not depart from the scope of this specification.

[0036] Figure 2 This is an exemplary block diagram of a system for generating simulated electroencephalogram (EEG) signals according to some embodiments of this specification.

[0037] like Figure 2 As shown, the simulated acquisition system 200 for generating EEG signals may include a first acquisition module 210, a second acquisition module 220, a third acquisition module 230, and a fourth acquisition module 240.

[0038] The first acquisition module 210 can be used to obtain noise-free simulated EEG signals based on the basic signal and the first parameter set. These simulated EEG signals are used as noise-free EEG signals in the human body for analysis and research. For more information on the basic signal, the first parameter set, and the simulated EEG signals, please refer to [link to relevant documentation]. Figure 3 And related descriptions. In some embodiments, the first acquisition module 210 can also be used to acquire an initial first parameter set; obtain an intermediate simulated EEG signal based on the initial first parameter set and the base signal; acquire a reference EEG signal; construct a loss function based on the intermediate simulated EEG signal and the reference EEG signal; iteratively update the parameters of the initial first parameter set based on the loss function until a preset condition is met, and obtain the first parameter set. For more information on the initial first parameter set, intermediate simulated EEG signal, and reference EEG signal, see [link to relevant documentation]. Figure 6And related descriptions. In some embodiments, the first parameter group includes a region of interest (ROI) transfer matrix and a signal source transfer matrix. The elements in the ROI transfer matrix represent the transfer relationship when a signal propagates between different ROIs, and the elements in the signal source transfer matrix represent the transfer relationship when a signal source and an electrode point propagate within the same ROI. Different ROIs represent different functional areas of the human brain. The first acquisition module 210 can also be used to determine the target basic signal of each signal source in the virtual brain model based on the ROI transfer matrix and the basic signal emitted by that signal source; to determine the initial sub-simulated EEG signal corresponding to each ROI in the virtual brain model based on the target basic signal of each signal source in that ROI and the signal source transfer matrix; and to obtain simulated EEG signals based on the initial sub-simulated EEG signals corresponding to each ROI. For more information on the ROI transfer matrix, signal source transfer matrix, target basic signal, and initial sub-simulated EEG signals, see [link to documentation]. Figure 4 And its related descriptions.

[0039] The second acquisition module 220 can be used to obtain a target noise signal based on the initial noise signal and the second parameter set. The target noise signal is used as the noise signal in the acquired EEG signals to participate in the analysis and research of the EEG signals. For more information on the initial noise signal, the second parameter set, and the target noise signal, please refer to [link to relevant documentation]. Figure 3 And related descriptions. In some embodiments, the initial noise signal includes simulated brain noise signal and electrode point noise signal. The second parameter set includes region of interest transfer matrix and electrode point transfer matrix. The elements in the region of interest transfer matrix represent the transfer relationship of the signal when propagating between different regions of interest, and the elements in the electrode point transfer matrix represent the transfer relationship of the signal when propagating between different electrode points. The second acquisition module 220 can also be used to determine a first sub-noise signal based on the region of interest transfer matrix and the simulated brain noise signal; determine a second sub-noise signal based on the electrode point transfer matrix and the electrode point noise signal; and obtain a target noise signal based on the first sub-noise signal and the second sub-noise signal. For more information on simulated brain noise signal, electrode point noise signal, electrode point transfer matrix, first sub-noise signal, and second sub-noise signal, see [link to documentation]. Figure 5 And its related descriptions.

[0040] In some embodiments, the second acquisition module 220 may include sub-modules such as a first determination module, a second determination module, and a signal processing module. The first determination module is used to determine the signal transfer relationship of the object under test. The signal transfer relationship characterizes the signal change relationship when the initial noise signal propagates between various locations of the object under test. More information about the initial noise signal, the object under test, and the signal transfer relationship can be found in [link to relevant documentation]. Figure 7 And related descriptions. In some embodiments, the first determining module can also be used to input an initial test signal at each location of the test subject, wherein the order of magnitude of the initial test signal is larger than the order of magnitude of the EEG signal of the test subject; to acquire signals from other locations of the test subject to obtain acquired test signals; and to determine signal transfer relationships based on the initial test signals and the acquired test signals. More information about the initial test signals and the acquired test signals can be found in [link to relevant documentation]. Figure 8 And its related description. The second determining module is used to determine the target noise signal based on the initial noise signal and the signal transition relationship. For more information on the target noise signal, please refer to... Figure 7 And related descriptions. In some embodiments, the signal transfer relationship includes multiple simple signal transfer relationships, which characterize the signal change relationship when the first simple noise signal propagates between various locations of the object under test. The first simple noise signal is obtained by transforming an initial noise signal. The second determining module can also be used to transform the initial noise signal into multiple first simple noise signals with different frequencies and intensities. For each first simple noise signal, the first simple noise signal is processed based on the simple signal transfer relationship corresponding to it to determine a second simple noise signal corresponding to the first simple noise signal. The second simple noise signals are then synthesized to obtain the target noise signal. More information on simple signal transfer relationships, first simple noise signals, and second simple noise signals can be found in [link to relevant documentation]. Figure 10 And its related description. The signal processing module is used to remove target noise signals from the acquired EEG signals from the subject, obtaining noise-free target EEG signals. For more information on acquired and target EEG signals, please refer to... Figure 7 And its related descriptions.

[0041] The third acquisition module 230 can be used to acquire the first weight of the simulated EEG signal and the second weight of the target noise signal. For more information on the first and second weights, please refer to [link to relevant documentation]. Figure 3 And its related descriptions.

[0042] The fourth acquisition module 240 can be used to obtain a target simulated acquired EEG signal based on the simulated EEG signal, the target noise signal, the first weight, and the second weight. This target simulated acquired EEG signal is used as the acquired EEG signal for analysis and research, and to reduce the acquisition of real EEG signals. For more information on target simulated acquired EEG signals, please refer to [link to relevant documentation]. Figure 3 And its related descriptions.

[0043] It should be noted that the above description of the analog EEG signal generation system 200 and its modules is for convenience only and should not be construed as limiting this specification to the scope of the embodiments described. It is understood that those skilled in the art, after understanding the principles of this system, may arbitrarily combine the various modules or construct subsystems connected to other modules without departing from these principles. In some embodiments, Figure 2 The first acquisition module 210, the second acquisition module 220, the third acquisition module 230, and the fourth acquisition module 240 disclosed herein can be different modules within a single system, or a single module can implement the functions of two or more of the aforementioned modules. For example, the modules can share a single storage module, or each module can have its own separate storage module. Such variations are all within the scope of protection of this specification.

[0044] Figure 3 This is a schematic diagram illustrating the generation of simulated electroencephalogram (EEG) signals according to some embodiments of this specification. In some embodiments, process 300 may be executed by server 110. Figure 3 As shown, process 300 may include the following steps:

[0045] Step 310: Based on the basic signal and the first parameter set, obtain a noise-free simulated EEG signal. In some embodiments, step 310 may be performed by the first acquisition module 210.

[0046] The basic signal can refer to a simple signal with a pattern of signal changes. Analog EEG signals can be used as noise-free EEG signals in the human body for analysis and research. The basic signal can include, but is not limited to, sine, cosine, and square wave signals. The first acquisition module 210 can input the characteristics of the basic signal into the analog signal generator to obtain the basic signal. The characteristics of the basic signal can include, but are not limited to, its frequency and amplitude. These characteristics can be preset so that the analog EEG signal obtained based on the basic information conforms to the basic characteristics of a real EEG signal. For example, the frequency of the basic signal can be preset to be less than 100Hz, and the amplitude range can be -200uV to +200uV.

[0047] The first parameter set can refer to a set of parameters that characterizes the influence of factors such as the region of interest (ROI), signal source, and electrode points installed on the virtual brain model on the generation of the basic signal during its propagation. In some embodiments, the basic signal can be processed using multiple different first parameter sets to obtain multiple different simulated EEG signals. Different ROIs in the virtual brain model can represent various functional areas of the human brain. For example, a ROI in the virtual brain model can represent the frontal lobe of the human brain, which is responsible for tasks related to attention and short-term memory.

[0048] It should be understood that when a signal propagates in the human brain, different brain structures in different regions of interest, different signal sources, and the electrode points used to collect the EEG signal all affect the signal, causing changes. In some embodiments of this specification, the influence of the region of interest, signal source, and electrode points on the signal can be determined using a first parameter set, so that the obtained simulated EEG signal is closer to the real noise-free EEG signal.

[0049] In some embodiments, the first parameter set may include the changes in signal propagation between different regions of interest (ROIs) in the virtual brain model. The first parameter set may include a region-of-interest (ROI) transfer matrix, which represents the changes in the basic signal propagation between different ROIs in the virtual brain model. For more information on ROI transfer matrices, see [link to relevant documentation]. Figure 4 And its related descriptions.

[0050] In some embodiments, the first parameter set may further include the variation of a signal emitted from a signal source in a region of interest within the virtual brain model as it propagates between electrode points within that region of interest. The first parameter set may also include a signal source transfer matrix, which represents the variation of a fundamental signal emitted from a signal source in a region of interest within the virtual brain model as it propagates between electrode points within that region of interest. For more information on signal source transfer matrices, see [link to relevant documentation]. Figure 4 And its related descriptions.

[0051] In some embodiments, the elements in the first parameter set can be fixed values ​​or range values. When the first parameter set is a range value, when processing the basic signal based on the first parameter set, a specific value can be randomly selected from the range value to process the basic signal, thereby obtaining a simulated electroencephalogram (EEG) signal. The first parameter set is different for different human bodies or different virtual brain models.

[0052] In some embodiments, the first parameter set can be obtained in various ways. For example, it can be obtained via a network. Another example is that it can be obtained through random initialization.

[0053] In some embodiments, the first parameter set is obtained based on a first preset parameter range. The first parameter set can be obtained by setting a first preset parameter range and iteratively updating the range of the first preset parameter range. The first preset parameter range can refer to the preset range of each element in the first parameter set. The first preset parameter range can be obtained in various ways; for example, it can be determined by randomly selecting an initial range. Another example is that it can be determined by expanding the range of an existing first parameter set. The parameter set corresponding to the first preset parameter range can be the initial first parameter set. In some embodiments, an intermediate simulated EEG signal can be obtained based on the initial first parameter set and the base signal; a loss function can be constructed based on the intermediate simulated EEG signal and the reference EEG signal; the initial first parameter set is iteratively updated based on the loss function until a preset condition is met, thus obtaining the first parameter set. For more details on the embodiments for obtaining the first parameter set described above, please refer to... Figure 6 And its related descriptions.

[0054] In some embodiments of this specification, the approximate range of the first parameter group can be determined by setting a first preset parameter range for the first parameter group, thereby reducing the amount of calculation required to determine the first parameter group and determining the first parameter group more quickly.

[0055] Simulated EEG signals can refer to signals that simulate the EEG signals of the human brain. In some embodiments, the base signal can be processed based on a first set of parameters to determine the influence of the region of interest (ROI), signal source, and electrode points installed on the virtual brain model on the base signal during acquisition, thereby obtaining simulated EEG signals. Noise-free simulated EEG signals are obtained. In some embodiments, for each ROI in the virtual brain model, based on the signal source transfer matrix and the base signal, an initial sub-simulated EEG signal corresponding to that ROI is determined; based on the ROI transfer matrix and the initial sub-simulated EEG signals corresponding to each ROI, the simulated EEG signal is obtained. For more details on the above embodiments, see [link to relevant documentation]. Figure 4 And its related descriptions.

[0056] In some embodiments, by emitting basic signals from multiple signal sources in a virtual brain model and superimposing these basic signals, the acquired simulated EEG signal can possess the basic characteristics of an EEG signal. These basic characteristics may include, but are not limited to, amplitude characteristics (e.g., amplitude range) and frequency characteristics (e.g., frequency range). For example, the basic characteristics of an EEG signal may include an amplitude range of -200µV to +200µV and a frequency less than 100Hz. Three signal sources in the virtual brain model can emit three sinusoidal signals with amplitudes of 10µV, 30µV, and 100µV, and frequencies of 5Hz, 20Hz, and 40Hz, respectively. Superimposing these three sinusoidal signals yields a more complex simulated EEG signal. This simulated EEG signal has an amplitude less than 100µV and three frequency components, all less than 100Hz. Therefore, this simulated EEG signal possesses the aforementioned basic characteristics of an EEG signal.

[0057] It should be understood that the simulated EEG signal is directly obtained by processing the basic signal through the first parameter group. Therefore, the simulated EEG signal does not contain noise signals caused by the activity of the acquisition device or other parts of the human body when the human EEG signal is acquired.

[0058] Step 320: Obtain the target noise signal based on the initial noise signal and the second parameter set. In some embodiments, step 320 may be performed by the second acquisition module 220.

[0059] Initial noise signal refers to the noise signal that has not been propagated during the acquisition of EEG signals. Target noise signal can be used as the noise signal in the acquired EEG signals to participate in the analysis and research of EEG signals.

[0060] During the acquisition of electroencephalogram (EEG) signals, various factors can affect the acquisition, resulting in the inclusion of various noise signals in the acquired EEG signals. Correspondingly, the initial noise signal can include multiple noise signals. In some embodiments, the initial noise signal can be determined through acquisition or simulation. For more information on determining the initial noise signal, see [link to relevant documentation]. Figure 5 And its related descriptions.

[0061] The second parameter set can refer to a set of parameters that characterize the influence of factors such as the region of interest in the virtual brain model, the signal source, and the electrode points installed on the virtual brain model on the generation of the initial noise signal during its propagation.

[0062] Similar to the first parameter set, the second parameter set may also include a region-of-interest (ROI) transfer matrix. In some embodiments, the second parameter set may further include an electrode transfer matrix, which represents the changes in signal propagation between electrodes in a virtual brain model. For more information on electrode transfer matrices, see [link to documentation]. Figure 4 And its related descriptions.

[0063] In some embodiments, the second parameter set can be obtained in various ways. For example, the region-of-interest transition matrix can be obtained via a network.

[0064] In some embodiments, the second parameter set is obtained based on a second preset parameter range. The second parameter set can be obtained by setting a second preset parameter range and iteratively updating the range of the second preset parameter range. The second preset parameter range can refer to a preset range of elements in the second parameter values. The second preset parameter range can be obtained in various ways, for example, by randomly selecting an initial range. The parameter set corresponding to the second preset parameter range can be an initial second parameter set. In some embodiments, the second parameter set can be determined based on the initial second parameter set. For more information on determining the second parameter set based on the initial second parameter set, please refer to [link to relevant documentation]. Figure 6 And related explanations.

[0065] In some embodiments, an initial test signal may be input from each electrode point of the test subject, wherein the order of magnitude of the initial test signal is larger than the order of magnitude of the EEG signal of the test subject; signals may be acquired from other electrode points of the test subject to obtain acquired test signals; based on the initial test signals and the acquired test signals, the signal transfer relationship between the electrode point and other electrode points may be determined; and based on the signal transfer relationship between each electrode point and other electrode points, an electrode point transfer matrix may be determined. For more details on the embodiments for determining the electrode point transfer matrix described above, please refer to... Figure 8 And its related descriptions.

[0066] The target noise signal can refer to the noise signal in the simulated acquired EEG signal. In some embodiments, the initial noise signal can be processed based on a second set of parameters to determine the influence of the region of interest (ROI), signal source, and electrode points installed on the virtual brain model on the initial noise signal during acquisition, thereby obtaining the target noise signal. In some embodiments, a first sub-noise signal is determined based on the ROI transfer matrix and the simulated brain noise signal; a second sub-noise signal is determined based on the electrode point transfer matrix and the electrode point noise signal; and the target noise signal is obtained based on the first and second sub-noise signals. For more details on the above embodiments, see [link to relevant documentation]. Figure 5 And its related descriptions.

[0067] Step 330: Obtain the first weight of the simulated EEG signal and the second weight of the target noise signal. In some embodiments, step 330 may be performed by the third acquisition module 230.

[0068] The first weight can refer to the proportion of the simulated EEG signal in the target simulated EEG signal. The second weight can refer to the proportion of the target noise signal in the target simulated EEG signal. The first and second weights can be predetermined by analyzing the ratio of EEG signal to noise signal in the actual acquired EEG signal.

[0069] Step 340: Based on the simulated EEG signal, the target noise signal, the first weight, and the second weight, obtain the target simulated EEG signal.

[0070] Targeted simulated EEG signal acquisition refers to the simulated acquisition of EEG signals. Targeted simulated EEG signal acquisition can be used as a reference for EEG signal analysis and research, and to reduce the acquisition of real EEG signals. Just as real EEG signals contain both EEG signals and noise signals, targeted simulated EEG signal acquisition can also contain simulated EEG signals and targeted noise signals.

[0071] In some embodiments, the target simulated EEG signal acquisition can be determined by formula (1):

[0072] x(t)=αx s (t)+βx n (t) (1)

[0073] Where x(t) represents the target simulated acquisition of EEG signals, x s (t) represents the simulated electroencephalogram (EEG) signal, x n (t) represents the target noise signal, and α and β represent the first weight and the second weight, respectively.

[0074] Some embodiments in this specification, by setting a first parameter group and a second parameter group, allow for the determination of the influence of factors such as the region of interest in the virtual brain model, the signal source, and the electrode points installed on the virtual brain model on the signal, thereby obtaining more realistic target-simulated EEG signals. The simulated EEG signals, target noise signals, and target-acquired EEG signals obtained through these embodiments can all be used for EEG signal analysis and research, avoiding the problem of insufficient research data in EEG signal analysis and research. Simultaneously, target-acquired EEG signals can also reduce the acquisition of real EEG signals, lowering the cost of acquiring real EEG signals in EEG signal analysis and research.

[0075] Figure 4 This is an exemplary flowchart illustrating the acquisition of noise-free simulated electroencephalogram (EEG) signals according to some embodiments of this specification. In some embodiments, process 400 may be executed by the first acquisition module 210. Figure 4 As shown, process 400 may include the following steps:

[0076] Step 410: For each signal source in the virtual brain model, determine the target basic signal of the signal source based on the region of interest transfer matrix and the basic signal emitted by the signal source.

[0077] The target base signal can be the signal generated by a base signal source after being affected by other regions of interest.

[0078] In some embodiments, the first parameter group may include a region-of-interest (ROI) transfer matrix. The elements of the ROI transfer matrix characterize the transfer relationships when a signal propagates between different ROIs. The ROI transfer matrix can be an n*n matrix, where n represents the number of ROIs. In some embodiments, the elements of the ROI transfer matrix can be real numbers, which are adjustment parameters between the corresponding ROIs and can be used to determine the transfer relationships when a signal propagates between the corresponding ROIs. For example, an element of the ROI transfer matrix can be a real number α, and the target base signal can be obtained based on the following formula:

[0079]

[0080] in, This represents the target fundamental signal of the i-th signal source in the j-th region of interest. This represents the fundamental signal emitted by the signal source, where α is the adjustment parameter for the signal to propagate from the l-th region of interest to the j-th region of interest. This represents the fundamental signal emitted by the i-th signal source in the l-th region of interest.

[0081] In some embodiments, the elements of the region-of-interest (ROI) transfer matrix can also be transfer functions, which characterize the influence of the corresponding ROI on the signal. These transfer functions can be linear or nonlinear. For information on how to obtain the ROI transfer matrix, please refer to [link to relevant documentation]. Figure 6 And its related descriptions.

[0082] In some embodiments, the target base signal can be obtained based on the following formula:

[0083]

[0084]

[0085] in, This represents the target fundamental signal of the i-th signal source in the j-th region of interest. This indicates the basic signal emitted by the signal source. This represents the influence of other regions of interest on the signal source, where T represents the region of interest transfer matrix. lThis represents the l-th row in the region of interest transition matrix. It can be the average of the signals emitted by m signal sources in the l-th region of interest. This represents the fundamental signal emitted by the i-th signal source in the l-th region of interest.

[0086] Step 420: For each region of interest in the virtual brain model, based on the target base signal of each signal source in the region of interest and the signal source transfer matrix, determine the initial sub-simulated EEG signal corresponding to the region of interest.

[0087] The initial sub-simulated EEG signal can be a signal emitted from a region of interest in a virtual brain model. The initial sub-simulated EEG signal can also be a signal formed by superimposing the target baseline signals from multiple signal sources within that region of interest under the influence of electrode points.

[0088] In some embodiments, the first parameter group may further include a signal source transfer matrix. The elements of the signal source transfer matrix can characterize the transfer relationship during propagation between a signal source and an electrode point within the same region of interest. The signal source transfer matrix can be an m*k matrix, where m represents the number of signal sources in the region of interest, and k represents the number of electrode points in the virtual brain model. Similar to the region of interest transfer matrix, the elements of the signal source transfer matrix can be real numbers, which are adjustment parameters between the corresponding signal source and electrode point, and can be used to determine the transfer relationship during signal propagation between the corresponding signal source and electrode point. The elements of the signal source transfer matrix can also be transfer functions, which can characterize the influence of the corresponding electrode point on the target basic signal of the corresponding signal source. For information on how to obtain the signal source transfer matrix, please refer to [link to relevant documentation]. Figure 6 And its related descriptions.

[0089] In some embodiments, the initial sub-simulated EEG signal can be obtained based on the following formula:

[0090]

[0091] in, This represents the initial sub-simulated EEG signal for the j-th region of interest. Let A represent the target fundamental signal of the i-th signal source in the j-th region of interest, and let A represent the signal source transfer matrix. represents the i-th row in the signal source transfer matrix, and m represents the number of signal sources in the region of interest.

[0092] Step 430: Obtain simulated EEG signals based on the initial sub-simulated EEG signals corresponding to each region of interest.

[0093] In some embodiments, simulated EEG signals can be obtained by superimposing initial sub-simulated EEG signals corresponding to each region of interest. The simulated EEG signals can be obtained based on the following formula:

[0094]

[0095] Among them, X s (t) represents the simulated electroencephalogram (EEG) signal. The initial simulated EEG signal for the j-th region of interest is represented by n, where n represents the number of regions of interest.

[0096] Figure 5 This is an exemplary flowchart illustrating the acquisition of a target noise signal according to some embodiments of this specification. In some embodiments, process 500 may be executed by a second acquisition module 220. Figure 5 As shown, process 500 may include the following steps:

[0097] Step 510: Determine the first sub-noise signal based on the region of interest transfer matrix and the simulated brain noise signal.

[0098] The initial noise signal can include simulated brain noise signals and electrode point noise signals. Simulated brain noise signals refer to noise signals within the human brain in simulated acquired EEG signals, while electrode point noise signals refer to noise signals from the electrodes in the acquired EEG signals. Simulated brain noise signals can include background noise signals and pink noise signals. Background noise signals and pink noise signals can be obtained through historical data or the network.

[0099] In some embodiments, the second parameter group may also include a region of interest transfer matrix.

[0100] The first sub-noise signal can be a noise signal determined based on the background noise signal and the pink noise signal. The first sub-noise signal can be determined based on the following formula:

[0101]

[0102]

[0103]

[0104] in, Let A represent the noise signal introduced by the background noise signal in the first sub-noise signal, and let A represent the signal source transfer matrix. This represents the i-th row of the signal source transfer matrix, where m represents the number of signal sources in the region of interest, and n represents the number of regions of interest in the virtual brain model. x represents the background noise signal emitted by the i-th signal source in the j-th region of interest; 1 / f (t) represents the noise signal introduced by the pink noise signal in the first sub-noise signal, and f represents the frequency of the pink noise signal. x1(t) represents the pink noise signal emitted by the i-th signal source in the j-th region of interest; x1(t) represents the first sub-noise signal; θ0 and θ1 represent the weight parameters corresponding to the background noise signal and the noise signal brought by the pink noise signal in the first sub-noise signal, respectively, which can be determined according to the preset settings.

[0105] Step 520: Determine the second sub-noise signal based on the electrode point transfer matrix and the electrode point noise signal.

[0106] In some embodiments, the second parameter set may include an electrode transfer matrix. The elements of the electrode transfer matrix characterize the transfer relationships of signals propagating between different electrode points. The electrode transfer matrix can be a k*k matrix, where k is the number of electrode points in the virtual brain model. Similar to the region-of-interest transfer matrix, the elements of the electrode transfer matrix can be real numbers, which represent adjustment parameters between the corresponding electrode points and can be used to determine the transfer relationships of signals propagating between the corresponding electrode points. The elements of the electrode transfer matrix can also be conduction functions, which characterize the influence of the corresponding electrode point on the electrode point noise signal.

[0107] Each element in the matrix can be a transfer function, which characterizes the influence of the corresponding motor point on the electrode point's noise signal. It should be understood that elements located on the diagonal of the electrode point transfer matrix can be 1, indicating that the electrode point has no effect on itself. For more information on obtaining the electrode point transfer matrix, please refer to [link to documentation / reference]. Figure 8 And its related descriptions.

[0108] The second sub-noise signal can be a noise signal determined based on the electrode point noise signal. In some embodiments, the second sub-noise signal can be determined according to the following formula:

[0109]

[0110] x2(t)=θ2ε(t) (11) where ε(t) represents the noise signal brought by the electrode point noise signal in the second sub-noise signal, σ p (t) represents the electrode noise signal at the p-th electrode point, and H represents the electrode transfer matrix. pθ2 represents the p-th row in the electrode point transfer matrix, k represents the number of electrode points; x2(t) represents the second sub-noise signal, and θ2 represents the weighting parameter of the noise signal brought by the electrode point noise signal in the second sub-noise signal, which can be determined by presetting.

[0111] Step 530: Obtain the target noise signal based on the first sub-noise signal and the second sub-noise signal.

[0112] In some embodiments, the target noise signal can be determined according to the following formula:

[0113] x n (t)=x1(t)+x2(t) (12)

[0114] Where, x n x1(t) represents the target noise signal, x2(t) represents the first sub-noise signal, and x2(t) represents the second sub-noise signal.

[0115] Some embodiments in this specification, by setting up region of interest transfer matrices, signal source transfer matrices, and electrode point transfer matrices, can clearly define the influence of factors such as signal source, region of interest, and electrode points on the signal, making the acquired simulated EEG signals and target noise signals more realistic and more consistent with the actual acquired EEG signals.

[0116] It is worth noting that in modern technology, machine learning models, specifically denoising models, can be used to denoise acquired EEG signals, resulting in noise-free EEG signals. However, this requires a large amount of training data to train the denoising model, including both the acquired EEG signals and the denoised acquired EEG signals. In reality, available EEG signal data is scarce and costly to acquire. Furthermore, since the inherent EEG signals in the human body are unknown, the noise signals within the acquired EEG signals are also unknown, making denoising impossible. Therefore, obtaining the aforementioned training data is difficult in practice. Some embodiments in this specification can generate a large number of target simulated acquired EEG signals and noise-free simulated EEG signals to simulate the acquired and denoised acquired EEG signals respectively. This data can then be used as training data to train the denoising model, resulting in a better generalization effect and more accurate output from the trained denoising model.

[0117] Figure 6 This is an exemplary flowchart illustrating the acquisition of a first parameter set according to some embodiments of this specification. In some embodiments, process 600 may be executed by server 110 or first acquisition module 210. Figure 6 As shown, process 600 may include the following steps:

[0118] Step 610: Obtain the initial first parameter set.

[0119] The initial first parameter set may refer to the parameter set corresponding to a first preset parameter range. Once the first preset parameter range is determined, the corresponding initial first parameter set can be determined. In some embodiments, the first parameter set may include a region of interest transfer matrix and a signal source transfer matrix; correspondingly, the initial first parameter set may include an initial region of interest transfer matrix and an initial signal source transfer matrix.

[0120] Step 620: Based on the initial first parameter set and the basic signal, obtain the intermediate simulated EEG signal.

[0121] Intermediate EEG signals can refer to simulated human EEG signals obtained by processing basic signals using an initial set of first parameters. In some embodiments, intermediate EEG signals can be obtained by processing basic signals based on the initial set of first parameters. The method for processing basic signals based on the initial set of first parameters can be found in the section on obtaining simulated EEG signals by processing basic signals based on the first set of first parameters in this specification. For example, the method for processing basic signals based on the initial set of first parameters can be found in... Figure 4 In formulas (2) to (5), the region of interest transfer matrix in the formula is replaced with the initial region of interest transfer matrix, and the signal source transfer matrix is ​​replaced with the initial signal source transfer matrix. The final signal can be an intermediate simulated EEG signal.

[0122] Step 630: Obtain reference EEG signal.

[0123] Reference EEG signals refer to noise-free EEG signals in the human body. Reference EEG signals can be obtained from historically collected human EEG signals or through the internet.

[0124] Step 640: Construct a loss function based on intermediate simulated EEG signals and reference EEG signals.

[0125] Step 650: Iteratively update the parameters of the initial first parameter group based on the loss function until the preset conditions are met, and obtain the first parameter group.

[0126] In some embodiments, the parameters of the initial first parameter set can be iteratively updated based on the loss function. The first parameter set is obtained when the loss function of the initial first parameter set and the reference EEG signal satisfies a preset condition. The preset condition may be that the difference between the intermediate simulated signal and the reference EEG signal is less than a threshold. Other preset conditions may also be present, such as loss function convergence or the number of iterations reaching a threshold. The method for updating the parameters of the initial first parameter set can be gradient descent. In some embodiments, a regularization term can also be added to the loss function to improve iteration efficiency and accelerate loss function convergence.

[0127] In some embodiments, an intermediate simulated EEG signal can be obtained based on an initial region of interest transfer matrix, an initial signal source transfer matrix, and a base signal. A loss function is constructed based on the intermediate simulated EEG signal and a reference EEG signal. The parameters of the initial region of interest transfer matrix and the initial signal source transfer matrix are iteratively updated based on the loss function. When the loss function meets a preset condition, the specific parameters in the region of interest transfer matrix and the signal source transfer matrix can be determined.

[0128] In some embodiments, the loss function may include a first loss term and a second loss term. The first loss term reflects the relationship between the initial region of interest (ROI) transfer matrix and the base signal emitted by the signal source, while the second loss term reflects the relationship between the initial signal source transfer matrix and the target base signal of each signal source within the ROI. The first and second loss terms can be combined in various ways. For example, weighted summation can be used, and different weights can be assigned to the first and second loss terms to reflect the different influences of various factors on the EEG signal.

[0129] Some embodiments of this specification can pre-set a large parameter range for the initial parameter set, namely the first preset parameter range, which can avoid more computation. Iterative updates based on this parameter range can obtain iterative calculation results more quickly, thereby obtaining the first parameter set.

[0130] Similar to obtaining the first parameter set based on a first preset parameter range, the second parameter set can be obtained based on a second preset parameter range. That is, an initial second parameter set corresponding to the second preset parameter range can be obtained, and an intermediate noise signal can be obtained based on the initial second parameter set and the initial noise signal; a reference noise signal can be obtained; a loss function can be constructed based on the intermediate noise signal and the reference noise signal; the initial second parameter set can be iteratively updated based on the loss function until a preset condition is met, thus obtaining the second parameter set. Here, the intermediate noise signal can refer to the noise signal in the simulated acquired EEG signal obtained after processing the initial noise signal using the initial second parameter set, and the reference noise signal can refer to the noise signal in the actual acquired EEG signal. More details about the above embodiments can be found in [reference needed]. Figure 6 Related descriptions.

[0131] Signal transfer relationships can be used to represent the transfer relationships of signals as they propagate between different electrode points. In some embodiments, signal transfer relationships may include electrode point transfer matrices, also known as signal transfer matrices. Signal transfer relationships may also include signal transfer models. In some embodiments, noise reduction processing can be performed on acquired EEG signals based on signal transfer relationships.

[0132] Figure 7This is an exemplary flowchart of a method for denoising acquired electroencephalogram (EEG) signals according to some embodiments of this specification. In some embodiments, process 700 may be executed by server 110. Figure 7 As shown, process 700 may include the following steps:

[0133] Step 710: Determine the signal transfer relationship of the object under test, wherein the signal transfer relationship characterizes the signal change relationship when the initial noise signal propagates between various locations of the object under test. In some embodiments, step 710 may be performed by the first determining module.

[0134] The object under test can refer to the object whose signal transfer relationship needs to be determined. The object under test can be a living organism, such as the head region of a human body. The object under test can also be a non-living organism, such as a virtual head model.

[0135] Initial noise signal refers to the noise signal in the acquired EEG signal before it has propagated. Acquired EEG signal refers to the EEG signal of the subject acquired through electrode points.

[0136] It should be understood that various factors (e.g., the operation of the EEG acquisition device or the physiological activities of the subject) can affect the acquisition process, resulting in various noise signals in the acquired EEG signals. Common noise signals in acquired EEG signals can include non-biological artifacts and biological artifacts. Non-biological artifacts can be noise signals from the external environment, equipment, etc., including but not limited to noise signals from the EEG acquisition device and noise signals from mains power interference. Non-biological artifacts can be obtained by consulting relevant materials. For example, the noise signals generated by a particular model of EEG acquisition device can be determined online. Biological artifacts can be noise signals generated by the physiological activities of the subject, including but not limited to electrocardiogram signals from heartbeats, electromyogram signals from muscle activity, and electrooculogram signals from eye movements. Biological artifacts can be acquired by acquiring the subject using relevant acquisition equipment. Due to individual differences, the biological artifact signals corresponding to different subjects should be different.

[0137] Signal transfer relationships can be used to determine how an initial noise signal changes as it propagates between different locations on a test object. In some embodiments, signal transfer relationships can be used to determine how the signal changes as it propagates from one location on the test object to another. For example, based on signal transfer relationships, the changes from position O on the test object can be determined. i A certain incoming signal S i Propagated to the location O of the object being measured j Then, it changed into signal S. jIn some embodiments, due to differences in the physiological conditions and brain structures of the test subjects, the signal transfer relationships differ among different test subjects. For example, the signal transfer relationships differ between the elderly and children. Similarly, the signal transfer relationships differ between a 26-year-old woman and a 26-year-old man.

[0138] In some embodiments, the signal transfer relationship can be represented as an N*N signal transfer matrix, where N is the number of electrodes on the object under test, and each electrode can represent a different position on the object under test. Each element in the signal transfer matrix can be the conduction function between corresponding two positions on the object under test. It should be understood that the conduction function between a certain position and itself in the signal transfer matrix can be 1, indicating that the function itself does not change when the signal does not propagate. In some embodiments, more information on obtaining the signal transfer matrix can be found in [link to documentation]. Figure 4 And its related descriptions.

[0139] In some embodiments, the signal transfer relationship can also be represented as a signal transfer model. The initial noise signal can be processed based on the signal transfer model to obtain the target noise signal. More information on signal transfer models can be found in [link to relevant documentation]. Figure 5 And its related descriptions.

[0140] In some embodiments, when the object under test is a virtual head model, the attenuation relationship of the signal at various positions in the virtual head model can be determined based on the material and shape of the virtual head model, thereby determining the signal transfer relationship of the virtual head model.

[0141] In some embodiments, an initial test signal may be input from a first location of the subject, wherein the order of magnitude of the initial test signal is larger than the order of magnitude of the subject's EEG signal; a signal may be acquired from a second location of the subject to obtain an acquired test signal; and a signal transfer relationship may be determined based on the initial test signal and the acquired test signal. For more details on the above embodiments, please refer to... Figure 4 And its related descriptions.

[0142] Step 720: Determine the target noise signal based on the initial noise signal and the signal transfer relationship. In some embodiments, step 720 may be performed by a second determining module.

[0143] The target noise signal can refer to the noise signal in the acquired electroencephalogram (EEG) signal. In some embodiments, the target noise signal can be a noise set, which includes noise signals acquired at various locations of the subject.

[0144] In some embodiments, after determining the initial noise signal, the basic characteristics corresponding to each initial noise signal can be determined. These basic characteristics may include, but are not limited to, the signal amplitude and variance. For each of the multiple electrode points, during EEG signal acquisition, the signal acquired at that electrode point can be dynamically evaluated. By comparing the basic characteristics of each acquired signal with the basic characteristics corresponding to each initial noise signal, it can be determined whether an initial noise signal exists in the signal acquired at that electrode point. When an initial noise signal is present in the signal acquired at that electrode point, it can be identified as the target noise signal in the EEG signal acquired at that electrode point. Simultaneously, the propagated initial noise signal acquired at other electrode points can be determined through signal transfer relationships, and the propagated initial noise signal can be identified as the target noise signal in the EEG signals acquired at other electrode points.

[0145] Step 730: Remove target noise signals from the acquired EEG signals collected from the subject to obtain noise-free target EEG signals. In some embodiments, step 730 may be performed by a signal processing module.

[0146] The target EEG signal refers to the noise-free EEG signal of the subject being tested. The target EEG signal can be used to analyze and diagnose brain diseases in the subject being tested.

[0147] In some embodiments, when the target noise signal acquired by a certain electrode point is the initial noise signal that has not been propagated at that electrode point, the initial noise signal can be directly removed from the acquired EEG signal acquired by that electrode point to obtain the target EEG signal of that electrode point.

[0148] In some embodiments, when the target noise signal acquired at a certain electrode point is the initial noise signal after propagation, the target EEG signal at that electrode point can be determined based on the following formula:

[0149]

[0150] Among them, S clean S represents the target EEG signal at electrode point j. obs The EEG signal is collected at this electrode point, and h is the signal transfer matrix. ij S represents the transfer function in the signal transfer matrix that propagates the signal from position i to position j. i Let be the initial noise signal originating from the electrode point at position i, and k be the number of initial noise signals.

[0151] In some embodiments, the initial noise signal at a certain location can be processed based on a signal transfer model to obtain the target noise signal from the acquired EEG signals at other locations. Then, the target noise signal is removed from the acquired EEG signals at other locations to obtain the target EEG signal at the corresponding location.

[0152] In some embodiments of this specification, by determining the signal transfer relationships as the signal propagates between different locations, the target noise signal after propagation can be accurately and quickly determined based on the initial noise signal and the signal transfer relationships. This allows for the removal of the target noise signal from the acquired EEG signal, resulting in a noise-free target EEG signal. Compared to existing methods for denoising acquired EEG signals, the cost is significantly reduced. Furthermore, determining the corresponding signal transfer relationships for each subject ensures that the determined signal transfer relationships are more accurate and reflect the actual situation of the subject. This approach accurately and completely removes the target noise signal from the acquired EEG signal without removing the target EEG signal itself.

[0153] Figure 8 This is an exemplary flowchart illustrating the acquisition of signal transfer relationships according to some embodiments of this specification. In some embodiments, process 800 may be executed by a first determining module. Figure 8 As shown, process 800 may include the following steps:

[0154] Step 810: For each location of the subject being tested, an initial test signal is input from that location, wherein the order of magnitude of the initial test signal is greater than the order of magnitude of the subject's EEG signal.

[0155] The initial test signal can refer to the test signal input to the test subject. The initial test signal can be used to test the changes in the signal as it propagates between different electrode points. In some embodiments, the order of magnitude of the initial test signal is larger than that of the human EEG signal. For example, the order of magnitude of the human EEG signal is μV, and the order of magnitude of the initial test signal can be three orders of magnitude larger than the human EEG signal, i.e., the selected initial test signal can be of the order of magnitude of mV. In some embodiments, the initial test signal can include multiple signals of different frequencies and intensities. In some embodiments, the initial test signal can be determined by pre-setting. For example, the initial test signal can be a pre-set pulse signal, the order of magnitude of which is larger than that of the EEG signal of the test subject. Each location of the test subject can refer to each electrode point of the test subject, and the initial test signal can be input at one electrode point of the test subject.

[0156] It should be understood that because the initial test signal is on a much larger scale than the EEG signal of the subject, the acquired signal is less affected by the EEG signal, thus ensuring the accuracy of the test results. At the same time, the initial test signal should not exceed the magnitude of the signal that the human body can tolerate to ensure the safety of the subject; for example, the amplitude range of the initial test signal can be 20mV to 200mV.

[0157] Step 820: Collect signals from other locations of the object under test to obtain the collected test signals.

[0158] Acquiring a test signal can refer to the signal obtained from the propagation of the initial test signal input at the aforementioned location, acquired at other locations. These other locations can be other electrode points. Correspondingly, acquiring a test signal can include signals acquired at other electrode points based on those locations. For example, electrode points A to D can be set on the object under test, where the initial test signal is input at electrode point A, and the other locations can refer to motor points B, C, and D. Correspondingly, signals can be acquired from motor points B, C, and D to obtain the acquired test signal.

[0159] It should be understood that the initial test signal will be affected by other electrode points when it propagates between the electrode points, so the acquired test signal may be different from the initial test signal.

[0160] Step 830: Determine the signal transfer relationship based on the initial test signal and the acquired test signal.

[0161] In some embodiments, modeling or various data analysis algorithms, such as regression analysis and discriminant analysis, can be used to analyze and process the initial test signal and the acquired test signal to determine the signal transfer relationship.

[0162] In some embodiments, when the elements of the signal transfer matrix can be real numbers, the elements of the matrix can be determined based on the following formula:

[0163] S j (t)=βs i (t) (14)

[0164] Among them, S j (t) represents the acquired test signal collected at electrode point j, β is the adjustment parameter for the signal propagation from electrode point i to electrode point J, and s i (t) represents the initial test signal input from electrode point i.

[0165] In some embodiments, the signal transfer relationship may include multiple simple signal transfer relationships. A simple signal transfer relationship can refer to the relationship of a simple signal with relatively regular signal changes as it propagates between multiple locations of the object under test. In some embodiments, the input initial test signal can be transformed into multiple initial simple test signals of different frequencies and intensities, and the acquired test signals collected at other electrode points can be transformed into multiple acquired simple test signals of different frequencies and intensities. For each initial simple test signal, based on the initial simple test signal and its corresponding acquired simple test signal, a simple signal transfer relationship corresponding to the initial simple test signal is determined. These simple signal transfer relationships are then synthesized to obtain a signal transfer relationship containing multiple simple signal transfer relationships. The initial simple test signal can be a simple signal obtained after transforming the initial test signal, and the acquired simple test signal can be a simple signal obtained after transforming the acquired test signal. The signal transformation method can be Fourier transform. For example, both the initial simple test signal and the acquired simple test signal can be sinusoidal signals.

[0166] In some embodiments, the initial test signal input at position i can be Fourier transformed to obtain a set S of M initial simple test signals with different phases and frequencies, where S is an N*M matrix and N is the number of electrode points. When no signal is input at other locations on the test object and the initial test signal has not been propagated, each element in the i-th row of S represents a separate initial simple test signal, and other positions in S can be 0, indicating that no signal is present at other locations. After the initial test signal propagates, Fourier transforms are performed on the acquired test signals at other locations to obtain a set S′ containing multiple acquired simple test signals, where S′ is also an N*M matrix, and each element in S′ represents the acquired simple test signal acquired at the corresponding electrode point, which is the propagated signal of the corresponding initial simple test signal. Then, based on S and S′, the signal transfer matrix H can be calculated to ensure that the following formula holds:

[0167] S′=HS (15)

[0168] Here, the elements of H are all transfer functions, representing the changes in a signal as it propagates at a corresponding location. For example, h ij The elements in H can represent the propagation function of the signal from position i to position j. In some embodiments, the propagation function can take various forms. As shown in the following formula, the propagation function h... ij This can be equivalent to an FIR (Finite Impulse Response) filter:

[0169] y[n] = b0x[n] + b1x[n-1] + ... + b L x[nL] (16)

[0170]

[0171] Where y[n] is the acquired simple test signal, x[n] is the initial simple test signal, n is the signal length, and L is the number of parameters, which can be determined in advance. When L≤M, the aforementioned transfer function h can be determined by solving a system of linear equations. ij All parameters in.

[0172] In some embodiments, the signal transfer model can also be trained based on the initial test signal and the acquired test signal to obtain a trained signal transfer model. More information about signal transfer models can be found in [link to relevant documentation]. Figure 9 And its related descriptions.

[0173] It should be understood that the brain structure of the test subject is almost fixed; therefore, even if the intensities of two signals are different, the signal change relationship during propagation between the locations of the test subject should be the same. Thus, some embodiments of this specification can determine the signal transfer relationship of the test subject by analyzing and processing the initial test signal and the acquired test signal. This signal transfer relationship can also characterize the signal change relationship when the initial noise signal propagates between various locations of the test subject.

[0174] Some embodiments in this specification, for each test object, can quickly determine the signal transfer relationship of the test object through the initial test signal and the acquired test signal, ensuring the accuracy of the signal transfer relationship, thereby avoiding errors caused by individual differences of the test object during noise reduction.

[0175] Figure 9 This is a schematic diagram of a signal transfer model shown in some embodiments of this specification.

[0176] In some embodiments, the initial noise signal can be processed based on a signal transfer model to obtain the target noise signal. For example... Figure 9 As shown, the input to the signal transfer model 930 may include an initial noise signal 910 and the position 920 of the initial noise signal, and the output may include a target noise signal 940. In some embodiments, the signal transfer model may include, but is not limited to, a deep neural network model, a support vector machine model, etc.

[0177] In some embodiments, the signal transfer model can be obtained by training a machine learning model using training samples. For example... Figure 9The signal transfer model can be obtained by training the initial signal transfer model 950 using training samples 960 and labels 970. The initial signal transfer model can be a signal transfer model without set parameters. The training samples 960 may include the initial test signal and its location, and the labels 970 may include the acquired test signal. The method for obtaining the training samples and labels can be found in [reference needed]. Figure 8 The training process involves inputting multiple sets of training samples 960, each labeled 970, into an initial signal transfer model 950. A loss function is constructed based on the output of the initial signal transfer model 950 and the labels 970. The parameters of the initial signal transfer model 950 are iteratively updated based on the loss function until preset conditions are met, at which point training ends, and the trained signal transfer model 930 is obtained. Preset conditions may include, but are not limited to, the loss function being less than a threshold, convergence, or the training period reaching a threshold.

[0178] In some embodiments, the signal transfer model may include an object information embedding layer and a signal determination layer. The object information embedding layer is used to extract features from the shape and hair information of the object under test to obtain head features. The signal determination layer is used to process the head features, the initial noise signal, and the position information of the initial noise signal to obtain the target noise signal. The shape information may include, but is not limited to, the shape and size of the object under test, and the hair information may include, but is not limited to, the hair length, hair hardness, hair curvature, and hair oiliness of the object under test. The shape and hair information of the object under test can be determined through pre-setting.

[0179] like Figure 10 As shown, the signal transfer model 930 may further include an object information embedding layer 930-1 and a signal determination layer 930-2 connected in sequence. The input to the object information embedding layer 930-1 may include shape information 1010 and hair information 1020, and the output is head features 931. The input to the signal determination layer may include head features 931, an initial noise signal 910, and the position 920 of the initial noise signal, and the output may be the target noise signal 940. The object information embedding layer may be a Naive Bayes model, and the signal determination layer may be a deep neural network model.

[0180] In some embodiments, an object information embedding layer can be obtained through training: training samples may include historical shape information and historical hair condition information of the test object, and labels may include the historical head features of the aforementioned test object. The test object can refer to the object used to acquire its data for training the object information embedding layer. Similar to the test object, the test object can be a living or non-living object. The aforementioned training samples and labels can be obtained by analyzing relevant information about manually described test objects. Multiple sets of training samples can be input into the initial object information embedding layer. A loss function is constructed based on the output of the initial object information embedding layer and the labels. The parameters of the initial object information embedding layer are iteratively updated based on the loss function until a preset condition is met, thus obtaining a trained object information embedding layer. The preset condition may include, but is not limited to, the loss function being less than a threshold, convergence, or the training period reaching a threshold.

[0181] In some embodiments, a signal determination layer can be obtained through training: training samples may include historical head features, the initial test signal of the tested object, and the location of the initial test signal; labels may include the collected test signal. The methods for obtaining the aforementioned training samples and labels can be found in the preceding text of this specification. Multiple sets of training samples can be input into the initial signal determination layer. A loss function is constructed based on the output of the initial signal determination layer and the labels. The parameters of the initial signal determination layer are iteratively updated based on the loss function until preset conditions are met, thus obtaining a trained signal determination layer. Preset conditions may include, but are not limited to, the loss function being less than a threshold, convergence, or the training period reaching a threshold.

[0182] It should be understood that the training data used to train the object information embedding layer and the signal determination layer correspond to different objects. The trained object information embedding layer can extract head features from all objects (including test objects and objects under test). Since the object information embedding layer is a universal model for all objects under test, it can extract head features from different objects under test by training on relevant data from the test object. The trained signal determination layer, however, is used to determine the signal transition relationship of a specific object under test. As mentioned earlier, the corresponding signal transition relationship differs depending on the object under test. Therefore, the signal determination layer trained on a specific object under test can only be used to determine the signal transition relationship of that object. When it is necessary to determine the signal transition relationship of another object under test, training can be performed based on the trained object information embedding layer, using the initial test signal of that object under test, its location, and the acquired test signal.

[0183] Some embodiments in this specification utilize signal transfer models to more quickly and conveniently determine the signal transfer relationships of the tested subject, facilitating denoising of the acquired EEG signals. Furthermore, some embodiments in this specification can extract relevant features of the tested subject's head, determining the impact of differences in head-related information on the signal transfer relationships, thus making the denoising of the acquired EEG signals more accurate.

[0184] Figure 11 This is an exemplary flowchart illustrating the determination of a target noise signal according to some embodiments of this specification. In some embodiments, process 1100 may be executed by a second determining module.

[0185] In some embodiments, the signal transfer relationship may include multiple simple signal transfer relationships. A simple signal transfer relationship characterizes the signal change relationship as a first simple noise signal propagates between various locations of the object under test. More information on simple signal transfer relationships can be found at [link to relevant documentation]. Figure 8 And its related description. The first simple noise signal is obtained by transforming the initial noise signal.

[0186] When the signal transfer relationship includes multiple simple signal transfer relationships, process 1100 can be executed to determine the target noise signal. For example... Figure 11 As shown, process 1100 may include the following steps:

[0187] Step 1110: Transform the initial noise signal into multiple first simple noise signals with different frequencies and intensities.

[0188] The first simple noise signal can be a simple signal obtained by transforming the initial noise signal. For example, the first simple noise signal can include any one or a combination of sine wave signals, cosine wave signals, and square wave signals. The initial noise signal can be processed using Fourier transform, Laplace transform, and discrete cosine transform to obtain multiple first simple noise signals with different frequencies and intensities.

[0189] Step 1120: For each first simple noise signal, process the first simple noise signal based on the simple signal transfer relationship corresponding to the first simple noise signal to determine the second simple noise signal corresponding to the first simple noise signal.

[0190] The second simple noise signal can refer to the noise signal when the first simple noise signal propagates to other locations of the object being measured.

[0191] In some embodiments, for each first simple noise signal, a second simple noise signal corresponding to the first simple noise signal at other locations can be determined based on the first simple noise signal, its input position, and simple signal transfer relationships. The propagation function from the input position to other locations can be determined based on the simple signal transfer relationships and the input position of the first simple noise signal. Based on these propagation functions and the first simple noise signal, the second simple noise signal at each of the other locations can be determined.

[0192] Step 1130: Synthesize each of the second simple noise signals to obtain the target noise signal.

[0193] In some embodiments, for each location of the measured object, the various second simple noise signals at that location can be synthesized to obtain the target noise signal at that location. The synthesis method can be the inverse transform of the transformation method described above that processes the initial noise signal into a first simple noise signal to obtain the target noise signal. For example, when the initial noise signal is processed into a first simple noise signal based on Fourier transform, the various second simple noise signals can be synthesized based on inverse Fourier transform to obtain the target noise signal.

[0194] Some embodiments in this specification can conveniently obtain target noise signals by utilizing simple noise signals and simple signal transfer relationships, thereby simplifying calculations and improving computational efficiency.

[0195] This specification provides an apparatus for generating simulated electroencephalogram (EEG) signals, including a processor, which is used to execute the method for generating simulated EEG signals as described in any one of the embodiments of this specification.

[0196] This specification provides a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the method for generating simulated electroencephalogram (EEG) signals as described in any one of the embodiments of this specification.

[0197] This specification also provides an embodiment of a device for acquiring and denoising electroencephalogram (EEG) signals, including a processor, which is used to execute the method for acquiring and denoising EEG signals as described in any one of the embodiments of this specification.

[0198] This specification provides a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the method for denoising acquired EEG signals as described in any one of the embodiments of this specification.

[0199] It should be noted that the above descriptions of the various processes are for illustrative purposes only and do not limit the scope of this specification. Those skilled in the art can make various modifications and changes to the processes under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0200] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments of this specification. Furthermore, this specification uses specific terms to describe embodiments of this specification. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different locations in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0201] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of some embodiments of the invention that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments described herein. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0202] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0203] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may be changed depending on the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this specification are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

[0204] For each patent, patent application, patent application publication, and other material, such as articles, books, specifications, publications, and documents, referenced in this specification, the entire contents of which are incorporated herein by reference. This excludes historical application documents that are inconsistent with or conflict with the content of this specification, as well as documents that limit the broadest scope of the claims in this specification (currently or subsequently appended to this specification). It should be noted that in the event of any inconsistency or conflict between the descriptions, definitions, and / or terminology used in the supplementary materials to this specification and the content of this specification, the descriptions, definitions, and / or terminology used in this specification shall prevail.

[0205] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for generating simulated electroencephalogram (EEG) signals, characterized in that, include: Based on the basic signal and the first parameter set, a noise-free simulated EEG signal is obtained. The simulated EEG signal is used as a noise-free EEG signal in the human body to participate in the analysis and research of EEG signals. Based on the initial noise signal and the second parameter set, a target noise signal is obtained. This target noise signal is used as the noise signal in the collected EEG signals to participate in the analysis and research of the EEG signals. The second parameter set includes a region-of-interest transfer matrix and an electrode point transfer matrix. The initial noise signal includes simulated brain noise signals and electrode point noise signals. Obtaining the target noise signal based on the initial noise signal and the second parameter set includes: Based on the region of interest transfer matrix and the simulated brain noise signal, a first sub-noise signal is determined, wherein the elements in the region of interest transfer matrix represent the transfer relationship when the signal propagates between different regions of interest; Based on the electrode point transfer matrix and the electrode point noise signal, a second sub-noise signal is determined, wherein the elements in the electrode point transfer matrix represent the transfer relationship when the signal propagates between different electrode points. The target noise signal is obtained based on the first sub-noise signal and the second sub-noise signal; Obtain the first weight of the simulated EEG signal and the second weight of the target noise signal; Based on the simulated EEG signal, the target noise signal, the first weight, and the second weight, a target simulated EEG signal is obtained. The target simulated EEG signal is used as the acquired EEG signal to participate in the analysis and research of the EEG signal and to reduce the acquisition of real EEG signals.

2. The method as described in claim 1, characterized in that, The first parameter group includes the region of interest transfer matrix and the signal source transfer matrix. The elements in the signal source transfer matrix represent the transfer relationship between the signal source and the electrode point in the same region of interest during propagation. The different regions of interest represent different functional areas of the human brain. The process of obtaining noise-free simulated EEG signals based on the basic signal and the first parameter set includes: For each signal source in the virtual brain model, the target basic signal of the signal source is determined based on the region of interest transfer matrix and the basic signal emitted by the signal source. For each region of interest in the virtual brain model, based on the target base signal of each signal source in the region of interest and the signal source transfer matrix, the initial sub-simulated EEG signal corresponding to the region of interest is determined; The simulated EEG signals are obtained based on the initial sub-simulated EEG signals corresponding to each of the regions of interest.

3. The method as described in claim 1, characterized in that, The first parameter set is obtained based on a first preset parameter range, and the second parameter set is obtained based on a second preset parameter range.

4. The method as described in claim 1, characterized in that, The elements in the electrode point transfer matrix are transfer functions.

5. A system for generating simulated electroencephalogram (EEG) signals, characterized in that, The system includes: The first acquisition module is used to obtain noise-free simulated EEG signals based on the basic signal and the first parameter set. The simulated EEG signals are used as noise-free EEG signals in the human body to participate in the analysis and research of EEG signals. The second acquisition module is used to obtain a target noise signal based on an initial noise signal and a second parameter set. The target noise signal is used as a noise signal in the collected EEG signals from the human body for analysis and research of the EEG signals. The second parameter set includes a region-of-interest transfer matrix and an electrode point transfer matrix. The initial noise signal includes simulated brain noise signals and electrode point noise signals. The second acquisition module is further used for: Based on the region of interest transfer matrix and the simulated brain noise signal, a first sub-noise signal is determined, wherein, The elements in the region of interest transfer matrix represent the transfer relationship of the signal when it propagates between different regions of interest; Based on the electrode point transfer matrix and the electrode point noise signal, a second sub-noise signal is determined, wherein the elements in the electrode point transfer matrix represent the transfer relationship of the signal when it propagates between different electrode points; based on the first sub-noise signal and the second sub-noise signal, the target noise signal is obtained. The third acquisition module is used to acquire the first weight of the simulated EEG signal and the second weight of the target noise signal; The fourth acquisition module is used to obtain a target simulated acquired EEG signal based on the simulated EEG signal, the target noise signal, the first weight, and the second weight. The target simulated acquired EEG signal is used as the acquired EEG signal to participate in the analysis and research of the EEG signal and to reduce the acquisition of real EEG signals.

6. The system as described in claim 5, characterized in that, The first parameter group includes the region of interest transfer matrix and the signal source transfer matrix. The elements in the signal source transfer matrix represent the transfer relationship between the signal source and the electrode point in the same region of interest during propagation. The different regions of interest represent different functional areas of the human brain. The first acquisition module is further used for: For each signal source in the virtual brain model, the target basic signal of the signal source is determined based on the region of interest transfer matrix and the basic signal emitted by the signal source. For each region of interest in the virtual brain model, based on the target base signal of each signal source in the region of interest and the signal source transfer matrix, the initial sub-simulated EEG signal corresponding to the region of interest is determined; The simulated EEG signals are obtained based on the initial sub-simulated EEG signals corresponding to each of the regions of interest.

7. The system as described in claim 5, characterized in that, The first parameter set is obtained based on a first preset parameter range, and the second parameter set is obtained based on a second preset parameter range.

8. The system as described in claim 5, characterized in that, The elements in the electrode point transfer matrix are transfer functions.

9. A device for generating simulated electroencephalogram (EEG) signals, comprising a processor, characterized in that, The processor is used to execute the method for generating simulated electroencephalogram (EEG) signals as described in any one of claims 1 to 4.

10. A computer-readable storage medium storing computer instructions, characterized in that, After the computer reads the computer instructions in the storage medium, the computer executes the method for generating simulated electroencephalogram (EEG) signals as described in any one of claims 1 to 4.

Images