Method for assessing motor imagery states, implantable rehabilitation training device

Abstract

The present application belongs to the technical field of physiological electrical signals, and particularly relates to a method for evaluating motor imagery state and an implantable rehabilitation training device. First, test electroencephalogram signals of a user in the process of alternation of a "motor task" and a "rest task" are acquired, then a balanced data set is constructed by using the test electroencephalogram signals, and zero-point balance parameters are acquired based on the balanced data set, then the synthetic features of real-time electroencephalogram signals are acquired by using the zero-point balance parameters, and scaling coefficients are acquired by using motor state data in the test electroencephalogram signals; the evaluation index of the real-time electroencephalogram signals is acquired by combining the scaling coefficients and the synthetic features of the real-time electroencephalogram signals, and is used for evaluating the motor imagery state in rehabilitation training.

Description

Technical Field

[0001] This invention belongs to the field of physiological electrical signal technology, specifically relating to a method for assessing motor imagery and an implantable rehabilitation training device. Background Technology

[0002] Typically, implantable rehabilitation training devices include electrode sensors, a processor, and training peripherals, with the electrode sensors permanently implanted intracranially. In these technologies, patients engage in motor imagery according to a training task paradigm. The electrode sensors collect the electroencephalogram (EEG) signals during motor imagery, the processor receives these signals and interprets them into instructions for the training task, and the training peripherals execute the training actions according to these instructions. Because the EEG signals during motor imagery directly determine the decoding results and the execution results of the training actions, and because motor imagery is autonomously implemented and controlled by the patient, and because different patients have different motor imagery, and even the same patient's motor imagery varies in different states, it is difficult to uniformly assess a patient's motor imagery state during rehabilitation training. For example, in rehabilitation training, when the execution results of the training actions are inconsistent with the requirements of the training task, it is difficult to determine which part of the process went wrong, and it is impossible to further adjust the training plan. Summary of the Invention

[0003] This invention provides a method for assessing motor imagery and an implantable rehabilitation training device to solve the problem of the difficulty in uniformly assessing motor imagery in rehabilitation training.

[0004] To address the aforementioned technical problems, this invention provides a method for evaluating motor imagery states based on electroencephalogram (EEG) signals, comprising: constructing a balance dataset based on test EEG signals, including motion-state data and resting-state data; obtaining zero-point balance parameters based on the balance dataset; obtaining a scaling factor by combining the zero-point balance parameters and the motion-state data in the balance dataset; obtaining synthetic features of real-time EEG signals based on the zero-point balance parameters; and obtaining an evaluation index by combining the scaling factor and the synthetic features of real-time EEG signals.

[0005] Furthermore, constructing the balanced dataset includes: slicing the test EEG signals into data slices; labeling the motion and resting state data in the data slices; balancing the data slices to ensure that the number of motion and resting state data slices is the same; and using the motion and resting state data as samples for the balanced dataset.

[0006] Furthermore, obtaining the zero-point balance parameters includes: obtaining low-frequency and high-frequency samples from the balance dataset; and obtaining the mean of the average residuals of all low-frequency samples. ; Obtain the mean of the average residuals of all high-frequency band samples. ; Calculate the synthetic features of the balanced dataset ,Right now Obtain the zero-point equilibrium parameter, i.e., the mean. and mean Incorporating synthetic features The formula and let the expected value ,get .

[0007] Furthermore, obtaining the scaling factor includes: acquiring each motion data sample from the balanced dataset to construct a motion dataset; and calculating the synthetic features of each sample in the motion dataset. ,Right now ;in, This represents the average residual of each high-frequency band sample in the dynamic dataset. Represents the average residual of each low-frequency sample in the dynamic dataset; based on each synthetic feature Constructing dynamic synthetic feature sets ; Set dynamic synthetic feature set The value at the percentile x is denoted as ; Set the feedback intensity of the evaluation indicators ,and ; Use the inverse hyperbolic tangent function to obtain the scaling factor ,Right now .

[0008] Furthermore, obtaining the synthetic features of real-time EEG signals includes: constructing a real-time dataset based on real-time EEG signals; obtaining low-frequency and high-frequency samples from the real-time dataset; and obtaining the mean of the average residuals of the low-frequency samples. Mean of the average residuals based on high-frequency band samples ; Calculate the synthetic features of real-time EEG signals ,Right now The evaluation index is obtained by normalization using the hyperbolic tangent function, i.e. .

[0009] Furthermore, it also includes: assessing the state of motor imagery, that is, determining whether the assessment index of the real-time EEG signal is greater than zero; if so, the motor imagery is assessed as a motor state; if not, the motor imagery is assessed as a resting state.

[0010] The present invention also provides an implantable rehabilitation training device, comprising: an electrode sensor located intracranially for collecting real-time electroencephalogram (EEG) signals and sending them to a host computer; a host computer equipped with a processor for running the rehabilitation training method; a training peripheral located at the user's affected area for executing training tasks according to preset instructions; and the processor for judging the consistency between the motor imagery state and the training task to obtain feedback results.

[0011] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals.

[0012] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals.

[0013] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals.

[0014] The beneficial effects of this invention are as follows: The method and implantable rehabilitation training device for assessing motor imagery based on electroencephalogram (EEG) signals first acquire test EEG signals from the user during the alternation of "motor tasks" and "resting tasks." Then, a balanced dataset is constructed using the test EEG signals, and zero-point balance parameters are obtained based on the balanced dataset. The synthetic features of the real-time EEG signals are then obtained using the zero-point balance parameters, and a scaling factor is obtained using the motion data in the test EEG signals. Finally, an evaluation index of the real-time EEG signals is obtained by combining the scaling factor and the synthetic features of the real-time EEG signals, which is used to assess the motor imagery state during rehabilitation training. By balancing positive and negative samples, the model avoids bias towards the class with more samples. The evaluation index measures the user's motor imagery state during rehabilitation training, thereby making the feedback results more standardized and objective.

[0015] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. To make the foregoing objects, features, and advantages of the invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 It is a flowchart for assessing the state of motor imagery.

[0018] Figure 2 This is a flowchart for obtaining the evaluation index.

[0019] Figure 3This is a flowchart of rehabilitation training methods.

[0020] Figure 4 This is a schematic diagram of the principle of implantable rehabilitation training equipment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] See Figure 1 and Figure 2 At least one embodiment provides a method for assessing motor imagery states based on electroencephalogram (EEG) signals, comprising: constructing a balance dataset based on test EEG signals, including motion-state data and resting-state data; obtaining zero-point balance parameters based on the balance dataset; obtaining a scaling factor by combining the zero-point balance parameters and the motion-state data in the balance dataset; obtaining synthetic features of real-time EEG signals based on the zero-point balance parameters; and obtaining an evaluation index by combining the scaling factor and the synthetic features of real-time EEG signals. Specifically, it includes the following steps.

[0023] Step S1: Construct a category-balanced calibration dataset based on the test EEG signals.

[0024] Constructing a balanced dataset involves: slicing the EEG signals from rehabilitation training; labeling the motor and resting state data within the slices; balancing the data by ensuring that the number of slices for motor and resting state data is the same; and using the motor and resting state data as samples for the balanced dataset.

[0025] Specifically, testing protocols can be designed for rehabilitation training, followed by the collection and slicing of EEG signals. From these slices, two types of segments—"motor task" and "resting state"—corresponding to the training task are labeled and extracted: motor data and resting state data. Undersampling is performed on the larger sample size (usually resting state), while resampling is performed on the smaller sample size (usually motor state) to ensure that the sample sizes of the two classes in the balanced dataset are strictly consistent. This solves the parameter shift problem caused by excessive resting state data in traditional calibration, providing a data foundation for obtaining unbiased zero-point balance parameters.

[0026] Step S2: Test the decoupling of non-periodic components and extraction of pure spectrum residuals from EEG signals or real-time EEG signals.

[0027] Generally, the 1 / f signal can be understood as the non-periodic component of the EEG power spectrum, exhibiting a power-law decay in power with increasing frequency (approximately a straight line on a double logarithmic scale). It reflects background neural activity or instrument noise in the brain and is usually considered the "background" in physiological signal analysis. Different individuals have different "1 / f slopes" in their EEG power spectra. Even if two people have the same neural oscillation intensity, their absolute power ratios will differ significantly due to different background noise slopes, causing the universal model to fail. Simultaneously, changes in electrode impedance or environmental noise can alter the overall power level, leading to unstable feedback values ​​and an inability to guarantee stable low values ​​in the "resting state" and stable high values ​​in the "movement state." Therefore, the average residual of both real-time and test EEG signals needs to be fitted with background noise and the spectral residual calculated. By fitting and removing the 1 / f non-periodic component, the pure frequency band oscillation residual can be extracted, effectively eliminating the influence of individual differences in background noise and solving the problem of poor parameter generalization ability caused by individual differences in background EEG (1 / f signal). The specific processing includes steps S21 and S22.

[0028] Step S21, background noise fitting.

[0029] The EEG signal or balanced dataset is sliced ​​to obtain data time windows. Feature extraction is performed on each data time window, the power spectral density is calculated and transformed to a log-log spectral coordinate system, and a "1 / f noise" feature extraction strategy is adopted, that is, robust linear regression is used to fit the non-periodic background trend line of the EEG signal. In this process, frequency bands containing physiological oscillations (such as alpha waves and beta waves) and power frequency interference are automatically masked, and only the background frequency band is used to fit the slope and intercept of the background noise (i.e., fitting the 1 / f curve). This can effectively overcome the baseline drift problem caused by differences in skull thickness and electrode impedance between individuals and reduce the variability between subjects.

[0030] Step S22: Calculate the spectral residual.

[0031] Generally, the low-frequency band can be the μ-band or β-band, and the high-frequency band can be the High-gamma band. The spectral residual can be understood as the periodic / oscillatory component remaining after subtracting the fitted 1 / f non-periodic component from the original power spectrum, reflecting the true neural oscillation intensity of a specific frequency band (such as beta waves or High-gamma waves). By subtracting the fitted background trend line (i.e., the 1 / f curve) from the original power spectrum, the "spectral residual" unaffected by the baseline level is extracted. Then, the average residual of the μ-band and β-band (characterizing desynchronization) is extracted. Its global mean is denoted as Extract the average residual from the High-γ band (characterizing synchronization). Its global mean is denoted as The average residual or its mean is used as the activation indicator for the actual execution of motion imagination.

[0032] Step S3: Obtain zero-bias automatic calibration of global statistics based on the test EEG signals.

[0033] Obtaining the zero-point balancing parameters includes: obtaining low-frequency and high-frequency samples from the balancing dataset; and obtaining the mean of the average residuals of all low-frequency samples. ; Obtain the mean of the average residuals of all high-frequency band samples. Constructing synthetic features for a balanced dataset ,Right now Obtain the zero-point equilibrium parameter, i.e., the mean. and mean Incorporating synthetic features The formula and let the expected value ,get The optimal γ value can be directly derived using the statistical mean of all similar samples in a balanced dataset. This automatically balances the contribution ratio of high-frequency enhancement and low-frequency suppression, forcing the synthesized features to be optimized. The mean is strictly zero in a globally statistical sense, thus achieving automatic alignment of the feedback zero point. In other words, this process can achieve feature synthesis. By centering the evaluation index at zero in a statistical sense, it is ensured that the average response of the evaluation index is neutral (i.e., zero) in the global (including dynamic and resting states), thus eliminating the offset caused by the different magnitudes of the frequency band energy itself.

[0034] Step S4: Obtain nonlinear dynamic mapping of quantile anchors based on the test EEG signals.

[0035] Obtaining the scaling factor includes: calculating the synthetic features of each sample in the motion dataset. ,Right now ;in, This represents the average residual of each high-frequency band sample in the dynamic dataset. Represents the average residual of each low-frequency sample in the dynamic dataset; based on each synthetic feature Constructing dynamic synthetic feature sets ; Set dynamic synthetic feature set The value at the percentile x is denoted as ; Set the feedback intensity of the evaluation indicators , ; Use the inverse hyperbolic tangent function to obtain the scaling factor ,Right now The scaling factor Precisely map high quantiles of features to feedback intensity This ensures that users with varying neural signal intensities receive standardized feedback during strenuous exercise. In practice, to guarantee users receive appropriately strong feedback (neither too weak nor too saturated) during exercise attempts, x can be set to the 90th percentile, and the expected feedback intensity can be defined. It is 80% of full scale.

[0036] Step S5: Obtain the evaluation index of real-time EEG signals.

[0037] Acquiring synthetic features from real-time EEG signals includes: constructing a real-time dataset based on real-time EEG signals; acquiring low-frequency and high-frequency samples from the real-time dataset; and acquiring the mean of the average residuals of the low-frequency samples. Mean of the average residuals based on high-frequency band samples ; Constructing synthetic features of real-time EEG signals ,Right now The evaluation index is obtained by normalization using the hyperbolic tangent function (Tanh), i.e. Used to provide real-time feedback on the brain's motor / resting state and to assess indices. The range is usually between [-1, 1].

[0038] Step S6: Real-time assessment of motor imagery state using EEG signals.

[0039] Optionally, steps S1 to S5 can ensure the evaluation index. When the global average response (including both dynamic and resting states) is neutral (i.e., zero), assessing the state of motor imagery can be understood as judging whether the assessment index of the real-time EEG signal is greater than zero; if yes, then the motor imagery is assessed as dynamic; if no, then the motor imagery is assessed as resting.

[0040] In summary, this method for assessing motor imagery states based on electroencephalogram (EEG) signals has the following effects.

[0041] (1) Strong robustness to individual differences: By subtracting the background noise of the fit, feature extraction no longer depends on the absolute power value, but on the "degree of protrusion relative to the background". This means that even if different patients have different skull thicknesses and electrode impedances, as long as the brain produces physiological oscillations, the algorithm can accurately capture them without the need for manual baseline calibration.

[0042] (2) Automated and accurate parameter fitting: Compared with the traditional gradient descent iterative search for parameters, this method is faster and the result (evaluation index) is uniquely determined.

[0043] (3) Intuitive controllability: By mapping the high score points of the movement state to specific assessment indices and displaying them as a visual progress bar, patients are guided in real time. This ensures that when users try to exercise, the assessment index can be stably in the high score area, while it falls back to near zero value when resting. This deterministic mapping can greatly enhance the user's confidence in rehabilitation training.

[0044] In some embodiments, see Figure 4 Furthermore, an implantable rehabilitation training device is provided, comprising: an electrode sensor located intracranially for acquiring real-time electroencephalogram (EEG) signals and sending them to a host computer; a host computer equipped with a processor for running the rehabilitation training method; the processor being equipped with a data processing module and a feedback module; the data processing module being used to acquire an evaluation index of the real-time EEG signal, evaluate the motor imagery state corresponding to the real-time EEG signal based on the evaluation index, and determine the consistency between the motor imagery state and the training task, thereby obtaining feedback results of the real-time EEG signal; the feedback module being used to feed back the feedback results to a display module or a prompting module to prompt the patient to maintain the current motor imagery state or adjust the motor imagery state; and a training peripheral located at the user's affected area, which executes training tasks according to program instructions for rehabilitation training.

[0045] Specifically, the implantable rehabilitation training device includes structural components and hardware circuitry; the operating platform of the method is the host computer; the user interface of the human-computer interaction system is the display module; signal input is provided by electrode sensors, amplifiers, and analog-to-digital conversion modules. Details are as follows.

[0046] (1) Electrode sensor, which is implanted in the body for a long time to collect raw simulated EEG signals.

[0047] (2) Lower-level machine (MCU / ARM / DSP hardware + embedded software), storing / forwarding digital signals. This includes an in-vivo machine and an external machine. The in-vivo machine is wired to the electrode sensors to transmit digital signals. The external machine communicates wirelessly with the in-vivo machine via coils and is wirelessly powered. An amplifier module (hardware analog circuit) amplifies the original analog EEG signal; an analog-to-digital converter module (hardware) converts the amplified analog EEG signal into a digital signal at a fixed sampling rate, i.e., test EEG signal and real-time EEG signal. The amplifier module and analog-to-digital converter module can be located on the in-vivo machine or integrated on the electrode sensors.

[0048] (3) The host and its processor, including PC hardware + PC software or APP hardware + APP software, receive digital signals sent by the external unit. The processor on the host contains a data processing module and a feedback module, which run as software, process digital signals and display them in conjunction with the UI workflow on the display module.

[0049] Optionally, when the host computer includes a user terminal and a management terminal, the host can be located at either the user terminal or the management terminal. The user terminal or management terminal can be understood as a cloud platform, computer, server, tablet, mobile phone, etc., with functions such as storage, data exchange, and analysis. The method or rehabilitation training method based on EEG signals can run on computer software or an app. The data processing module and feedback module on it can be understood as the process of inputting parameters or options through the software's interface and executing corresponding method steps through program instructions.

[0050] (4) Display module: Works with the human-computer interaction system to present the UI operation interface. Users can operate the UI to set parameters, display evaluation indexes, display motion imagery status and feedback results, etc.

[0051] (5) Training peripherals can be understood as flexible wearable glove robots that perform training actions according to preset program instructions and are worn on the trainer's hands.

[0052] Specifically, when patients use implanted rehabilitation training devices, the peripheral device executes training actions according to preset program instructions for the training task, while the patient simultaneously engages in motor imagery. The current state of the training action, such as a kinetic or resting state, can be identified or marked by corresponding program instructions. If the program instructions remain unchanged, the state of the training action and its corresponding time point can be considered fixed. Therefore, as long as the training action and motor imagery are synchronized, the consistency between the motor imagery and the training task can be determined using the state of motor imagery at the same moment. For example, at a certain training moment, if the program instructions executed by the peripheral device determine that the current training action is in a kinetic state, and the motor imagery state obtained through real-time EEG signals is also in a kinetic state, the feedback result is to maintain the motor imagery state. Since the peripheral device executes training actions according to preset program instructions, rather than according to analyzed EEG signals, errors are generally rare, and errors in the training paradigm are easily identified. If a discrepancy occurs between the state of motor imagery and the state of the training task, it is easy to determine that the patient needs to adjust their motor imagery.

[0053] In some embodiments, a computer device is also provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals or steps of a rehabilitation training method.

[0054] The processor can be a central processing unit (CPU), an ASIC, or one or more integrated circuits configured to implement embodiments of the present invention. In specific implementations, if the memory and processor are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a PCI bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be categorized as an address bus, a data bus, a control bus, etc. If the memory and processor are integrated onto a single chip, they can communicate with each other through an internal interface.

[0055] In some embodiments, a computer-readable storage medium is also provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals or the steps of a rehabilitation training method. The storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0056] In some embodiments, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of a method for assessing motor imagery states based on electroencephalogram (EEG) signals or the steps of a rehabilitation training method. The computer program may include program code, which includes computer operation instructions and may be stored in a computer-readable storage medium. Based on this understanding, when the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be implemented in the form of a software product or sold or used as an independent product, the computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0057] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0058] Based on the above-described preferred embodiments of the present invention, and through the above description, those skilled in the art can make various changes and modifications without departing from the technical concept of the present invention. That is, the technical scope of the present invention is not limited to the contents of the specification.

Claims

1. A method for assessing motor imagery states based on electroencephalogram (EEG) signals, characterized in that, include: A balance dataset was constructed based on test EEG signals, including dynamic and resting state data; Obtain zero-point equilibrium parameters based on the equilibrium dataset; The scaling factor is obtained by combining the zero-point equilibrium parameters and the motion data in the equilibrium dataset; Synthetic features of real-time EEG signals are obtained based on zero-point balance parameters; Evaluation index for obtaining synthetic features of real-time EEG signals by combining scaling factor and scaling factor; Building a balanced dataset includes: Data slices were made from the tested electroencephalogram (EEG) signals; Mark the dynamic and resting data in the data slice; Balancing processing ensures that the number of slices for dynamic data and resting data is the same; Using dynamic and resting state data as samples for the equilibrium dataset; The zero-point equilibrium parameters obtained from the equilibrium dataset include: Obtain low-frequency and high-frequency samples from a balanced dataset; Obtain the mean of the average residuals of all low-frequency samples. ; Obtain the mean of the average residuals of all high-frequency band samples. ; Calculate the synthetic features of a balanced dataset ,Right now ; Obtain the zero-point equilibrium parameter, i.e., the mean. and mean Incorporating synthetic features The formula and let the expected value ,get .

2. The method according to claim 1, characterized in that, Obtaining the scaling factor includes: Obtain motion state data samples from the equilibrium dataset and construct a motion state dataset; Calculate the synthetic features of each sample in the motion dataset. ,Right now ;in, This represents the average residual of each high-frequency band sample in the dynamic dataset. This represents the average residual of each low-frequency sample in the dynamic dataset; Based on each synthetic feature Constructing dynamic synthetic feature sets ; Setting dynamic synthesis feature sets The value at the percentile x is denoted as ; Set the feedback intensity of the evaluation indicators ,and ; Obtaining the scaling factor using the inverse hyperbolic tangent function ,Right now .

3. The method according to claim 2, characterized in that, Synthetic features for acquiring real-time EEG signals include: Constructing a real-time dataset based on real-time EEG signals; Obtain low-frequency and high-frequency samples from the real-time dataset; Obtain the mean of the average residuals of the low-frequency band samples. ; Mean of the average residuals based on high-frequency band samples ; Calculate the synthetic features of real-time EEG signals ,Right now ; The evaluation index is obtained by normalization using the hyperbolic tangent function, i.e. 。 4. The method according to claim 1 or 3, characterized in that, Also includes: Assess the state of motor imagery, i.e. Determine if the evaluation index of the real-time EEG signal is greater than zero; if so, assess the motor imagery as motor. If not, then assess the motion imagery as resting.

5. An implantable rehabilitation training device, characterized in that, include: Electrode sensors, located inside the skull, are used to collect real-time electroencephalogram (EEG) signals and send them to the host computer. The host computer is equipped with a processor and runs the method described in any one of claims 1-4 to obtain an evaluation index of real-time EEG signals to evaluate the state of motor imagery. The training peripheral is located at the user's affected area and executes training tasks according to preset instructions; as well as The processor is used to determine the consistency between the motion visualization state and the training task in order to obtain feedback results.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method as described in any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-4.

Images