An epidural electroencephalogram electrode interface stability evaluation method and device

Abstract

The application provides an epidural electroencephalogram electrode interface stability evaluation method and device, and relates to the technical field of brain-computer interfaces. The method comprises the following steps: performing short-term stability evaluation on the electrode interface stability of the epidural electroencephalogram; performing power-law index long-term stability evaluation on the electrode interface stability of the epidural electroencephalogram; and performing sub-band energy long-term stability evaluation on the electrode interface stability of the epidural electroencephalogram. The device executes the above method. The epidural electroencephalogram electrode interface stability evaluation method provided in the embodiment of the application can quantitatively and accurately evaluate the dynamic change of the brain-computer interface electrode interface.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface technology, specifically to a method and apparatus for evaluating the stability of the electrode interface in epidural electroencephalography (EEG). Background Technology

[0002] Implantable brain-computer interfaces (BCIs) can decode motor intentions in real time. However, most existing implantable BCIs are prone to signal attenuation or even electrode failure after long-term implantation due to changes in physical contact (such as displacement or deformation) and biochemical state (such as glial cells encapsulating the electrodes). This also leads to frequent system calibration and high maintenance costs, which is the core challenge for implantable BCIs to move from the laboratory to the home environment.

[0003] Epidural electroencephalography (EEG), as a novel recording method, places electrodes outside the dura mater of the brain, maintaining high spatial resolution and wide bandwidth signal acquisition while avoiding damage to the brain's internal environment, thus enabling long-term stable recording. However, current technologies lack a systematic method for quantitatively assessing the dynamic changes at the brain-computer interface electrode interface. Summary of the Invention

[0004] To address the problems in the prior art, embodiments of the present invention provide a method and apparatus for evaluating the stability of the electrode interface in epidural electroencephalography (EEG), which can at least partially solve the problems existing in the prior art.

[0005] On one hand, this invention proposes a method for assessing the stability of the electrode interface in epidural electroencephalography (EEG), comprising:

[0006] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0007] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0008] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0009] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0010] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0011] The step of obtaining the power-law exponent based on the first power spectral density and each frequency point in the first high-frequency band includes:

[0012] Based on the logarithmic values ​​of each frequency point and the corresponding logarithmic values ​​of the power spectral density at each frequency point, a linear fit is performed on the frequency points and the power spectral density to obtain the optimal fitted line.

[0013] The slope of the fitted line is determined based on the optimal fitted line, and the absolute value of the slope of the fitted line is taken to obtain the power law exponent.

[0014] The step of obtaining the power spectral subband energy based on the second power spectral density and the upper and lower bounds of the target spectral subband includes:

[0015] The power spectrum subband energy is calculated using the following formula:

[0016]

[0017] in, f1 is the energy of the power spectral sub-band, f2 is the upper bound of the target spectral sub-band, f1 is the lower bound of the target spectral sub-band, and dfP2(f) represents the integration over the second power spectral density.

[0018] The method for assessing the stability of the electrode interface in epidural electroencephalography further includes:

[0019] Based on the power spectrum subband energy of each channel in each experiment, calculate the relative energy percentage corresponding to each channel in each experiment;

[0020] Based on the relative energy percentage vector of each channel, which is composed of the relative energy percentage of each channel and the relative energy percentage vector of each experiment, calculate the similarity between any two channel relative energy percentage vectors.

[0021] The spatial stability of different sub-band signals is determined by comparing the similarity calculation results with the preset similarity threshold.

[0022] The method for assessing the stability of the electrode interface in epidural electroencephalography further includes:

[0023] Acquire EEG data under limb movement task state, and determine the task power spectrum sub-band energy of the signal characteristic frequency band during the task prompt period and the baseline power spectrum sub-band energy of the same frequency band before the task prompt based on the time when the task prompt appears in the EEG data;

[0024] The signal-to-noise ratio of the EEG data is calculated based on the task power spectrum energy and the baseline power spectrum energy.

[0025] The third long-term rate of change of the signal-to-noise ratio is obtained within the target time period. Based on the third long-term comparison result between the third long-term rate of change and the third preset long-term threshold, the long-term stability of the signal-to-noise ratio of the characteristic signal quality of epidural EEG is evaluated.

[0026] The third preset long-term threshold is determined based on the signal-to-noise ratio.

[0027] The method for assessing the stability of the electrode interface in epidural electroencephalography further includes:

[0028] The event-related spectral perturbation of the EEG data is calculated based on the task power spectral band energy and the baseline power spectral band energy.

[0029] The spectral disturbance index is calculated based on the time-frequency characteristics of the event-related spectral disturbance.

[0030] The fourth long-term rate of change of the spectral perturbation index is obtained within the target time period. Based on the fourth long-term comparison result between the fourth long-term rate of change and the fourth preset long-term threshold, the long-term stability of the characteristic signal quality of epidural EEG is evaluated.

[0031] The fourth preset long-term threshold is determined based on spectral perturbations.

[0032] On one hand, this invention proposes a device for assessing the stability of the electrode interface in epidural electroencephalography (EEG), comprising:

[0033] The first acquisition unit is used to acquire epidural resting-state EEG data, and acquire a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the epidural resting-state EEG data.

[0034] The second acquisition unit is used to acquire the power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and to acquire the power spectral subband energy based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0035] The first evaluation unit is used to evaluate the short-term stability of the electrode interface of epidural EEG based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0036] The second evaluation unit is used to acquire the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy within the target time period, and to evaluate the long-term stability of the electrode interface of the epidural EEG based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold; and to evaluate the long-term stability of the electrode interface of the epidural EEG based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold.

[0037] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0038] In another aspect, embodiments of the present invention provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the following method:

[0039] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0040] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0041] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0042] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0043] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0044] This invention provides a computer-readable storage medium, comprising:

[0045] The computer-readable storage medium stores a computer program that, when executed by a processor, implements the following method:

[0046] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0047] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0048] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0049] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0050] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0051] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the following method:

[0052] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0053] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0054] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0055] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0056] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0057] The present invention provides a method and apparatus for evaluating the stability of the electrode interface in epidural electroencephalography (EEG). The method acquires resting-state EEG data from the epidural brain; obtains a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the resting-state EEG data; obtains a power law exponent based on the first power spectral density and each frequency point in the first high-frequency band; obtains the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; evaluates the short-term stability of the electrode interface in the epidural EEG based on a short-term comparison of the power law exponent with a preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode; and obtains a first long-term rate of change of the power law exponent within a target time period. The second long-term rate of change of the power spectrum subband energy, and the power law exponential long-term stability assessment of the electrode interface stability of epidural EEG based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold; the subband energy long-term stability assessment of the electrode interface stability of epidural EEG based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold; wherein, the target time period is the total number of implantation days, with days as the basic unit, after a preset number of days after implantation of the brain-computer interface, the first preset long-term threshold is determined according to the power law exponent, and the second long-term preset threshold is determined according to the target spectrum subband, the second high-frequency band, and the power spectrum subband energy; the lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band, which can quantitatively and accurately assess the dynamic changes of the brain-computer interface electrode interface. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0059] Figure 1 This is a flowchart illustrating a method for evaluating the stability of the electrode interface in epidural electroencephalography (EEG) according to an embodiment of the present invention.

[0060] Figure 2 This is a schematic diagram of a method for calculating two indicators, the power law exponent and the power spectral band energy, from the power spectral density of epidural resting-state EEG data, provided in an embodiment of the present invention.

[0061] Figure 3 This is a schematic diagram of the process for evaluating the stability of the epidural electrode interface using the power law exponent of epidural resting-state EEG data, provided in an embodiment of the present invention.

[0062] Figure 4 This is a schematic diagram of the process for evaluating the stability of the epidural electrode interface using the power spectrum band energy of resting-state epidural EEG data, provided in an embodiment of the present invention.

[0063] Figure 5 This is the time-frequency diagram of the average of multiple trials of hand imaginary grasping provided in the embodiments of the present invention.

[0064] Figure 6 This is a schematic diagram of the process for evaluating the long-term trend of signal quality by using the signal-to-noise ratio of epidural EEG data when the subject imagines grasping with their hand, as provided in an embodiment of the present invention.

[0065] Figure 7 This is a schematic diagram of the process for assessing the long-term trend of signal quality by using the spectral perturbation index of epidural EEG data when the subject imagines grasping with their hand, as provided in an embodiment of the present invention.

[0066] Figure 8 This is a schematic diagram of the structure of an epidural electroencephalogram electrode interface stability assessment device provided in an embodiment of the present invention.

[0067] Figure 9 This is a schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments and descriptions of the present invention are used to explain the present invention, but are not intended to limit the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other.

[0069] To facilitate understanding of the technical solution provided in this application, the relevant content of the technical solution will be explained below. Epidural brain-computer interfaces (BCIs) have a lower likelihood of infection and electrode failure due to the preservation of the intact dura mater, possessing the potential for long-term stability of the electrode interface and decoding algorithm. However, currently, there is a lack of quantitative evaluation methods for the stability and characteristic signals of epidural EEG electrode interfaces. Therefore, this application uses epidural resting-state EEG data to extract indicators such as power law exponent and power spectral band energy to evaluate the long-term stability of the electrode interface; and uses epidural task-state EEG data to extract indicators such as signal-to-noise ratio and frequency band perturbation to evaluate long-term changes in signal quality and measure the decodeability of the BCI signal.

[0070] Figure 1 This is a flowchart illustrating a method for assessing the stability of the electrode interface in epidural electroencephalography (EEG) according to an embodiment of the present invention. Figure 1As shown, the method for evaluating the stability of the electrode interface in epidural electroencephalography provided in this embodiment of the invention includes:

[0071] Step S1: Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, acquire the first power spectral density corresponding to the first high-frequency band and the second power spectral density corresponding to the target spectral sub-band.

[0072] Step S2: Obtain the power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and obtain the power spectral subband energy based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0073] Step S3: Based on the short-term comparison results between the power law exponent and the preset short-term threshold, the short-term stability of the electrode interface of the epidural EEG is evaluated. The upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of the scalp EEG and the power law exponent of the subdural electrode.

[0074] Step S4: Obtain the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy within the target time period. Based on the first long-term comparison result between the first long-term rate of change and the first preset long-term threshold, perform a power law exponent long-term stability assessment on the electrode interface stability of the epidural EEG. Based on the second long-term comparison result between the second long-term rate of change and the second preset long-term threshold, perform a subband energy long-term stability assessment on the electrode interface stability of the epidural EEG.

[0075] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0076] In step S1 above, the device acquires epidural resting-state EEG data, and based on the epidural resting-state EEG data, acquires a first power spectral density corresponding to a first high-frequency band, and a second power spectral density corresponding to a target spectral sub-band. The device can be a computer device, such as a server, that executes this method. The acquisition, storage, use, and processing of data in this application's technical solution all comply with relevant regulations. Figure 2 As shown, the power-law exponent calculates the absolute value of the slope of the power spectral density curve, while the power spectral subband energy calculates the normalized energy within a specific frequency band. The following sections describe the specific procedures for evaluating the long-term stability of the electrode interface based on the power-law exponent and power spectral subband energy indices.

[0077] Based on epidural resting-state EEG data, preprocessing such as averaging rereference, high-pass filtering, and power frequency notch filtering can be performed first.

[0078] It is worth noting that resting-state data can be collected under conditions intended for relaxation, such as... Figure 3 As shown, the data truncation length can be selected to be 30 seconds or more to obtain reliable power law exponent and power spectrum subband energy results.

[0079] Further, resting-state data segments of 100 seconds or longer can be collected to improve the reliability of the analysis results.

[0080] Based on the preprocessed EEG data, the frequency f and the corresponding first power spectral density P1(f) are obtained by spectral estimation methods including but not limited to the Welch method.

[0081] In this embodiment of the invention, the spectrum sub-bands are divided as follows: alpha (8-12Hz), beta (15-30Hz), low-gamma (35-50Hz), and high-gamma (55-200Hz) bands. The spectrum division is not limited to the above method; the upper and lower boundaries of the sub-bands may vary slightly depending on individual differences among test subjects. Taking alpha as an example, the target spectrum sub-band is 8-12Hz, and the second power spectral density is denoted as P2(f).

[0082] In step S2 above, the device obtains the power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and obtains the power spectral subband energy based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0083] By calculating the frequency points and the corresponding first power spectral density values, the power law exponent of the first high-frequency band signal is calculated. Since the power spectral density P1(f) of the non-periodic component of the EEG signal satisfies the power law distribution shown in the following equation within a specific frequency range:

[0084] P1(f)~Af -χ

[0085] Where A represents the overall intensity coefficient of the non-periodic component, and χ is the power law exponent, used to describe the decay rate of the first power spectral density in the frequency domain: the power law exponent of scalp EEG is about 1, the power law exponent of subdural electrodes is about 4, and the power law exponent of epidural electrodes is between scalp EEG and subdural EEG, and is closer to subdural EEG, with a normal range between 2.5 and 4.

[0086] Furthermore, the power law exponent reflects the statistical regularity of asynchronous activity in a neuronal population. To avoid interference from the obvious periodic oscillation components in the low-frequency band on the accuracy of the power law exponent calculation, a high-frequency signal is selected to calculate the power law exponent of the non-periodic oscillation components. The lower limit of the preferred frequency band for calculating the power law exponent is 50Hz; further, the frequency range of 70-200Hz can be selected.

[0087] The step of obtaining the power-law exponent based on the first power spectral density and each frequency point in the first high-frequency band includes:

[0088] Based on the logarithmic values ​​of each frequency point and the corresponding logarithmic values ​​of the power spectral density at each frequency point, a linear fit is performed on the frequency points and the power spectral density to obtain the optimal fitted line.

[0089] The slope of the fitted line is determined based on the optimal fitted line, and the absolute value of the slope is taken to obtain the power law exponent. The explanation is as follows:

[0090] In a double logarithmic coordinate system, within a specific frequency band, the optimal fitting line is obtained by linearly fitting the frequency points and the power spectral density P1(f) corresponding to each frequency point using the logarithmic value of each frequency point and the logarithmic value of each frequency point using the least squares method.

[0091] The slope of the fitted line is calculated by extracting the optimal fitted line, and the absolute value of the slope of the fitted line is taken to obtain the power law exponent.

[0092] The step of obtaining the power spectral subband energy based on the second power spectral density and the upper and lower bounds of the target spectral subband includes:

[0093] The power spectrum subband energy is calculated using the following formula:

[0094]

[0095] in, Let f2 be the upper bound of the target spectral sub-band, f1 be the lower bound of the target spectral sub-band, and dfP2(f) represent the integration over the second power spectral density. Based on the second power spectral density P2(f), the average energy within the same sub-band is calculated and converted to dB units. The power spectral sub-band energy reflects the energy intensity of the EEG data power spectral density within a specific frequency range, reflecting the intensity of synchronous activity of a neuronal population within a specific frequency range.

[0096] In step S3 above, the device performs a short-term stability assessment of the electrode interface stability of the epidural EEG based on a short-term comparison of the power-law exponent with a preset short-term threshold. The upper and lower bounds of the preset short-term threshold are determined based on the power-law exponents of the scalp EEG and the subdural electrodes. Referring to the example above, the upper bound of the preset short-term threshold is 4, and the lower bound is 2.5. If the power-law exponent is within the range of the upper and lower bounds, it indicates that the epidural EEG signal is normal and the electrode interface is stable.

[0097] If the power law exponent of an electrode is outside the threshold range in the calculation results, an early warning will be issued. In this case, it may be necessary to check and confirm the electrode displacement through imaging or other means.

[0098] In step S4 above, the device acquires the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy within the target time period. Based on the first long-term comparison result between the first long-term rate of change and the first preset long-term threshold, the device performs a power law exponent long-term stability assessment on the electrode interface stability of the epidural EEG. Based on the second long-term comparison result between the second long-term rate of change and the second preset long-term threshold, the device performs a subband energy long-term stability assessment on the electrode interface stability of the epidural EEG.

[0099] The target time period is the total number of implantation days, measured in days, after a preset number of days following brain-computer interface implantation. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the target spectral sub-band, the second high-frequency band, and the power spectral sub-band energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band. The preset number of days can be set independently based on actual conditions and can be selected as 100 days. Based on EEG data spanning a long period, the power law exponent of each epidural resting-state EEG data collection is calculated with the data implantation days as the independent variable. To exclude the influence of tissue healing during the repair period on signal stability assessment, long-term rate of change analysis should begin at least 100 days after implantation. The total number of implantation days can be the total number of implantation days collected daily or not collected daily. For example, if data is collected on Monday, Wednesday, Friday, and Sunday of a week, the total number of implantation days is not seven days but four days.

[0100] Linear regression was used to fit the trend of the power law exponent with the number of implantation days, and the slope of the long-term trend fitting line was obtained. The fitting formula for the long-term trend linear regression model is as follows:

[0101] y = βx + b + ε

[0102] Where y represents the observed frequency band power index, x represents the number of days after implantation, β represents the long-term power change rate, b is the intercept of the fitted line, and ε is the fitting error.

[0103] Calculate the 95% confidence interval of the slope and multiply the slope value by 100. Define the first long-term rate of change of the power law exponent (unit: / 100 days) as a visual assessment of the stability of the electrode interface.

[0104] To further improve data acquisition accuracy, the power-law exponent of each epidural resting-state EEG data acquisition can be calculated using the following method:

[0105] Based on the power law exponent calculation results of each electrode signal, the average value of the power law exponents of multiple electrode signals in a single acquisition can also be calculated as one of the epidural resting-state EEG data acquisitions.

[0106] The first preset long-term threshold can be selected as 0.5. When the long-term change rate of the power law exponent is close to zero, it indicates that the electrode interface remains stable in the long term. The long-term stability threshold of the power law exponent is ±0.5 / 100 days. That is, when the change of the power law exponent in 100 days exceeds 0.5, it indicates that the long-term stability of the electrode interface is insufficient.

[0107] To analyze the long-range rate of change of power spectrum band energy, specifically, using the number of days after EEG data implantation as the independent variable, the power spectrum band energy of each electrode at each acquisition was calculated to assess the impact of tissue healing on the power spectrum band results during the repair period. Long-range rate of change analysis was initiated at least 100 days after implantation.

[0108] A first-order linear regression model was used to fit the long-term trend of energy variation of each power spectrum subband with the number of days since implantation.

[0109] Based on the fitting results, multiplying them by 100 defines the long-term rate of change of power spectrum subband energy (unit: dB / 100 days); the percentage of long-term change of power spectrum subband energy is compared with a preset threshold.

[0110] The second high-frequency band can be selected as a band with a lower limit above 30Hz. It's worth noting that two different thresholds are set for the high-frequency and low-frequency bands. Because low-frequency signals include alpha and beta components (typically below 30Hz), their amplitudes are inherently larger, and they are affected by the test state in the resting state, resulting in greater long-term fluctuations than the high-frequency band. For the evaluation of low-frequency subband energy, a threshold of ±5dB / 100 days is set. If the long-term rate of change of the low-frequency subband energy is within the threshold range, the low-frequency subband energy is considered to be long-term stable.

[0111] For the high-frequency subband energy, it is relatively stable in the resting state. The threshold is set at ±2dB / 100 days. If the long-term change rate of the high-frequency subband energy is within the threshold range, the high-frequency subband energy is determined to be stable in the long term.

[0112] Furthermore, the long-term stability of the electrode is determined based on the long-term stability of the subband energy.

[0113] The above-mentioned high-frequency and low-frequency band division ranges may vary slightly for different test subjects, with a typical value of around 30Hz, which is the upper limit of the beta band.

[0114] The method for assessing the stability of the electrode interface in epidural EEG also includes:

[0115] Based on the power spectrum subband energy of each channel in each experiment, calculate the relative energy percentage corresponding to each channel in each experiment;

[0116] Based on the relative energy percentage vector of each channel, which is composed of the relative energy percentage of each channel and the relative energy percentage vector of each experiment, calculate the similarity between any two channel relative energy percentage vectors.

[0117] The spatial stability of different sub-band signals is determined by comparing the similarity calculation results with the preset similarity threshold.

[0118] like Figure 4 As shown, the relative energy percentage corresponding to each channel in each experiment can be calculated using the following formula.

[0119]

[0120] in, P represents the relative power percentage of channel i in the j-th experiment. i,j Let the power spectrum band energy of channel i in the j-th experiment be, such as Figure 5 As shown, taking an example with 8 channels, the energy of each electrode channel can be used to measure the proportion of energy in the total energy of all channels, thus measuring the long-term changes in the spatial distribution characteristics of the signal. Referring to the example above, the vector of relative energy proportions of the 8 channels, combined with the relative energy proportions of the channels in the m-th experiment, is denoted as ρ. m The vector of relative energy percentages of the eight channels, along with the relative energy percentages of the channels in the nth experiment, is denoted as ρ. n Then calculate the similarity between the relative energy percentage vectors of each pair of channels. m,n This can be expressed by the following formula:

[0121]

[0122] The similarity between the relative energy percentage vectors of each pair of channels can reflect the spatial distribution of signal energy in different experimental sessions.

[0123] The preset similarity threshold can be set according to the actual situation. It can be selected as 0.9. The specific similarity can be selected as cosine similarity. The cosine similarity average of the energy of the power spectrum of all data across the long term is calculated. When the cosine average value exceeds 0.9, the spatial distribution of the signal is stable in the long term, reflecting the spatial stability of signals in different sub-bands.

[0124] For low-frequency signals, the spatial stability of their power spectral band energy is usually lower than that of high-frequency signals. Typically, the spatial stability of signals below 30 Hz is weaker than that of signals above 30 Hz.

[0125] The method for assessing the stability of the electrode interface in epidural EEG also includes:

[0126] Acquire EEG data during a limb movement task, and determine the task power spectrum sub-band energy of the signal characteristic frequency band during the task prompt period, as well as the baseline power spectrum sub-band energy of the same frequency band before the task prompt, based on the time of the task prompt appearance in the EEG data; for example Figure 5 As shown, the dashed line indicates the time when the task prompt appears, the right side of the dashed line represents the task power spectrum energy of the signal characteristic frequency band during the task prompt period, and the left side of the dashed line represents the baseline power spectrum energy of the same frequency band before the task prompt.

[0127] The signal-to-noise ratio of the EEG data is calculated based on the task power spectrum energy and the baseline power spectrum energy.

[0128] The third long-term rate of change of the signal-to-noise ratio is obtained within the target time period. Based on the third long-term comparison result between the third long-term rate of change and the third preset long-term threshold, the long-term stability of the signal-to-noise ratio of the characteristic signal quality of epidural EEG is evaluated.

[0129] The third preset long-term threshold is determined based on the signal-to-noise ratio.

[0130] Taking EEG data from an epidural hand-image grasping task as an example, such as Figure 5 As shown, when imagining grasping with the hand, the main manifestations are event-related desynchronization (ERD) with beta and low gamma and event-related synchronization (ERS) with high gamma above 50Hz.

[0131] like Figure 6 As shown, based on the EEG signals of the subjects' imagined hand movements, the subjects need to imagine grasping or relaxing their left or right hand in a familiar way, and the number of times can be greater than 5 times, and can be selected as 10-15 times, so as to improve the reliability of the results by averaging the number of trials in the future; the captured task-state EEG data are preprocessed by averaging rereference, high-pass filtering, power frequency notch filtering, etc.

[0132] Based on the preprocessed EEG data, the EEG signal within a few seconds after the task prompt appears is used as the task signal, and the EEG data for a certain period of time before the task prompt appears is used as the resting baseline signal.

[0133] Based on the task signal and resting baseline signal, the signal-to-noise ratio (SNR) of the EEG data in the task is calculated using the following formula:

[0134]

[0135] Among them, P task P represents the power spectral subband energy of the signal's characteristic frequency band during the task prompt period. baseline The signal-to-noise ratio (SNR) indicates the power spectrum subband energy within the same frequency band during the resting baseline period before the prompt. The SNR measures the degree of enhancement (ERS) or suppression (ERD) of the mission signal relative to the resting baseline signal.

[0136] Because the subjects' EEG data show a decrease in low-frequency energy and an increase in high-frequency energy after each task prompt, and the response begins to decay 3 seconds after the task prompt appears, this embodiment uses the EEG data segment from 0 to 2.5 seconds after the task prompt as the task signal.

[0137] To ensure the reliability of the results, the duration of the resting baseline data needs to be similar to that of the task data. In this embodiment, -3 to -0.5 seconds before the task prompt is used as the resting baseline.

[0138] Since signal quality is positively correlated with the degree of ERS and negatively correlated with the degree of low-frequency ERD, this embodiment performs high-frequency energy signal-to-noise ratio calculation. The characteristic frequency band is preferably the high gamma band (55-200Hz), which may vary slightly for different subjects.

[0139] Based on EEG data collected from different dates, the signal-to-noise ratio of high-frequency energy during the subject's hand grasping task can be calculated on different dates.

[0140] Using linear regression, with the number of implantation days as the independent variable and the signal-to-noise ratio of high-frequency signals during the hand-image grasping task performed by the subjects on the corresponding dates as the dependent variable, linear regression fitting was performed.

[0141] The slope of the fitted straight line multiplied by 100 is used as the long-term rate of change of the signal-to-noise ratio (unit: dB / 100 days).

[0142] Comparing the calculated results with the threshold ±1dB / 100 days, when the long-term rate of change of the signal-to-noise ratio exceeds the threshold, it indicates that the long-term change of the signal-to-noise ratio is large, suggesting that the brain-computer interface decoding model may need to be recalibrated.

[0143] Meanwhile, if the long-term rate of change of the signal-to-noise ratio exceeds the threshold, it indicates a significant change in the quality of the characteristic signal. When the long-term rate of change of the signal-to-noise ratio is greater than 1 dB / 100 days, it indicates a significant improvement in the signal quality of the brain-computer interface, which may predict beneficial plasticity changes in the cortex induced by training. However, when the long-term rate of change of the signal-to-noise ratio is less than -1 dB / 100 days, it indicates a significant deterioration in the signal quality of the brain-computer interface, and it is necessary to check factors that may induce signal quality deterioration, such as electrode interface, user status, and system hardware.

[0144] The method for assessing the stability of the electrode interface in epidural EEG also includes:

[0145] The event-related spectral perturbation of the EEG data is calculated based on the task power spectral band energy and the baseline power spectral band energy.

[0146] The spectral disturbance index is calculated based on the time-frequency characteristics of the event-related spectral disturbance.

[0147] The fourth long-term rate of change of the spectral perturbation index is obtained within the target time period. Based on the fourth long-term comparison result between the fourth long-term rate of change and the fourth preset long-term threshold, the long-term stability of the characteristic signal quality of epidural EEG is evaluated.

[0148] The fourth preset long-term threshold is determined based on spectral perturbations.

[0149] like Figure 7 As shown, based on the EEG signals of the subjects' imagined hand movements, the subjects need to imagine grasping or relaxing their left or right hand in a familiar way, and the number of times can be greater than 5 times, and can be selected as 10-15 times, so as to improve the reliability of the results by averaging the number of trials in the future; the captured task-state EEG data are preprocessed by averaging rereference, high-pass filtering, power frequency notch filtering, etc.

[0150] Based on the preprocessed EEG data, the EEG signal within a few seconds after the task prompt appears is used as the task signal, and the EEG data for a certain period of time before the task prompt appears is used as the resting baseline signal.

[0151] The event-related spectral perturbation of EEG data in the task is calculated using the following formula:

[0152]

[0153] Among them, P task P represents the power spectral subband energy of the signal's characteristic frequency band during the task prompt period. baseline This indicates the power spectrum band energy within the same frequency band during the resting baseline period before the prompt.

[0154] The time-frequency characteristics of ERSP can be obtained by means of methods including but not limited to wavelet transform, which reflect the characteristics of data in the time and frequency domains relative to the resting baseline under task conditions, and measure the intensity of ERS and ERD at different times and in different subbands.

[0155] The spectral perturbation index was calculated based on the time-frequency characteristics of ERSP. After each task prompt, the subject's EEG data showed a decrease in low-frequency energy and an increase in high-frequency energy, and the response began to decay 3 seconds after the task prompt appeared.

[0156] The Spectrum Perturbation Index (SPI) is shown in the following formula:

[0157]

[0158] Among them, t max and t min These represent the upper and lower time bounds for selecting the time-frequency features of ERSP, respectively. max and f min These are the upper and lower frequency bounds for selecting the time-frequency characteristics of the ERSP.

[0159] according to Figure 5 In this embodiment, the ERSP time-frequency map of EEG data from 0 to 2.5 seconds after the task prompt is selected for quantitative analysis. In this embodiment, the average intensity of high-frequency ERS is calculated, and the characteristic frequency band is preferably the high gamma band (55-200Hz), which may vary slightly for different subjects. The time period is selected from 0 to 2.5 seconds, and the average is calculated to obtain the SPI index in the subject's word conversation.

[0160] After calculating the spectral perturbation index of all trials in a single experimental session, the average of the trials is performed.

[0161] Based on the changes in spectral perturbation indices in long-term task-oriented EEG data, a first-order linear model was used to fit the spectral perturbation indices, with the number of implantation days as the independent variable and the corresponding spectral perturbation indices as the dependent variable.

[0162] The long-term rate of change ( / 100 days) of the spectral perturbation index SPI is calculated by multiplying the fitting slope by 100, and the quality changes of the characteristic signal are quantitatively evaluated.

[0163] A threshold is determined based on the selected frequency band according to the spectral perturbation index. Taking the high-gamma band signal mentioned above as an example, the threshold is set to ±0.5 / 100 days. When the long-term rate of change of the spectral perturbation index exceeds this threshold range, it indicates that the high-frequency ERS response in the patient's task state has changed significantly, suggesting that the model should be recalibrated in a timely manner to make the model more suitable for the signal characteristics under the patient's task state.

[0164] Similarly, if the spectral perturbation index exceeds the threshold and is higher than 0.5 / 100 days, it indicates a significant improvement in the signal quality of the brain-computer interface; while if the spectral perturbation index exceeds the threshold and is lower than 0.5 / 100 days, it indicates a significant deterioration in the signal quality of the brain-computer interface, and it is necessary to check the factors that may lead to the deterioration of the characteristic signal quality.

[0165] The present invention provides a method for evaluating the stability of the electrode interface in epidural electroencephalography (EEG). This method acquires epidural resting-state EEG data; obtains a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the epidural resting-state EEG data; obtains a power law exponent based on the first power spectral density and each frequency point in the first high-frequency band; obtains the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; performs a short-term stability evaluation of the electrode interface stability of the epidural EEG based on a short-term comparison of the power law exponent with a preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode; and acquires the first long-term rate of change of the power law exponent and the... The second long-term rate of change of the power spectrum subband energy is used to evaluate the long-term stability of the epidural EEG electrode interface based on the first long-term comparison result between the first long-term rate of change and the first preset long-term threshold. The second long-term rate of change is used to evaluate the long-term stability of the subband energy based on the second long-term comparison result between the second long-term rate of change and the second preset long-term threshold. The target time period is the total number of implantation days, measured in days, after a preset number of days following brain-computer interface implantation. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the target spectrum subband, the second high-frequency band, and the power spectrum subband energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band, enabling quantitative and accurate evaluation of the dynamic changes in the brain-computer interface electrode interface.

[0166] Further, obtaining the power-law exponent based on the first power spectral density and each frequency point in the first high-frequency band includes:

[0167] Based on the logarithmic values ​​of each frequency point and the corresponding logarithmic values ​​of the power spectral density, a linear fit is performed on the frequency points and the power spectral density to obtain the optimal fitting line; the above embodiments can be referred to for explanation, and will not be repeated here.

[0168] The slope of the fitted line is determined based on the optimal fitted line, and the absolute value of the slope is taken to obtain the power law exponent. This can be referred to the above embodiment for further explanation, and will not be repeated here.

[0169] Further, obtaining the power spectral subband energy based on the second power spectral density and the upper and lower bounds of the target spectral subband includes:

[0170] The power spectrum subband energy is calculated using the following formula:

[0171]

[0172] in, Let f1 be the energy of the power spectral sub-band, f2 be the upper bound of the target spectral sub-band, f1 be the lower bound of the target spectral sub-band, and dfP2(f) represent the integration over the second power spectral density. Refer to the above embodiments for further details.

[0173] Furthermore, the method for assessing the stability of the epidural EEG electrode interface also includes:

[0174] Based on the power spectrum band energy of each channel in each experiment, calculate the relative energy percentage corresponding to each channel in each experiment; refer to the above embodiments for explanation, and will not be repeated here.

[0175] The similarity between any two channel relative energy percentage vectors is calculated based on the relative energy percentage vectors corresponding to each channel and the relative energy percentage vectors corresponding to each experiment; this can be referred to the above embodiments for explanation, and will not be repeated here.

[0176] The spatial stability of different sub-band signals is determined by comparing the similarity calculation results with the preset similarity threshold. This can be referred to the above embodiments for further explanation, and will not be repeated here.

[0177] Furthermore, the method for assessing the stability of the epidural EEG electrode interface also includes:

[0178] EEG data under limb movement task state is acquired, and the task power spectrum band energy of the signal characteristic frequency band during the task prompt period and the baseline power spectrum band energy of the same frequency band before the task prompt are determined based on the time when the task prompt appears in the EEG data; the above embodiments can be referred to for explanation, and will not be repeated here.

[0179] The signal-to-noise ratio of the EEG data is calculated based on the task power spectrum band energy and the baseline power spectrum band energy; this can be referred to the above embodiments for explanation, and will not be repeated here.

[0180] The third long-term rate of change of the signal-to-noise ratio is obtained within the target time period. Based on the third long-term comparison result between the third long-term rate of change and the third preset long-term threshold, the long-term stability of the signal-to-noise ratio of the characteristic signal quality of epidural EEG is evaluated. This can be referred to the above embodiment for explanation, and will not be repeated here.

[0181] The third preset long-term threshold is determined based on the signal-to-noise ratio. This can be referred to the above embodiments for further explanation and will not be repeated here.

[0182] Furthermore, the method for assessing the stability of the epidural EEG electrode interface also includes:

[0183] The event-related spectral perturbation of the EEG data is calculated based on the task power spectrum band energy and the baseline power spectrum band energy; this can be referred to the above embodiments for explanation, and will not be repeated here.

[0184] The spectral disturbance index is calculated based on the time-frequency characteristics of the event-related spectral disturbance; this can be referred to the above embodiments for explanation, and will not be repeated here.

[0185] The fourth long-term rate of change of the spectral perturbation index is obtained within the target time period. Based on the fourth long-term comparison result between the fourth long-term rate of change and the fourth preset long-term threshold, the long-term stability of the characteristic signal quality of epidural EEG is evaluated. This can be referred to the above embodiment for explanation, and will not be repeated here.

[0186] The fourth preset long-term threshold is determined based on spectral perturbations. This can be referred to the above embodiments for further explanation and will not be repeated here.

[0187] Figure 8 This is a schematic diagram of the structure of an epidural electroencephalogram electrode interface stability assessment device provided in an embodiment of the present invention, as shown below. Figure 8 As shown, the epidural electroencephalogram (EEG) electrode interface stability assessment device provided in this embodiment of the invention includes a first acquisition unit 801, a second acquisition unit 802, a first assessment unit 803, and a second assessment unit 804, wherein:

[0188] The first acquisition unit 801 is used to acquire epidural resting-state EEG data, and acquire a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the epidural resting-state EEG data; the second acquisition unit 802 is used to acquire a power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and acquire the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; the first evaluation unit 803 is used to perform a short-term stability evaluation of the electrode interface stability of the epidural EEG based on the short-term comparison result between the power law exponent and a preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of the scalp EEG and the power law exponent of the subdural electrode; the second evaluation unit 804 is used to evaluate the electrode interface stability of the epidural EEG during the target time period. The first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy. The target time period is the total number of implantation days after a preset number of days after the implantation of the brain-computer interface, with days as the basic unit. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectrum subband, the second high-frequency band, and the power spectrum subband energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0189] Specifically, the first acquisition unit 801 in the device is used to acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, acquire a first power spectral density corresponding to a first high-frequency band, and a second power spectral density corresponding to a target spectral sub-band; the second acquisition unit 802 is used to acquire the power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and acquire the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; the first evaluation unit 803 is used to perform a short-term stability evaluation of the electrode interface stability of the epidural EEG based on the short-term comparison result of the power law exponent and a preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of the scalp EEG and the power law exponent of the subdural electrode; the second evaluation unit 804 is used to... Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy. The target time period is the total number of implantation days after a preset number of days after the implantation of the brain-computer interface, with days as the basic unit. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectrum subband, the second high-frequency band, and the power spectrum subband energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0190] The electrode interface stability assessment device for epidural EEG provided in this embodiment of the invention acquires epidural resting-state EEG data, obtains a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the epidural resting-state EEG data; obtains a power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and obtains the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; performs a short-term stability assessment of the electrode interface stability of the epidural EEG based on the short-term comparison results of the power law exponent and a preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode; and obtains the first long-term rate of change of the power law exponent and the... The second long-term rate of change of the power spectrum subband energy is used to evaluate the long-term stability of the epidural EEG electrode interface based on the first long-term comparison result between the first long-term rate of change and the first preset long-term threshold. The second long-term rate of change is used to evaluate the long-term stability of the subband energy based on the second long-term comparison result between the second long-term rate of change and the second preset long-term threshold. The target time period is the total number of implantation days, measured in days, after a preset number of days following brain-computer interface implantation. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the target spectrum subband, the second high-frequency band, and the power spectrum subband energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band, enabling quantitative and accurate evaluation of the dynamic changes in the brain-computer interface electrode interface.

[0191] The embodiments of the present invention provide an electrode interface stability assessment device for epidural electroencephalography that can be used to execute the processing flow of the above-described method embodiments. Its functions will not be repeated here, but can be referred to the detailed description of the above-described method embodiments.

[0192] Figure 9 This is a schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention, such as... Figure 9 As shown, the computer device includes: a memory 901, a processor 902, and a computer program stored in the memory 901 and executable on the processor 902. When the processor 902 executes the computer program, it implements the following method:

[0193] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0194] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0195] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0196] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0197] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0198] This embodiment discloses a computer program product, which includes a computer program that, when executed by a processor, implements the following method:

[0199] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0200] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0201] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0202] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0203] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0204] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the following method:

[0205] Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band;

[0206] The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband.

[0207] The short-term stability of the electrode interface of epidural EEG is evaluated based on the short-term comparison results between the power law index and the preset short-term threshold; the upper and lower bounds of the preset short-term threshold are determined based on the power law index of scalp EEG and the power law index of subdural electrodes.

[0208] Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy.

[0209] The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second long-term preset threshold is determined based on the target spectral sub-band, the second high-frequency band, and the energy of the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band.

[0210] Compared with existing technologies, the present invention provides a method for evaluating the stability of the electrode interface in epidural EEG. This method involves acquiring resting-state epidural EEG data, obtaining a first power spectral density corresponding to a first high-frequency band, and a second power spectral density corresponding to a target spectral sub-band based on the resting-state EEG data; obtaining a power law exponent based on the first power spectral density and each frequency point in the first high-frequency band; obtaining the power spectral sub-band energy based on the second power spectral density and the upper and lower bounds of the target spectral sub-band; and performing a short-term stability evaluation of the electrode interface stability of the epidural EEG based on a short-term comparison of the power law exponent with a preset short-term threshold. The upper and lower bounds of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode. The method acquires the power law exponent within a target time period. The first long-term rate of change and the second long-term rate of change of the power spectrum sub-band energy are used to evaluate the long-term stability of the electrode interface of the epidural EEG using a power law exponential method, based on the first long-term comparison result between the first long-term rate of change and the first preset long-term threshold. The second long-term rate of change and the second preset long-term threshold are used to evaluate the long-term stability of the electrode interface of the epidural EEG using sub-band energy, based on the second long-term comparison result between the second long-term rate of change and the second preset long-term threshold. The target time period is the total number of implantation days, measured in days, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the target spectrum sub-band, the second high-frequency band, and the power spectrum sub-band energy. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band, enabling quantitative and accurate evaluation of the dynamic changes in the brain-computer interface electrode interface.

[0211] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0212] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0213] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0214] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0215] In the description of this specification, the references to terms such as "an embodiment," "a specific embodiment," "some embodiments," "for example," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0216] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for evaluating the stability of the electrode interface in epidural electroencephalography (EEG), characterized in that, include: Acquire epidural resting-state EEG data, and based on the epidural resting-state EEG data, obtain a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band; The power law exponent is obtained based on the first power spectral density and each frequency point in the first high-frequency band, and the power spectral subband energy is obtained based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband. Based on the short-term comparison results between the power law exponent and the preset short-term threshold, the short-term stability of the electrode interface of epidural EEG is evaluated. Within the target time period, the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy are obtained. Based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the power law exponent. Based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold, the long-term stability of the electrode interface of the epidural EEG is evaluated by the subband energy. The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the energy of the target spectral sub-band, the second high-frequency band, and the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band. The upper and lower limits of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode.

2. The method for evaluating the stability of the electrode interface in epidural electroencephalography according to claim 1, characterized in that, The step of obtaining the power-law exponent based on the first power spectral density and each frequency point in the first high-frequency band includes: Based on the logarithmic values ​​of each frequency point and the corresponding logarithmic values ​​of the power spectral density at each frequency point, a linear fit is performed on the frequency points and the power spectral density to obtain the optimal fitted line. The slope of the fitted line is determined based on the optimal fitted line, and the absolute value of the slope of the fitted line is taken to obtain the power law exponent.

3. The method for evaluating the stability of the electrode interface in epidural electroencephalography according to claim 1, characterized in that, The step of obtaining the power spectral subband energy based on the second power spectral density and the upper and lower bounds of the target spectral subband includes: The power spectrum subband energy is calculated using the following formula: in, The power spectrum carries energy. This is the upper bound value in the target spectral sub-band. This is the lower bound value in the target spectral sub-band. This indicates that the second power spectral density is integrated.

4. The method for evaluating the stability of the electrode interface in epidural electroencephalography according to claim 1, characterized in that, The method for assessing the stability of the electrode interface in epidural EEG also includes: Based on the power spectrum subband energy of each channel in each experiment, calculate the relative energy percentage corresponding to each channel in each experiment; Based on the relative energy percentage vector of each channel, which is composed of the relative energy percentage of each channel and the relative energy percentage vector of each experiment, calculate the similarity between any two channel relative energy percentage vectors. The spatial stability of different sub-band signals is determined by comparing the similarity calculation results with the preset similarity threshold.

5. The method for evaluating the stability of the electrode interface in epidural electroencephalography according to any one of claims 1 to 4, characterized in that, The method for assessing the stability of the electrode interface in epidural EEG also includes: Acquire EEG data under limb movement task state, and determine the task power spectrum sub-band energy of the signal characteristic frequency band during the task prompt period and the baseline power spectrum sub-band energy of the same frequency band before the task prompt based on the time when the task prompt appears in the EEG data; The signal-to-noise ratio of the EEG data is calculated based on the task power spectrum energy and the baseline power spectrum energy. The third long-term rate of change of the signal-to-noise ratio is obtained within the target time period. Based on the third long-term comparison result between the third long-term rate of change and the third preset long-term threshold, the long-term stability of the signal-to-noise ratio of the characteristic signal quality of epidural EEG is evaluated. The third preset long-term threshold is determined based on the signal-to-noise ratio.

6. The method for evaluating the stability of the electrode interface in epidural electroencephalography according to claim 5, characterized in that, The method for assessing the stability of the electrode interface in epidural EEG also includes: The event-related spectral perturbation of the EEG data is calculated based on the task power spectral band energy and the baseline power spectral band energy. The spectral disturbance index is calculated based on the time-frequency characteristics of the event-related spectral disturbance. The fourth long-term rate of change of the spectral perturbation index is obtained within the target time period. Based on the fourth long-term comparison result between the fourth long-term rate of change and the fourth preset long-term threshold, the long-term stability of the characteristic signal quality of epidural EEG is evaluated. The fourth preset long-term threshold is determined based on spectral perturbations.

7. A device for assessing the stability of the electrode interface in epidural electroencephalography (EEG), characterized in that, include: The first acquisition unit is used to acquire epidural resting-state EEG data, and acquire a first power spectral density corresponding to a first high-frequency band and a second power spectral density corresponding to a target spectral sub-band based on the epidural resting-state EEG data. The second acquisition unit is used to acquire the power law exponent based on the first power spectral density and each frequency point in the first high-frequency band, and to acquire the power spectral subband energy based on the second power spectral density and the upper and lower bound values ​​in the target spectral subband. The first evaluation unit is used to evaluate the short-term stability of the electrode interface of epidural EEG based on the short-term comparison results between the power law exponent and the preset short-term threshold. The second evaluation unit is used to acquire the first long-term rate of change of the power law exponent and the second long-term rate of change of the power spectrum subband energy within the target time period, and to evaluate the long-term stability of the electrode interface of the epidural EEG based on the first long-term comparison result of the first long-term rate of change and the first preset long-term threshold; and to evaluate the long-term stability of the electrode interface of the epidural EEG based on the second long-term comparison result of the second long-term rate of change and the second preset long-term threshold. The target time period is the total number of implantation days, with days as the basic unit, after a preset number of days following the implantation of the brain-computer interface. The first preset long-term threshold is determined based on the power law exponent, and the second preset long-term threshold is determined based on the energy of the target spectral sub-band, the second high-frequency band, and the power spectral sub-band. The lower limit of the second high-frequency band is lower than the lower limit of the first high-frequency band. The upper and lower limits of the preset short-term threshold are determined based on the power law exponent of scalp EEG and the power law exponent of the subdural electrode.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 6.

Images