A sleep therapy effect evaluation device based on cerebral cortex brain electrical source stimulation parameter modulation

Abstract

This invention discloses a device for evaluating the therapeutic effect of sleep modulation based on cortical brain energy. It utilizes an EEG acquisition system to obtain EEG and electrooculography (EOG) data of sleep following both true and false stimulation. Independent component analysis (ICA) is used to preprocess the acquired data to filter out signal artifacts. Brain energy imaging is used to calculate the distribution of neural activity and extract the time series of sleep-related cortical brain energy. The YASA algorithm is used to segment sleep based on the time series of sleep-related cortical brain energy and the EOG data, and the difference in sleep efficiency after true and false stimulation is calculated. A correlation model between stimulation parameters and the difference in sleep efficiency is established. The device includes a data acquisition module and a processing module. This invention provides a scientific basis for evaluating the therapeutic effect of sleep modulation and is of great significance for the optimal selection of stimulation parameters, which can improve the therapeutic effect of sleep modulation.

Description

Technical Field

[0001] This invention relates to the field of sleep modulation therapy evaluation technology, and in particular to a sleep therapy evaluation device based on stimulation parameters of brain cortex power sources. Background Technology

[0002] Treatment for insomnia typically includes sleep-inducing medications and cognitive behavioral therapy (CBT). Hypnotics for insomnia carry the risk of dependence and potential side effects such as headaches, nausea, and short-term memory loss. Furthermore, CBT is only effective for some patients. Therefore, non-invasive physical modulation is a promising method for modulating sleep and improving sleep efficiency. Among these, transcutaneous electrical stimulation (TES) is worthy of further research and promotion due to its low cost, ease of operation, safety, and effectiveness.

[0003] While some effective stimulation parameters for transcutaneous electrical stimulation (TES) in modulating sleep have been identified, their selection is based on experience. How different stimulation parameters (current intensity, stimulation site, duration, stimulation frequency, etc.) affect sleep modulation remains unclear, and the optimal stimulation parameters cannot yet be determined. Furthermore, the efficacy of TES in regulating sleep is mostly assessed by changes in scores on relevant scales before and after treatment. This subjective assessment method can introduce biases and lead to a disconnect between stimulation parameters and behavioral effects. Therefore, a method for assessing the efficacy of TES in modulating sleep based on cortical brain electrical activity is needed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a device for evaluating the therapeutic effect of sleep modulation by stimulation parameters based on the brain cortex power source, which can be used to evaluate the effect of transcutaneous electrical stimulation on sleep modulation and can provide important guidance for the selection of stimulation parameters for sleep modulation.

[0005] To address the aforementioned technical problems, this invention provides a method for evaluating sleep efficacy based on stimulation parameters modulated by cerebral cortex electrical activity, comprising the following steps:

[0006] Step 1: Use an EEG acquisition system to collect sleep EEG and EOG data for 60 minutes after 30 minutes of real stimulation and sham stimulation;

[0007] Step 2: Preprocess the EEG and EEG data collected in Step 1: downsampling, notch filtering to remove power frequency and harmonic interference, and independent component analysis to remove artifact interference.

[0008] Step 3: Calculate the estimated location, direction, and intensity of neural activity sources in the brain from the EEG data preprocessed in Step 2.

[0009] Step 4: Extract the time series of sleep-related cortical brain power sources from the neural activity sources calculated in Step 3;

[0010] Step 5: Use the YASA algorithm to perform sleep staging on the time series of sleep-related cortical brain energy and electrooculography signals extracted in Step 4, and calculate the difference in sleep efficiency after true stimulation and false stimulation.

[0011] Step 6: Establish a correlation model between the difference in sleep efficiency and stimulation parameters through data fitting;

[0012] Step 7: Evaluate the therapeutic effect of sleep modulation based on the model obtained in Step 6.

[0013] Preferably, in step 1, the EEG acquisition system includes a 256-lead Ag / AgCl electrode cap, four integrated bipolar leads, a 256-lead amplifier, and a data cable. The electrode distribution conforms to the international 10-20 system. The bipolar leads are used to measure vertical and horizontal electrooculography, and the data cable connects the amplifier and the recording computer.

[0014] Preferably, step 1, collecting sleep EEG and EEG data for 60 minutes after 30 minutes of real and sham stimulation, specifically includes the following steps:

[0015] Step 11: Select an EEG cap that fits your head circumference appropriately;

[0016] Step 12: Treat the skin under the bilateral mastoid electrodes with a scrub;

[0017] Step 13: Correctly wear the electrode cap and attach the electrooculography electrodes;

[0018] Step 14: Inject conductive paste. First, inject the grounding electrode and the reference electrode, and then inject the other electrodes.

[0019] Step 15: After injecting all channels, wait for the conductive paste to settle for a period of time and the impedance to drop to 5kΩ, then simultaneously collect EEG and EEG data.

[0020] Preferably, step 2, the preprocessing of the EEG and EEG data collected in step 1, specifically includes the following steps:

[0021] Step 21: Downsample the signal to 250Hz;

[0022] Step 22: Use a Butterworth filter to filter out 50Hz power frequency and harmonic interference;

[0023] Step 23: Use independent component analysis to remove artifacts from electrooculography (EOG) signals, electrocardiogram (ECG) signals, and blinking artifacts from EOG signals.

[0024] Preferably, step 3, which involves reverse-engineering the estimated location, direction, and intensity information of the brain's neural activity sources from the preprocessed EEG data of step 2, specifically includes the following steps:

[0025] Step 31: Use the boundary element method to solve the positive EEG problem;

[0026] Step 32: The minimum norm imaging method is used to solve the EEG inverse problem. The high underdeterminism of the inverse problem is solved by introducing a regularization term in the form of source covariance.

[0027] Preferably, in step 4, the sleep-related brain regions are the frontal lobe, occipital lobe, parietal lobe, and central region, and the brain atlas used is Desikan Atlas.

[0028] Preferably, in step 5, the sleep-related cortical brain energy time series and electrooculogram signals extracted in step 4 are used to perform sleep staging using the YASA algorithm, and the difference in sleep efficiency after true stimulation and false stimulation is calculated. This specifically includes the following steps:

[0029] Step 51: Read the time series of brain electrical activity in the sleep-related cortex (frontal lobe, occipital lobe, parietal lobe, and central region) after tracing the source;

[0030] Step 52: Select the sleep brain region and electrooculography channel, extract feature indicators for analysis, and obtain the sleep transition probability matrix and sleep stage prediction results; the feature indicators are ordered by importance as follows: EOG absolute power, EEG Petrosian fractal dimension, EEG absolute power, EEG Beta power, EEG fast Delta power, EEG Delta / Beta power ratio, EEG permutation entropy, and EOG Petrosian fractal dimension.

[0031] Step 53: Using the sleep transition probability matrix calculated in Step 52 and the sleep stage prediction results, calculate the percentage of sleep time:

[0032]

[0033] Preferably, in step 6, establishing a correlation model between the difference in sleep efficiency and stimulation parameters through data fitting specifically includes the following steps:

[0034] Step 61: Collect data, including the difference in sleep efficiency between real and fake stimuli (y) and stimulation parameters (stimulation intensity: x1, stimulation time: x2, stimulation frequency: x3).

[0035] Step 62: Plot the data using scatter plots. The y-axis represents the difference in sleep efficiency, and the x-axis represents the stimulation intensity, stimulation time, and stimulation frequency, respectively. By plotting the differences in stimulation intensity versus sleep efficiency, stimulation time versus sleep efficiency, and stimulation frequency versus sleep efficiency, the relationship between the difference in sleep efficiency and stimulation intensity, stimulation time, and stimulation frequency is determined to be linear.

[0036] Step 63: Check homoscedasticity by plotting the residuals using the fitted values. If the variance of the residuals remains constant within the range of the fitted values, homoscedasticity is satisfied.

[0037] Step 64: Check normality by plotting the residuals using a normal probability plot. The residuals are normally distributed, satisfying the normality assumption.

[0038] Step 65: Check the independence of the residuals by examining the residual plot according to the data order. There are no discernible patterns in the residuals, satisfying the independence assumption.

[0039] Step 66: Construct the model: y = 0.56 – 0.43x1 + 1.23x2 - 1.25x3 - 0.15x1 2 +1.03x2 2 +0.93x3 2 +0.91x1x2–0.36x1x3+0.12x2x3.

[0040] Preferably, in step 7, different stimulation parameters are designed according to the needs, and the changes in sleep efficiency under different stimulation parameters are predicted based on the constructed model and the designed stimulation parameters.

[0041] Correspondingly, a device for evaluating the therapeutic effect of sleep by modulating stimulation parameters based on the brain cortex power source includes: a data acquisition module for acquiring 60 minutes of EEG and EEG data after 30 minutes of true and false stimulation via transcutaneous electrical stimulation, and preprocessing the data by downsampling to 250Hz, filtering out power frequency and harmonic interference using a Butterworth notch filter, and removing EEG and ECG artifacts and blink artifacts from the EEG signal using independent component analysis;

[0042] The processing module employs the boundary element method to solve the forward EEG problem, using the Colin 27 brain model. It uses minimum norm imaging to solve the inverse EEG problem, where the regularization term is calculated from the source covariance of resting EEG data. Time series of brain energy sources in sleep-related brain regions (frontal, occipital, parietal, and central regions) are extracted using the Desikan Atlas brain atlas. The YASA algorithm is used to stage sleep using the extracted time series of sleep-related cortical brain energy sources and electrooculogram (EOG) signals, and the difference in sleep efficiency after true and false stimulation is calculated. The sleep modulation efficacy of the stimulation parameters is then evaluated based on the sleep efficiency.

[0043] The beneficial effects of this invention are as follows: This invention helps to evaluate the impact of stimulation parameters on sleep modulation, provides a theoretical basis and calculation model for the selection of stimulation parameters, and the selected stimulation parameters more effectively improve sleep efficiency. Attached Figure Description

[0044] Figure 1This is a schematic diagram of the experimental paradigm for sleep modulation of the present invention.

[0045] Figure 2 This is a schematic diagram of the method flow of the present invention.

[0046] Figure 3 This is a schematic diagram of the data processing flow of the present invention.

[0047] Figure 4 This is a schematic diagram illustrating the brain power imaging and sleep-related brain region selection process of the present invention.

[0048] Figure 5 This is a schematic diagram of the electronic device of the present invention.

[0049] Figure 6 This diagram illustrates the data acquisition system and real-time sleep efficacy evaluation system built for this invention. Detailed Implementation

[0050] A method for assessing the therapeutic effect of sleep by modulating stimulation parameters based on cortical brain electrical activity includes the following steps:

[0051] Step 1: Use an EEG acquisition system to collect sleep EEG and EOG data for 60 minutes after 30 minutes of real stimulation and sham stimulation;

[0052] Step 2: Preprocess the EEG and EEG data collected in Step 1: downsampling, notch filtering to remove power frequency and harmonic interference, and independent component analysis to remove artifact interference.

[0053] Step 3: Calculate the estimated location, direction, and intensity of neural activity sources in the brain from the EEG data preprocessed in Step 2.

[0054] Step 4: Extract the time series of sleep-related cortex from the neural activity sources calculated in Step 3;

[0055] Step 5: Use the YASA algorithm to perform sleep staging on the time series of sleep-related cortical brain energy and electrooculography signals extracted in Step 4, and calculate the difference in sleep efficiency after true stimulation and false stimulation.

[0056] Step 6: Establish a correlation model between the difference in sleep efficiency and stimulation parameters through data fitting;

[0057] Step 7: Evaluate the therapeutic effect of sleep modulation based on the model obtained in Step 6.

[0058] As Figure 1The diagram illustrates the experimental paradigm for transcutaneous electrical stimulation (TES) sleep modulation. Subjects received 30 minutes each of sham and real TES stimulation, followed by relaxation and natural sleep, while EEG signals were recorded for 60 minutes. Real stimulation consisted of 60 repetitions of the following phase: one 20-second stimulation phase and one 10-second stimulation-off phase. Sham stimulation lasted for one minute, with no electrical stimulation for the remaining 29 minutes. The stimulation current intensity was below the perceived threshold.

[0059] like Figure 2 The diagram shown is a schematic of the overall algorithm flow of this invention, which mainly includes several steps: acquisition of EEG and EEG data, preprocessing, brain power imaging, extraction of time series of sleep-related cortical neurons and EEG signals, sleep staging, calculation of sleep efficiency, and establishment of a correlation model between stimulation parameters and sleep efficiency.

[0060] The main steps involved in EEG data acquisition are as follows:

[0061] (1) Before collecting EEG data, first select an EEG cap with a head circumference that is appropriate;

[0062] (2) Use a scrub to treat the skin under the bilateral mastoid electrodes to reduce impedance and enable the electrodes to capture brain signals more clearly.

[0063] (3) Wear the electrode cap correctly, with the Cz electrode in the center of the head, and the tip of the nose and the center of the eyebrows on the central line;

[0064] (4) Use medical tape to attach the electrooculography electrodes. The horizontal electrooculography electrode is placed two fingers below the eyebrow, about 1 cm away from the outer corner of the eye, and the vertical electrooculography electrode is placed 2 cm away from the lower eyelid.

[0065] (5) Inject conductive paste. Slowly insert the syringe needle into the electrode at a 45-degree angle and inject about 5 ml of conductive paste. Rotate the needle around the electrode once to distribute the conductive paste evenly. Inject the ground electrode and reference electrode first, then inject the other electrodes;

[0066] (6) After all electrode channels are injected, wait for the conductive paste to settle for a period of time. Once the impedance drops to 5kΩ, collect EEG data at a frequency of 1000Hz.

[0067] Preprocessing methods for EEG and EEG signals include: downsampling to 250Hz; using a Butterworth filter to remove 50Hz power frequency and harmonic interference; and using independent component analysis to remove EEG and ECG artifacts, as well as blink artifacts in EEG signals.

[0068] Brain power imaging:

[0069] Brain electrical activity imaging projects the potentials acquired by electrodes onto the cerebral cortex. The calculation formula is as follows:

[0070] L = TH + N

[0071] Where T is the pilot field matrix, N is the noise matrix, L is the scalp potential recorded by the electrodes, and H is the density of cortical brain power.

[0072] First, the boundary element method was used to solve the positive EEG problem, and the brain model constructed was Colin 27.

[0073] The covariance matrix is ​​estimated from resting EEG data. The inverse EEG problem is solved by employing minimum paradigm imaging and introducing the covariance matrix. The formulas include:

[0074]

[0075] Where X is the covariance matrix and Z is the weight matrix.

[0076] After brain electrical activity imaging, time-series and electrooculography data of several sleep-related cortical neurons, including the frontal lobe, occipital lobe, parietal lobe, and central region, were extracted. The YASA algorithm was used to stage sleep, calculate sleep efficiency, and construct a correlation model between stimulation parameters and sleep efficiency.

[0077] Figure 3 The flowchart illustrates the data processing steps of this invention. Raw EEG and EEG data from sleep following both true and sham transcutaneous electrical stimulation (TES) are downsampled, with one sample taken every four points. An 8th-order Butterworth infinite impulse response filter with a center frequency of 50Hz is used to remove power frequency interference.

[0078] Figure 4 The flowchart for the brain power imaging and sleep-related cortical selection of this invention is as follows: 3D coordinates of the acquisition electrode locations are created and co-registered with the Colin 27 head volume. The boundary element method is used to segment the Colin 27 head volume. The pilot field matrix is ​​calculated when the source space is confined to gray matter. The minimum norm imaging method is applied to solve the EEG inverse problem, and the high underdeterminism of the inverse problem is addressed by introducing a regularization term in the form of source covariance. A DesikanAtlas is constructed, and sleep-related cortical areas—frontal lobe, occipital lobe, parietal lobe, and central region—are selected using region-of-interest filters.

[0079] Figure 5 The diagram below shows the modules of the electronic device of the present invention. The device 500 includes: a data acquisition module 501, used to acquire EEG signals after real and false stimuli; and a processing module 502, used to preprocess the acquired EEG data, perform EEG source imaging, extract the time series of sleep-related cortical neurons, perform sleep staging on the extracted time series of sleep-related cortical neurons and electrooculogram signals, and calculate the difference in sleep efficiency after real and false stimuli.

[0080] Figure 6 The diagram shows the data acquisition system and real-time sleep efficacy evaluation system built by this developer. Specifically, an EEG-based data acquisition system is built, which includes EEG sensors, EOG sensors, interface and control box modules, and software control interface. While acquiring data, sleep efficiency and brain region activation levels are visualized in real time, enabling real-time evaluation of sleep efficacy.

Claims

1. A device for evaluating sleep efficacy based on stimulation parameters modulated by brain cortex electrical activity, characterized in that, Includes: a data acquisition module that collects sleep EEG and EEG data for 60 minutes after 30 minutes of real and fake stimulation through an EEG acquisition system, and preprocesses the data by downsampling, notch filtering to remove power frequency and harmonic interference, and using independent component analysis to remove artifact interference; The processing module reverse-engineers the location, direction, and intensity of estimated intracranial neural activity sources from the preprocessed EEG data. It extracts the time series of sleep-related cortical brain activity sources from the calculated neural activity sources. Using the YASA algorithm, it performs sleep staging on the extracted time series of sleep-related cortical brain activity and EEG data, and calculates the difference in sleep efficiency after true stimulation and false stimulation. Through data fitting, it establishes a correlation model between the difference in sleep efficiency and stimulation parameters. Based on the obtained model, it evaluates the therapeutic effect of sleep modulation on stimulation parameters.

2. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The EEG acquisition system includes a 256-lead Ag / AgCl electrode cap, four integrated bipolar leads, a 256-lead amplifier, and a data cable. The electrode distribution conforms to the international 10-20 system. The bipolar leads are used to measure vertical and horizontal electrooculography, and the data cable connects the amplifier and the recording computer.

3. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The process of collecting sleep EEG and EEG data for 60 minutes after 30 minutes of real and sham stimulation includes the following steps: Step 11: Select an EEG cap that fits your head circumference appropriately; Step 12: Treat the skin under the bilateral mastoid electrodes with a scrub; Step 13: Correctly wear the electrode cap and attach the electrooculography electrodes; Step 14: Inject conductive paste. First, inject the grounding electrode and the reference electrode, and then inject the other electrodes. Step 15: After injecting all channels, wait for the conductive paste to settle for a period of time and the impedance to drop to 5kΩ, then simultaneously collect EEG and EEG data.

4. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The preprocessing of the collected EEG and EEG data includes the following steps: Step 21: Downsample the signal to 250Hz; Step 22: Use a Butterworth filter to filter out 50Hz power frequency and harmonic interference; Step 23: Use independent component analysis to remove artifacts from electrooculography (EOG) signals, electrocardiogram (ECG) signals, and blinking artifacts from EOG signals.

5. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The process of reverse-engineering the location, direction, and intensity of neural activity sources in the brain from preprocessed EEG data includes the following steps: Step 31: Use the boundary element method to solve the positive EEG problem; Step 32: The minimum norm imaging method is used to solve the EEG inverse problem. The high underdeterminism of the inverse problem is solved by introducing a regularization term in the form of source covariance.

6. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The brain regions associated with sleep are the frontal lobe, occipital lobe, parietal lobe, and central region, and the brain atlas used is the Desikan Atlas.

7. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, The YASA algorithm was used to segment sleep based on the extracted time-series and electrooculography data of sleep-related cortical brain energy, and the difference in sleep efficiency after true stimulation and sham stimulation was calculated. The specific steps included: Step 51: Read the time series of sleep-related cortical brain energy after tracing the source; Step 52: Select the sleep brain region and electrooculography channel, extract feature indicators for analysis, and obtain the sleep transition probability matrix and sleep stage prediction results; the feature indicators are ordered by importance as follows: EOG absolute power, EEG Petrosian fractal dimension, EEG absolute power, EEG Beta power, EEG fast Delta power, EEG Delta / Beta power ratio, EEG permutation entropy, and EOG Petrosian fractal dimension. Step 53: Using the sleep transition probability matrix calculated in Step 52 and the sleep stage prediction results, calculate the percentage of sleep time:

8. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, Establishing a correlation model between sleep efficiency differences and stimulation parameters through data fitting includes the following steps: Step 61: Collect data. Collect the difference in sleep efficiency between real and fake stimuli (y) and the stimulation parameters. The stimulation parameters are: stimulation intensity (x1), stimulation time (x2), and stimulation frequency (x3). Step 62: Plot the data. Use a scatter plot to plot the data. The y-axis represents the difference in sleep efficiency, and the x-axis represents the stimulation intensity, stimulation time, and stimulation frequency, respectively. By plotting the differences in stimulation intensity and sleep efficiency, stimulation time and sleep efficiency, and stimulation frequency and sleep efficiency, we can determine that the relationship between the difference in sleep efficiency and stimulation intensity, stimulation time, and stimulation frequency is linear. Step 63: Check homoscedasticity by plotting the residuals using the fitted values; the variance of the residuals remains constant within the range of the fitted values, thus satisfying homoscedasticity. Step 64: Check normality by plotting the residuals using a normal probability plot; the residuals are normally distributed, satisfying the normality assumption. Step 65: Check the independence of the residuals by examining the residual plot according to the data order; there are no discernible patterns in the residuals, satisfying the independence assumption; Step 66: Construct the model: y = 0.56 – 0.43x1 + 1.23x2 - 1.25x3 - 0.15x1 2 +1.03x2 2 +0.93x3 2 +0.91x1x2–0.36x1x3+0.12x2x3.

9. The sleep efficacy assessment device based on stimulation parameter modulation of cerebral cortex power sources as described in claim 1, characterized in that, Different stimulation parameters are designed according to needs. Based on the established model and the designed stimulation parameters, the changes in sleep efficiency under different stimulation parameters are predicted.

Images