Verification method for effect of sleep music closed-loop intervention on slow wave and memory

Abstract

The application discloses a kind of verification method of the influence of sleep music closed loop intervention on slow wave and memory, comprising the following specific steps: step 1), recruit subjects, let each subject participate in adaptive night experiment, stimulation night experiment and pseudo-stimulation night experiment;Step 2), after completing adaptive night experiment, stimulation night experiment and pseudo-stimulation night experiment are carried out;When pseudo-stimulation night experiment is carried out, no music stimulation is given to the subject during sleep, and when stimulation night experiment is carried out, sleep music closed loop control system is started, and music stimulation is given;Step 3), by trained deep learning model, the sleep staging of the electroencephalogram data of subject is carried out, and sleep index is calculated;Step 4), compare and analyze each sleep index of stimulation night experiment group and pseudo-stimulation night experiment group.The application systematically analyzes the verification method of the influence of music stimulation on slow wave sleep and the relationship between slow wave sleep and memory, to promote the in-depth study of human sleep.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sleep analysis, in particular to a verification method for influence of music closed-loop intervention on slow wave and memory during sleep. BACKGROUND

[0002] In recent years, according to statistics, about one fourth of people in the world have sleep problems, and the proportion of people with sleep disorders in China reaches 38%, and the proportion of sleep disorders among office workers is as high as 65%. However, the importance of night sleep for human beings is self-evident, especially in the slow wave sleep period (SWS). Studies have shown that the length and intensity of the slow wave sleep period will directly affect the physiological activities of the human body, as well as the maintenance of cognitive and memory functions, because the human body will perform self-repair and growth in this stage, and the consolidation of memory of the brain is of great significance. The sleep disorder population has a series of problems such as short deep sleep time and low quality, which will cause great harm to their learning ability, concentration, and emotional control ability.

[0003] At present, there are various deep sleep promoting products on the market, such as intelligent deep sleep bedroom, soft and hard intelligent control partition mattress and other appliances, and some natural supplements such as melatonin, B vitamin tablets, etc., which more or less improve the sleep quality and depth of sleep disorder patients, but the specific influence on the sleep process is lacking. In addition, in the existing literature, slow wave sleep enhancement stimulation is mainly divided into transcranial direct current stimulation (tDCS), transcranial magnetic stimulation (TMS) and sound stimulation (AS), but the former two ways belong to invasive stimulation and are expensive; and sound (music) stimulation also has a certain influence on slow wave sleep, but in the existing technology, the influence of sound (music) stimulation on slow wave is not fully shown, and the correlation between slow wave sleep proportion and memory is not studied.

[0004] Therefore, it is urgent to provide a verification method capable of systematically analyzing the influence of music stimulation on slow wave sleep and the relationship between slow wave sleep and memory, so as to promote in-depth research on human sleep. SUMMARY

[0005] The purpose of the present application is to provide a verification method capable of systematically analyzing the influence of music stimulation on slow wave sleep and the relationship between slow wave sleep and memory, so as to promote in-depth research on human sleep.

[0006] The purpose of the present application is achieved by the following technical solution: a verification method for influence of music closed-loop intervention on slow wave and memory during sleep, comprising the following specific steps:

[0007] Step 1), recruit subjects, let each subject participate in an adaptation night experiment, a stimulation night experiment and a pseudo-stimulation night experiment; all subjects first perform the adaptation night experiment, and during the adaptation night experiment, no music stimulation is given to the subjects during sleep;

[0008] Step 2), after completing the adaptation night experiment, randomly select half of the subjects to first perform the stimulation night experiment and then the pseudo-stimulation night experiment, and the other half of the subjects to first perform the pseudo-stimulation experiment and then the stimulation night experiment;

[0009] During the stimulation night experiment and the pseudo-stimulation night experiment, the subjects are first given a pre-sleep word association test before going to sleep; during the pseudo-stimulation night experiment, no music stimulation is given to the subjects during sleep, and the brain electrical data of the subjects during sleep is collected; during the stimulation night experiment, the brain electrical data of the subjects during sleep is collected, and at the same time, a sleep-in music closed-loop control system is started, which performs real-time sleep staging on the collected brain electrical signals; when it is detected that the subject is in deep sleep, the music playing device is controlled to give the subject music stimulation;

[0010] After the subjects finish sleeping, the subjects are given a post-sleep word association test, and the difference between the post-sleep word association test result and the pre-sleep word association test result is taken as the memory consolidation test result of the subject;

[0011] Step 3), through the trained deep learning model, the brain electrical data of the subjects is subjected to sleep staging, which is divided into four sleep stages: wakefulness, rapid eye movement sleep, light sleep and deep sleep; and the following sleep indicators are calculated: NREM total power, spindle number, slow wave power, slow wave amplitude and slow wave slope; wherein, the slow wave slope is the ratio of the difference between the ordinates (i.e. amplitude) of the wave peak and the wave trough of each slow wave to the difference between the abscissas of the wave peak and the wave trough;

[0012] Step 4), the subjects who perform the stimulation night experiment are classified into a stimulation night experiment group, and the subjects who perform the pseudo-stimulation night experiment are classified into a pseudo-stimulation night experiment group; the sleep indicators of the stimulation night experiment group and the pseudo-stimulation night experiment group are compared, and a significance test is performed; and the effect of music stimulation on sleep memory and the correlation between sleep memory and the proportion of deep sleep are analyzed.

[0013] As a preferred, in step 1), the experimental duration of each adaptation night experiment, stimulation night experiment and pseudo-stimulation night experiment is 7 days.

[0014] As a preferred, in step 2), the music playing device is a music pillow.

[0015] As a preferred, in step 2), the sleep-in music closed-loop control system includes a data quality detection, a real-time sleep staging module, a sleep depth calculation module and a real-time slow wave detection module.

[0016] The control method for the sleep music closed-loop control system is as follows:

[0017] S1: The real-time sleep staging module performs real-time sleep staging on the collected EEG data;

[0018] S2: Once the subject is detected to have entered a stable N3 phase, the data quality detection module detects the EEG data and removes abnormal data; then the sleep depth calculation module calculates the sleep depth; and adjusts the volume of the music stimulation according to the subject's sleep depth.

[0019] Among them, sleep depth is directly proportional to the logarithmic ratio of the power of the delta band and beta band EEG (obtained from the power spectrum after fast Fourier transform), and the calculation formula is as follows:

[0020]

[0021] in, Sleep depth, This refers to the EEG power in the delta band. This refers to the power of the beta-band EEG.

[0022] The volume of music stimulation The relationship between sleep depth and sleep duration is shown in the following formula:

[0023]

[0024] in, The preset initial volume;

[0025] S3: Simultaneously, the real-time slow wave detection module starts detecting the rising edge of the slow wave. After detecting the rising edge of the slow wave, the subject will be given a pink noise stimulus near the top of the rising edge of the slow wave. After giving two consecutive pink noise stimuli, cool for 2.5 seconds.

[0026] Among them, the slow wave rising edge that meets the conditions is: each input of 15 consecutive points, after smoothing, their arrangement shows a monotonically rising trend and an inflection point relative to the baseline, which is the slow wave rising edge that meets the conditions.

[0027] S4: Repeat step S3 until the subject falls asleep.

[0028] As a preferred approach, when testing for significant differences in slow wave amplitude and slope, the slow wave amplitude and slope of each slow wave in the raw EEG signal of each subject in both experimental groups are calculated, and the median and interquartile range of all slow wave amplitudes and slopes in the experimental groups are calculated; and the distribution of slow wave amplitude and slope between the two experimental groups is tested to determine if there are significant differences.

[0029] When testing for significant differences in the number of spindle waves, the average number of spindle waves detected in each experimental group was taken according to the number of subjects in that experimental group, and the significance of the number of spindle waves between the two experimental groups was tested.

[0030] When testing for significant differences in total power during the NREM period, EEG signal segments from the light and deep sleep phases of each subject were extracted, and the sum of the power of the five bands (delta, theta, alpha, sigma, and beta) in the EEG signal segments of each subject was calculated. The mean and variance of the sum of the power of each band were calculated for the two experimental groups according to the number of subjects, and the significance of the total power during the NREM period between the two experimental groups was tested.

[0031] When analyzing the effect of music stimulation on sleep memory, the mean and variance of the memory consolidation test results of the two experimental groups were calculated, and the significance of the memory consolidation test results between the two experimental groups was examined.

[0032] When analyzing the correlation between sleep memory and the proportion of deep sleep, the Pearson correlation coefficient method was used to calculate the correlation between the memory consolidation test results of each experimental group and the corresponding proportion of deep sleep, as well as the correlation between the memory consolidation test results of all subjects in both groups and the corresponding proportion of deep sleep.

[0033] Preferably, in step 4), a significant difference analysis is performed on the waveform changes near the initial musical stimulus. The specific method is as follows:

[0034] The original signal segments from 4 seconds before to 8 seconds after the first music stimulation time point were extracted from the EEG signals of each subject in the two experimental groups. The average signal of the original signal segments of the two experimental groups was calculated, and the phase of the average signal of the two experimental groups was aligned. The average signal is the average of the amplitude of each original signal.

[0035] The average signal segments from 1 second before to 5 seconds after the first stimulation time point of the average signal of the two experimental groups were taken. After aligning the average signal segments of the two experimental groups, a point-to-point test was conducted to examine whether there was a significant difference between the average signal segments of the two experimental groups.

[0036] As a preferred method, when verifying the effect of music stimulation on the power of deep sleep, the deep sleep EEG signal segments of each subject in the two experimental groups were extracted, and the deep sleep EEG signal segments were divided into bins with a fixed frequency width. The deep sleep EEG signal segments of each subject were divided into several band bins, and the power ratio of each band bin was calculated. The average and variance of the power ratio of the band bins were calculated for the two experimental groups according to the number of subjects, and the significance of the difference was tested between the two experimental groups.

[0037] Preferably, the frequency range of the EEG signal segment during deep sleep is 0.25Hz to 20Hz, and the fixed frequency bandwidth is 0.25Hz.

[0038] The beneficial effects of this invention are: This invention systematically analyzes the effects of music stimulation on slow-wave sleep and the verification method of the relationship between slow-wave sleep and memory, so as to promote in-depth research on human sleep. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the process of the present invention.

[0040] Figure 2 This is a schematic diagram of the control process of a closed-loop control system for sleep music.

[0041] Figure 3 This is a comparison of the changes and significance of the average signal segments from 1 second before to 5 seconds after the first stimulation time point in the two experimental groups.

[0042] Figure 4 A scatter plot comparing the difference in word association test results before and after sleep with the proportion of slow-wave sleep.

[0043] Figure 5 This is a comparison and significance chart showing the relative power percentage of each band during deep sleep in the two experimental groups. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0045] Those skilled in the art should understand that, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention.

[0046] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0047] like Figures 1-5 As shown, a method for verifying the effects of sleep music closed-loop intervention on slow waves and memory includes the following specific steps:

[0048] Step 1) Recruit participants and have each participant participate in the Adaptation Night Experiment (ADAP), Stimulation Night Experiment (STIM), and Sham Stimulation Night Experiment (SHAM). All participants first undergo the Adaptation Night Experiment, during which no musical stimulation is given to the participants during sleep. Each Adaptation Night Experiment, Stimulation Night Experiment, and Sham Stimulation Night Experiment lasts for 7 days.

[0049] Subjects prepared the experimental environment before going to sleep and informed others not to disturb them, so as to avoid interference with the experimental process.

[0050] Step 2) After completing the adaptation night experiment, half of the subjects were randomly selected to undergo the stimulation night experiment first and then the sham stimulation night experiment, while the other half of the subjects underwent the sham stimulation experiment first and then the stimulation night experiment.

[0051] In the stimulation night experiment and the sham stimulation night experiment, subjects were given a pre-sleep word association test before bedtime. In the sham stimulation night experiment, subjects were not given music stimulation during sleep, and their EEG data were collected during sleep. In the stimulation night experiment, subjects' EEG data were collected during sleep, and a closed-loop sleep music control system was activated. The closed-loop sleep music control system performed real-time sleep staging on the collected EEG signals. When the subject was detected to be in deep sleep, the music playback device was controlled to provide music stimulation to the subject.

[0052] After the subjects finished sleeping, EEG data collection and music stimulation were stopped. Then, a post-sleep word association test was administered. This test was conducted 30-40 minutes after the experiment. The difference between the post-sleep word association test result and the pre-sleep word association test result was used as the subject's memory consolidation test result. The word association test used in this invention is existing technology. The test score ranges from 0 to 100 points. The word association test result reflects the subject's memory level; a higher score indicates a higher memory level.

[0053] The sleep music closed-loop control system includes data quality detection, real-time sleep staging module, sleep depth calculation module, and real-time slow wave detection module.

[0054] like Figure 2 As shown, the control method of the sleep music closed-loop control system is as follows:

[0055] S1: The real-time sleep staging module performs real-time sleep staging on the collected EEG data; the real-time sleep staging module divides the subject's sleep into the wakefulness stage, non-rapid eye movement stage 1 (N1 stage), non-rapid eye movement stage 2 (N2 stage), non-rapid eye movement stage 3 (N3 stage), and rapid eye movement (REM stage);

[0056] S2: Once the subject is detected to have entered a stable N3 phase, the data quality detection module detects the EEG data and removes abnormal data; then the sleep depth calculation module calculates the sleep depth; and adjusts the volume of the music stimulation according to the subject's sleep depth.

[0057] Among them, sleep depth is directly proportional to the logarithmic ratio of the power of the delta band and beta band EEG (obtained from the power spectrum after fast Fourier transform), and the calculation formula is as follows:

[0058]

[0059] in, Sleep depth, This refers to the EEG power in the delta band. This refers to the power of the beta-band EEG.

[0060] The volume of music stimulation The relationship between sleep depth and sleep duration is shown in the following formula:

[0061]

[0062] in, The preset initial volume;

[0063] S3: Simultaneously, the real-time slow wave detection module starts detecting the rising edge of the slow wave. After detecting a slow wave rising edge that meets the conditions, the subject will be given a pink noise stimulus near the top of the rising edge of the slow wave. After giving two consecutive pink noise stimuli, the subject will be cooled for 2.5 seconds.

[0064] Among them, the slow wave rising edge that meets the conditions is: each input of 15 consecutive points, after smoothing, their arrangement shows a monotonically rising trend and an inflection point relative to the baseline, which is the slow wave rising edge that meets the conditions.

[0065] S4: Repeat step S3 until the subject falls asleep.

[0066] The music playback device used in this step is a music pillow. The music pillow connects to the computer via Bluetooth. Before conducting the stimulation night experiment, the operator needs to select music according to the subject's personal preferences and continuously adjust the volume until the subject reaches a comfortable state. This volume is used as the preset initial volume and is denoted as [insert initial volume here]. .

[0067] Step 3) Using the trained deep learning model, the subject's EEG data is divided into four sleep stages: wakefulness (WK), rapid eye movement (REM) sleep, light sleep, and deep sleep. The light and deep sleep stages are collectively referred to as NREM (non-rapid eye movement sleep). The proportion of each sleep stage in the entire sleep process is calculated, namely the proportion of wakefulness, REM sleep, light sleep, and deep sleep. The proportion of deep sleep is also called the slow-wave sleep proportion. The following sleep indicators are calculated: total power of NREM, number of spindle waves, slow-wave power, slow-wave amplitude, and slow-wave slope. The slow-wave slope is the ratio of the difference between the vertical coordinate (i.e., amplitude) of the peak and trough of each slow wave to the difference between the horizontal coordinate of the peak and trough.

[0068] Step 4) Assign the subjects who participated in the Stimulated Night Experiment to the Stimulated Night Experiment Group, and assign the subjects who participated in the Pseudo-Stimulated Night Experiment to the Pseudo-Stimulated Night Experiment Group; compare the various sleep indicators of the Stimulated Night Experiment Group and the Pseudo-Stimulated Night Experiment Group, and perform a significance test for the differences; and analyze the effect of music stimulation on sleep memory and the correlation between sleep memory and the proportion of deep sleep.

[0069] In the significance test for slow-wave amplitude and slope, the slow-wave amplitude and slope of each slow wave in the raw EEG signals of each subject in both experimental groups were calculated, and the median and interquartile range of all slow-wave amplitudes and slopes in each experimental group were calculated. The study then examined whether there were significant differences in slow-wave amplitude and slope between the two experimental groups. If a significant difference existed, it indicated that music stimulation during sleep had a significant impact on the slow-wave amplitude and slope of sleepers, and significantly affected slow-wave sleep (i.e., deep sleep); otherwise, it indicated that music stimulation during sleep had no significant impact on the slow-wave amplitude and slope of sleepers.

[0070] When conducting a significance test on the number of spindle waves, the average number of spindle waves detected in each experimental group was calculated based on the number of subjects in that group, and the difference in the number of spindle waves between the two experimental groups was then examined. If a significant difference exists, it indicates that music stimulation during sleep has a significant effect on the number of spindle waves in sleepers; otherwise, it indicates that music stimulation during sleep has no significant effect on the number of spindle waves in sleepers.

[0071] When testing for significant differences in total power during the NREM period, EEG signal segments from the light and deep sleep phases of each subject were extracted. The sum of the power of the five bands (delta, theta, alpha, sigma, and beta) in the EEG signal segments of each subject was calculated. The mean and variance of the sum of the band power were calculated for both experimental groups according to the number of subjects, and a significance test was performed between the two experimental groups. The frequencies of the delta band were between 0.5-4 Hz, the theta band between 4-8 Hz, the alpha band between 8-12 Hz, the sigma band between 12-16 Hz, and the beta band between 16-20 Hz.

[0072] In analyzing the impact of music stimulation on sleep memory, the mean and variance of the memory consolidation test results for the two experimental groups were calculated, and a significant difference was examined between the two groups. If a significant difference existed, it indicated that music stimulation during sleep had a significant impact on the sleeper's memory; otherwise, it indicated that music stimulation during sleep had no significant impact on the sleeper's memory.

[0073] When analyzing the correlation between sleep memory and the proportion of deep sleep (slow-wave sleep), the Pearson correlation coefficient method was used to calculate the correlation between the memory consolidation test results of each experimental group and the corresponding proportion of deep sleep, as well as the correlation between the memory consolidation test results of all subjects in both groups and the corresponding proportion of deep sleep. If the correlation is valid, it indicates that the proportion of deep sleep during sleep is closely related to the level of memory after sleep; otherwise, it indicates that there is no significant relationship between the proportion of deep sleep during sleep and the level of memory after sleep.

[0074] To verify the effect of music stimulation on power during deep sleep, EEG signal segments during deep sleep were extracted from each subject in both experimental groups. These segments were then divided into bins with a fixed frequency width. Each subject's deep sleep EEG signal segment was further divided into several band bins, and the power percentage of each band bin (i.e., the proportion of power of each band bin within the entire deep sleep EEG signal segment) was calculated. The mean and variance of the band bin power percentages were calculated for both experimental groups based on the number of subjects, and a significance test was performed between the two groups. The frequency range of the deep sleep EEG signal segment was 0.25 Hz to 20 Hz, and the fixed frequency width was 0.25 Hz.

[0075] In step 4), a significance analysis of the waveform changes near the initial musical stimulus is also performed. The specific method is as follows:

[0076] The original signal segments from 4 seconds before to 8 seconds after the first music stimulation time point were extracted from the EEG signals of each subject in the two experimental groups. The average signal of the original signal segments of the two experimental groups was calculated, and the phase of the average signal of the two experimental groups was aligned. The average signal is the average of the amplitude of each original signal.

[0077] The average signal segments from 1 second before to 5 seconds after the first stimulus time point were taken from the average signal segments of the two experimental groups. After aligning the average signal segments of the two experimental groups, a point-to-point method was used to examine whether there was a significant difference between the average signal segments of the two experimental groups. Specific results are as follows: Figure 3 As shown, Figure 3 The results show that there are several bands with significant differences between the average signal segments of the two experimental groups.

[0078] The method for testing the significance of differences is as follows:

[0079] Establish the null hypothesis H0 and the alternative hypothesis H1: H0: μ 对照组 = μ 干预组 H1: μ 对照组 ≠ μ 干预组 μ is the average value of the sample group; in this invention, the intervention group in the above hypothesis is the stimulation experiment group, and the control group in the above hypothesis is the sham stimulation experiment group.

[0080] The normality and homogeneity of variance of the test data between the control group and different music intervention groups were tested using the Shapiro-Wilk test and the Levene test. Specifically, if the output values ​​of both the Shapiro-Wilk test and the Levene test are greater than 0.05, it indicates that the sleep indicators of the control group and different music intervention groups satisfy the requirements of normal distribution and homogeneity of variance.

[0081] a: If the test data of the control group and the intervention group satisfy the normality of distribution and homogeneity of variance, perform a t-test and calculate the t-statistic. The formula for calculating the t-statistic is as follows:

[0082]

[0083] And calculate the degrees of freedom, the magnitude of which is + -2; where, These are the average values ​​of the control group and the intervention group, respectively. The variances of the control group and the intervention group are respectively. These are the sample sizes for the control group and the intervention group, respectively.

[0084] Based on the degrees of freedom and t-statistic, the probability p-value of the null hypothesis is obtained by referring to the table. If the probability p-value is <0.05, the difference in indicators between the intervention group and the control group is considered statistically significant; otherwise, the difference in indicators between the intervention group and the control group is considered not significant and not statistically significant.

[0085] b: If the test data of the control group and the intervention group do not meet the requirements of normal distribution and homogeneity of variance, then the Mann-Whitney U test should be performed. The test method is as follows:

[0086] The test data from the control group and the intervention group were mixed and arranged in ascending order to form a data set. The largest data point was ranked n1+n2, the next largest n1+n2-1, and so on, with the smallest data point ranked 1. If data points of the same size were encountered, their ranking was determined as the average of the rankings of these data points. For example, if the data was arranged in ascending order as 0, 1, 1, 1, 2, 3, 3, then the ranking of observation 1 would be (2+3+4) / 3 = 3, and the value of observation 3 would be (6+7) / 2 = 6.5. Therefore, the final data set would be 1, 3, 3, 3, 5, 6.5, 6.5.

[0087] Then, the sum of the ranks of the data from the control group and the intervention group was calculated separately. , And calculate the Mann-Whitney U statistic for the data sets. , ; , The calculation formula is as follows:

[0088]

[0089]

[0090] choose , The smallest of the three, based on its sample size and critical value U. α The critical value U α The null hypothesis is defined as the minimum U-value required to reject the null hypothesis (H0) when the sizes of the two sets of data are known. The probability p-value of the null hypothesis is obtained by looking up a table. If the probability p-value is <0.05, the null hypothesis is rejected, and the difference in indicators between the intervention group and the control group is considered significant and statistically meaningful; otherwise, the difference in indicators between the intervention group and the control group is considered not significant and not statistically meaningful.

[0091] When analyzing the correlation between sleep memory and the proportion of deep sleep (slow-wave sleep) using the Pearson correlation coefficient method, the method is as follows:

[0092] (1) Draw a scatter plot and corresponding trend line between the independent variable X (percentage of deep sleep) and the dependent variable Y (difference between the results of the pre- and post-sleep word association test, i.e., the results of the memory consolidation test), and observe whether there is a linear relationship between the two variables. The scatter plot and corresponding trend line are as follows: Figure 4 As shown;

[0093] (2) Test whether the two variables (i.e., the percentage of deep sleep and the difference between the results of the word association test before and after sleep) conform to or are approximately normally distributed according to the difference significance test method described above.

[0094] (3) Based on satisfying (1) and (2), calculate the Pearson correlation coefficient:

[0095]

[0096] and The average values ​​of the independent and dependent variables, respectively.

[0097] (4) Perform hypothesis testing on the correlation coefficient:

[0098] a. Establish the null hypothesis H0 and the alternative hypothesis H1:

[0099] H0: r = 0, H1: r ≠ 0

[0100] b. Calculate the t-statistic:

[0101]

[0102] n is the capacity of the samples in the group;

[0103] c. Calculate the degrees of freedom (n-2). Based on the degrees of freedom and the t-statistic, look up the table to obtain the probability p-value of the null hypothesis. If p < 0.05, we have sufficient confidence to reject the null hypothesis and believe that there is indeed a correlation between the independent and dependent variables; otherwise, we believe that the correlation is not significant and is caused by sampling error or other factors.

[0104] This invention systematically analyzes the effects of music stimulation on slow-wave sleep and verifies the relationship between slow-wave sleep and memory, in order to promote in-depth research on human sleep.

[0105] This invention is not limited to the preferred embodiments described above. Anyone can derive other products in various forms under the guidance of this invention. However, regardless of any changes in shape or structure, any technical solution that is the same as or similar to this application falls within the protection scope of this invention.

Claims

1. A method for verifying the effects of closed-loop intervention with sleep music on slow waves and memory, characterized in that, The specific steps include the following: Step 1) Recruit subjects and have each subject participate in the adaptation night experiment, the stimulation night experiment, and the sham stimulation night experiment; all subjects first undergo the adaptation night experiment, during which no musical stimulation is given to the subjects during sleep. Step 2) After completing the adaptation night experiment, half of the subjects were randomly selected to undergo the stimulation night experiment first and then the sham stimulation night experiment, while the other half of the subjects underwent the sham stimulation experiment first and then the stimulation night experiment. In the stimulation night experiment and the sham stimulation night experiment, subjects were given a pre-sleep word association test before bedtime. In the sham stimulation night experiment, subjects were not given music stimulation during sleep, and their EEG data were collected during sleep. In the stimulation night experiment, subjects' EEG data were collected during sleep, and a closed-loop sleep music control system was activated. The closed-loop sleep music control system performed real-time sleep staging on the collected EEG signals. When the subject was detected to be in deep sleep, the music playback device was controlled to provide music stimulation to the subject. After the subjects finished sleeping, they were given a post-sleep word association test. The difference between the post-sleep word association test results and the pre-sleep word association test results was used as the subjects' memory consolidation test results. Step 3) Using the trained deep learning model, the subject's EEG data is divided into four sleep stages: wakefulness, REM sleep, light sleep, and deep sleep. The following sleep indicators are calculated: total power of NREM sleep, number of spindle waves, slow wave power, slow wave amplitude, and slow wave slope. The two sleep stages of light sleep and deep sleep are collectively referred to as the NREM sleep stage. Step 4) Assign subjects who participated in the Stimulated Night Experiment to the Stimulated Night Experiment Group, and subjects who participated in the Pseudo-Stimulated Night Experiment to the Pseudo-Stimulated Night Experiment Group; compare various sleep indicators between the Stimulated Night Experiment Group and the Pseudo-Stimulated Night Experiment Group, and perform significance tests on the differences; analyze the effect of music stimulation on sleep memory and the correlation between sleep memory and the proportion of deep sleep; when analyzing the effect of music stimulation on sleep memory, calculate the mean and variance of the memory consolidation test results of the two experimental groups, and test whether there is a significant difference between the memory consolidation test results of the two experimental groups; when analyzing the correlation between sleep memory and the proportion of deep sleep, use the Pearson correlation coefficient method to calculate the correlation between the memory consolidation test results of each experimental group and the corresponding proportion of deep sleep, as well as the correlation between the memory consolidation test results of all subjects in both groups and the corresponding proportion of deep sleep.

2. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, In step 1), the duration of each adaptation night experiment, stimulation night experiment, and sham stimulation night experiment is 7 days.

3. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, In step 2), the music playback device is a music pillow.

4. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, In step 2), the sleep music closed-loop control system includes a data quality detection module, a real-time sleep staging module, a sleep depth calculation module, and a real-time slow wave detection module. The control method for the sleep music closed-loop control system is as follows: S1: The real-time sleep staging module performs real-time sleep staging on the collected EEG data; S2: Once the subject is detected to have entered a stable N3 phase, the data quality detection module detects the EEG data and removes abnormal data; then the sleep depth calculation module calculates the sleep depth; and adjusts the volume of the music stimulation according to the subject's sleep depth. The formula for calculating sleep depth is as follows: in, Sleep depth, This refers to the EEG power in the delta band. This refers to the power of the beta-band EEG. The volume of music stimulation The relationship between sleep depth and sleep duration is shown in the following formula: in, The preset initial volume; S3: Simultaneously, the real-time slow wave detection module starts detecting the rising edge of the slow wave. After detecting the rising edge of the slow wave, the subject will be given a pink noise stimulus near the top of the rising edge of the slow wave. After giving two consecutive pink noise stimuli, cool for 2.5 seconds. Among them, the slow wave rising edge that meets the conditions is: each input of 15 consecutive points, after smoothing, their arrangement shows a monotonically rising trend and an inflection point relative to the baseline, which is the slow wave rising edge that meets the conditions. S4: Repeat step S3 until the subject falls asleep.

5. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, When testing for significant differences in slow wave amplitude and slope, the slow wave amplitude and slope of each slow wave in the raw EEG signal of each subject in the two experimental groups were calculated, and the median and interquartile range of all slow wave amplitudes and slopes in the experimental groups were calculated. And examine whether there are significant differences in the distribution of slow wave amplitude and slow wave slope between the two experimental groups; When testing for significant differences in the number of spindle waves, the average number of spindle waves detected in each experimental group was taken according to the number of subjects in that experimental group, and the significance of the number of spindle waves between the two experimental groups was tested. When testing for significant differences in total power during the NREM period, EEG signal segments from the light and deep sleep phases of each subject were extracted, and the sum of the power of the five bands (delta, theta, alpha, sigma, and beta) in the EEG signal segments of each subject was calculated. The mean and variance of the sum of the power of each band were calculated for the two experimental groups according to the number of subjects, and the significance of the total power during the NREM period between the two experimental groups was tested.

6. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, In step 4), a significance analysis of the waveform changes near the initial musical stimulus is performed. The specific method is as follows: The original signal segments from 4 seconds before to 8 seconds after the first music stimulation time point were extracted from the EEG signals of each subject in the two experimental groups. The average signal of the original signal segments of the two experimental groups was calculated and the phase of the average signals of the two experimental groups was aligned. The average signal segments from 1 second before to 5 seconds after the first stimulation time point of the average signal of the two experimental groups were taken. After aligning the average signal segments of the two experimental groups, a point-to-point test was conducted to examine whether there was a significant difference between the average signal segments of the two experimental groups.

7. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 1, characterized in that, To verify the effect of music stimulation on power during deep sleep, the EEG signal segments during deep sleep were extracted from each subject in both experimental groups. The EEG signal segments during deep sleep were divided into bins with a fixed frequency width. Each subject's EEG signal segment during deep sleep was divided into several band bins, and the power proportion of each band bin was calculated. The mean and variance of the power proportion of the band bins were calculated for each experimental group according to the number of subjects, and a significance test was performed between the two experimental groups.

8. The method for verifying the effects of sleep music closed-loop intervention on slow waves and memory according to claim 7, characterized in that, The frequency range of the EEG signal segment during deep sleep is 0.25Hz to 20Hz, and the fixed frequency bandwidth is 0.25Hz.

Images