An intelligent evaluation method for the efficacy of sleeping patch products

Abstract

The present invention belongs to the field of intelligent medical technology, and specifically relates to an intelligent evaluation method for the effectiveness of a sleep patch product; the collected signal data is subjected to heart rate variability analysis, electrocardiogram signal and respiratory waveform correlation analysis, heart rate variability parameter combination and linear regression model prediction of heart rate variability. Then, the sleep quality and respiratory quality of the evaluation population are evaluated based on characteristic parameters and indicators; by analyzing the heart rate signal and calculating the heart rate variability parameters, the number of respiratory events is predicted based on a method combining the heart rate variability parameters with a linear regression model. The present application utilizes heart rate variability parameters in combination with a polysomnography monitor to predict and evaluate respiratory signals, and the indirect methods obtained by the two methods are used to compare and analyze the evaluation results of respiratory signals, thereby improving the accuracy of the data and providing a guarantee for an accurate evaluation of the therapeutic effect of the sleep patch product.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent medical technology, and specifically relates to an intelligent evaluation method for the effectiveness of a sleeping patch product. Background Art

[0002] When anxiety or depression occurs, the sympathetic nervous system of the human body is affected, triggering the release of a series of stress response hormones, which leads to sleep disorders. When long-term sleep disorders occur, they will lead to changes in the physiological characteristics and psychology of the human body, and further aggravate sleep disorders and sub-health conditions. Therefore, studying how to reduce the release of stress response hormones in the sympathetic nervous system has become the key to treating sleep disorders.

[0003] In the prior art, when sleep patch products are used to treat sleep disorders, they mainly use physiological parameter indicators such as heart rate and breathing to determine the effect of sleep intervention, and do not involve heart rate variability parameters. These parameter characteristics can objectively evaluate the degree of influence on the sympathetic nervous system, so there is still a lot of room for improvement in the treatment effect evaluation method. Due to the above reasons, methods and technologies that can quantitatively evaluate heart rate variability are needed to effectively evaluate the treatment effect of sleep patch products on specific sleep disorders, so as to more objectively evaluate the tension of the sympathetic nervous system.

[0004] The results of heart rate variability analysis are mainly based on frequency domain analysis, which can fully understand the tension of the human sympathetic nervous system and parasympathetic nervous system. In heart rate variability analysis, low frequency power (LF) mainly reflects the activity of sympathetic nerves; high frequency power (HF) mainly reflects the activity of parasympathetic nerves; low-high frequency ratio (LF / HF) reflects the balance between the sympathetic nervous system and the parasympathetic nervous system. Heart rate variability can also calculate the difference (RRin-RR) between adjacent RR time periods from time series analysis to calculate PNN50, to reflect the degree of change in the time series. Heart rate variability parameter indicators are the most direct and effective parameters to reflect the state of autonomic nervous function.

[0005] Therefore, in evaluating the effectiveness of sleep patch products in treating specific sleep disorders, the use of heart rate variability indicators and parameters for evaluation and research can further enhance the activity of the parasympathetic nerves at night and help improve the patient's sleep. At the same time, analyzing the frequency of various respiratory events, understanding the patient's awakening caused by various types of breathing, and evaluating the patient's degree of respiratory disorders also have direct reference significance for the treatment effect.

[0006] In view of the above reasons, when conducting intelligent evaluation of the therapeutic effect of sleep patch products on specific sleep disorders, multi-parameter indicators are used for quantitative evaluation; objective quantitative evaluation combined with the patient's subjective sleep experience is the key technical problem to be solved in this study.

[0007] In view of the above problems, the present invention proposes an intelligent evaluation method for the effectiveness of a sleeping patch product. Summary of the invention

[0008] In order to solve the problems of the prior art, the present invention aims to provide an intelligent evaluation method for the effectiveness of a sleep patch product based on the enhanced effects of heart rate variability and respiratory variability of sleep physiological signals, using data curves and statistical operations, targeting changes in frequency characteristics of a certain specific human body signal.

[0009] The present invention discloses a method for intelligently evaluating the effectiveness of a sleep patch product, comprising the following steps: equipment selection and preparation: selecting a suitable target population, installing and connecting corresponding equipment, collecting physiological data, and calibrating the equipment; data collection: conducting a sleep intervention on the target population for one week through a sleep patch product, collecting physiological signal data and sleep observation data, and preprocessing the data; data analysis: extracting physiological signals during the sleep period, extracting and calculating characteristic parameters that characterize sleep and breathing characteristic values; result evaluation and feedback: using an algorithm model to perform cluster analysis on sleep parameters during the sleep period, generating a corresponding report, and evaluating the results based on the characteristic performance of the clusters and the standard. Quasi-phase comparison, summarize the improvement of sleep characteristics and breathing characteristics, and generate feedback results; continuous monitoring and optimization: feedback results are fed back to the device to adjust and optimize the sleep patch; the specific steps of data analysis are: step one, data extraction, extract various physiological parameters of data samples from data points, and use clustering algorithm to classify, and obtain characteristic curves; step two, data preprocessing, denoising the characteristic curve obtained in step one; step three, data analysis, regression analysis of the data curve after data preprocessing, calculate the regression equation and curve fitting goodness of fit, and determine the correlation of the curves and the changes between the characteristic curves according to the fitting goodness of fit.

[0010] Specifically, the data preprocessing step includes: dividing the data point context, calculating the average heart rate 2 seconds before and 5 seconds after the data point as the background value of the current data point, and then calculating the absolute difference ,in Refers to the absolute deviation between the data point and the background value. Results that meet the absolute difference will be counted into the snapshot.

[0011] Specifically, in the data analysis of step 3, heart rate fluctuation index and heart rate variability analysis are performed; parameters of heart rate, electrocardiogram, pulse and respiratory physiological signals are extracted from each sleep period; HR variability parameters are calculated from the heart rate signal: taking the entire electrocardiogram of 1 minute, the heart rate variability time domain analysis parameters are: number of premature ventricular contractions , maximum degree of premature ventricular contraction ; The frequency domain analysis parameters of heart rate variability are ultra-low frequency , low frequency ,high frequency , , the transition characteristics from low frequency domain to high frequency domain: .

[0012] Specifically, RR, P-QRS, PR, PP intervals, RR interval, RR-RR interval, PR interval, and RR interval are calculated from the QT interval signal, and the variability of the RR interval is calculated and used as an objective indicator standard. The calculation formula for the RR interval variability indicator is:

[0013]

[0014]

[0015]

[0016]

[0017]

[0018]

[0019] in: is the RR interval sequence, is the standard deviation, is the first Segment value, The first arrive The standard deviation between segments, and is a natural number, is the mean of all RR intervals, .

[0020] Specifically, it also includes comparing the objective change results of the monitoring target with the standard value: when the calculated average RR interval value is <350ms and the standard deviation of the RR interval is <50ms, the monitoring target is judged to have a normal sinoatrial node signal; when the average RR interval is >500ms and the standard deviation of the RR interval is >120ms, the monitoring target is judged to have an abnormal sinoatrial node signal; when the average RR interval result is basically stable within the normal value range, and the RR time series variability SD1<50ms, SD2<75ms, it indicates that the conduction function of the sinoatrial node efferent nerve is within the normal range; when the average RR interval value basically shows a downward trend, and SD1 has an increasing trend, and SD2 increases significantly, it indicates that the sinoatrial node efferent function is weakened; perform heart rate time domain and frequency domain analysis through data extraction: if VM>4 in the frequency domain analysis result of heart rate variability, it means that the heart rate variability is large; VM<2 indicates that the heart rate variability is low; when the transition characteristic from low frequency to high frequency is HF>50%, the heart rate variability tends to improve, and when the high frequency is >80%, the heart rate variability is small.

[0021] Specifically, it also includes generating a characteristic curve diagram of heart rate changes in each sleep period from the test data in a visual manner: the characteristic curve diagram presents the characteristics of heart rate oscillations, with the horizontal and vertical axes representing the average heart rate per minute, the heart rate oscillation amplitude, and the frequency value of the heart rate RR interval. When the horizontal axis is constant, an increase in the heart rate amplitude and frequency proves that the heart rate variability increases, and the stronger the heart rate variability; when the frequency is constant, an increase in the amplitude indicates that the degree of heart rate oscillation increases, and vice versa, the degree of heart rate oscillation decreases.

[0022] Specifically, the data collection and data analysis steps are performed in two or more rounds.

[0023] Specifically, the first round of data collection and data analysis process includes: Step S1: Use a sleep patch to perform sleep intervention treatment on the human body for one week to obtain physiological signals and respiratory signals during sleep, and the physiological signals include heart rate, respiratory rate, and blood oxygen signal; Step S2: Convert the acquired physiological signals and respiratory signals during sleep into text files; Step S3: Use the relevant data functions of R language to analyze the text files to obtain characteristic parameters that characterize sleep and respiratory characteristics; The specific data processing steps of the data analysis include: Step S4: Filter the waveform after the characteristics; Step S5: Calculate the characteristic parameters based on the filtering results; The data processing performs electrocardiogram, heart rate, respiration and blood oxygen waveform processing; ECG signal processing, the steps include: selecting ECG signal, filtering, normalizing, then extracting R wave, extracting corresponding waveforms Q and T wave, calculating the deviation of QT time relative to the average heart rate, QT interval ratio and QT and RR correlation value, and performing feature analysis; heart rate, respiration, and blood oxygen signal processing, the steps include: heart rate signal denoising, respiration and blood oxygen windowing processing, calculating the mean of heart rate, respiration and blood oxygen, extracting RR interval, then calculating the RR interval mean and RR variance, performing variability of QT interval, RR interval and RR time series to obtain characteristic values; performing statistical analysis, algorithm model fit test and characteristic curve graphical analysis on the characteristic parameters.

[0024] Specifically, the second round of data collection and data analysis process also includes the following detailed steps: Step S5: After analyzing according to the data formula and algorithm model in Step 3, output the characteristic data; Step S6: Use the average heart rate mean and standard deviation in the characteristic data of the heart rate signal of the next day of the output data of Step S5 as parameters to judge the fluctuation of the sleep physiological signal; Step S7: Use the output data of Step S5 as parameters to judge the variability of the RR time series; Step S8: Based on the output data of Step S5, make a variability judgment on the RR interval; Step S9: Based on the output data of Step S5, judge the heart rate variability; and Step S10: Perform a graphical comparative analysis of the changes of each indicator in different sleep periods; Among them, in the step of using a sleep patch to intervene in the human body for one week, the sensitivity of the target population and different populations to the sleep intervention product is analyzed. The evaluation indicators of sleep data of different groups should be calibrated before execution, and the calibration includes the following steps: Step S11: In multi-target sleep intervention, target heart rate, respiratory waveform and blood oxygen signal are collected; Step S12: The signal collected in Step S13 is subjected to noise reduction processing to extract R wave peak information; Step S13: Gait detection is performed through the information of Step S12 to determine different sleep periods; Step S14: According to the sleep detection information in Step S13, the mean and standard deviation of each sleep period are calculated respectively; Step S15: The heart rate standard deviation and RR interval fluctuation indicators in Step S13 and Step S14 can be used as evaluation indicators for judging the effect of heart rate intervention on different groups of people, and the indicators can be adjusted according to the fluctuation of the group; wherein the sensitivity of different groups of people to heart rate variability includes different sensitivity of athletes and patients with depression.

[0025] Specifically, it also includes a long-term evaluation of the trend change of heart rate variability, the steps are: using the same test target group as the research object, comparing the evaluation indicators for four consecutive weeks, including various indicators of heart rate variability, electrocardiogram, respiration and blood oxygen signals, comparing the changes before and after, and judging the regulatory effect of the sleep patch on the heart rate of a specific population. The beneficial effect of the present invention is that the method of the present invention can automatically extract characteristic indicators related to heart rate and respiration, calculate the goodness of fit of the data, and perform adaptive heart rate fluctuations and the degree of respiratory disorders, calculate the heart rate fluctuation trend and the heart rate variability change according to the changes in the heart rate related indicator values, so as to obtain the heart rate change results and obtain the heart rate regulation effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is the overall logic block diagram of the present invention.

[0027] Figure 2 It is a logic block diagram of the heart rate variability analysis of the present invention.

[0028] Figure 3 It is a data preprocessing logic block diagram of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] The present invention provides a method for intelligently evaluating the effectiveness of a sleep patch product. The method provides a scientific and objective basis for evaluating the effectiveness of a sleep patch product by systematically collecting, analyzing and evaluating sleep-related physiological data. The technical solution of the present invention will be described in detail below in conjunction with specific embodiments.

[0031] like Figure 1-3 As shown, the intelligent assessment method of the present invention includes the following main steps: equipment selection and preparation: select a suitable target population, install and connect the corresponding equipment, collect physiological data, and calibrate the equipment; data collection: conduct a one-week sleep intervention on the target population through the sleep patch product, collect physiological signal data and sleep observation data, and pre-process the data; data analysis: extract physiological signals during the sleep period, extract and calculate characteristic parameters that characterize sleep and breathing characteristic values; result evaluation and feedback: use the algorithm model to perform cluster analysis on the sleep parameters during the sleep period, generate a corresponding report, and compare the clustered characteristic performance results with the standard , summarize the improvement of sleep characteristics and breathing characteristics, and generate feedback results; continuous monitoring and optimization: the feedback results are fed back to the device to adjust and optimize the sleep patch; the specific steps of data analysis are: step one, data extraction, extract the various physiological parameters of the data sample from the data points, and use the clustering algorithm to classify them to obtain the characteristic curve; step two, data preprocessing, denoising the characteristic curve obtained in step one; step three, data analysis, regression analysis of the data curve after data preprocessing, calculate the regression equation and curve fitting goodness, and determine the correlation of the curve and the changes between the characteristic curves according to the fitting goodness.

[0032] First, in the equipment selection and preparation step, it is necessary to select a suitable target population, and install and connect the corresponding monitoring equipment. Preferably, in one embodiment of the present invention, the monitoring equipment includes but is not limited to an electrocardiograph, a respiratory monitor, and an oximeter. These devices can collect a variety of physiological data including heart rate, respiratory rate, electrocardiogram, and blood oxygen saturation. After connecting the equipment, the equipment needs to be calibrated to ensure the accuracy of data collection.

[0033] Next, in the data collection step, the target population will be given a sleep intervention for one week through the sleep patch product. During this process, the monitoring device will continuously collect physiological signal data and sleep observation data. It is worth noting that in order to ensure the comprehensiveness and reliability of the data, it is recommended to collect at least 8 hours of continuous data every night.

[0034] Then, we move on to the data analysis step. This step is the core of the entire evaluation method and contains multiple sub-steps. First, the collected raw data needs to be preprocessed. This includes three main steps: data extraction, data preprocessing, and data analysis.

[0035] In the data extraction stage, samples of various physiological parameters are extracted from the original data points. Visual processing is performed on the heart rate, respiration, electrocardiogram and blood oxygen waveform. Preferably, in one embodiment of the present invention, a clustering algorithm is used to classify these parameters. For example, a K-means clustering algorithm can be used to classify the heart rate, respiration, electrocardiogram and blood oxygen waveform as a dimension to obtain the corresponding characteristic curve.

[0036] In the data preprocessing stage, such as filtering and noise reduction, regression analysis is performed on the preprocessed data, parameters of heart rate, electrocardiogram, pulse and respiratory physiological signals are extracted, heart rate fluctuation index and heart rate variability analysis are performed, and ECG signal curves are generated. This step is crucial to improve the accuracy of subsequent analysis. Specifically, the following method is used for data preprocessing: First, the average heart rate of each data point 2 seconds before and 5 seconds after is calculated, and it is used as the background value of the current data point. Then, the absolute difference |HRp-mean|>|HRp-mean±5| is calculated, where |HRp-mean| refers to the absolute deviation between the data point and the background value. Finally, the results that meet this condition are counted into the snapshot.

[0037] In addition, the heart rate curve is also subjected to noise processing. Preferably, in one embodiment of the present invention, the corresponding waveform curve is subjected to windowing processing according to the sleep stage, and then its frequency component is analyzed with 0.2 Hz as the order of Fourier transform, so as to obtain the respiratory wave spectrum and the energy value of the respiratory wave. According to the frequency segmentation curve plotting, the heart rate characteristic curve can be obtained.

[0038] In the data analysis phase, regression analysis is performed on the preprocessed data curve. The regression equation and curve fitting goodness are calculated, and the correlation of the curves and the changes between the characteristic curves are determined based on the fitting goodness. In particular, the heart rate fluctuation index and heart rate variability analysis are performed.

[0039] In heart rate variability analysis, HR variability parameters are calculated from the heart rate signal. Specifically, take the entire ECG signal for 1 minute and calculate the following parameters:

[0040] Time domain analysis parameters of heart rate variability: number of ventricular premature contractions PN and maximum extent of ventricular premature contractions PSa.

[0041] Heart rate variability frequency domain analysis parameters: ultra-low frequency SD1, low frequency SDF, high frequency SD2 and VM.

[0042] Transition characteristics from low frequency domain to high frequency domain: HFnor = SD2 * HF.

[0043] These parameters can fully reflect the changing characteristics of heart rate and provide an important basis for evaluating sleep quality. In addition, parameters such as RR, P-QRS, PR, PP interval, RR interval, RR-RR interval, PR interval and RR interval are calculated from the QT interval signal. In particular, the variability of the RR interval is calculated as an objective indicator standard. The calculation formula of the RR interval variability index is as follows:

[0044]

[0045]

[0046]

[0047]

[0048]

[0049]

[0050]

[0051] in: is the RR interval sequence, is the standard deviation, is the first Segment value, The first arrive The standard deviation between segments, and is a natural number, is the mean of all RR intervals, .

[0052] Preferably, in one embodiment of the present invention, Deviation between time and average heart rate ( Time interval), The time interval ratio and Time and The specific calculation formula is as follows:

[0053] QT interval ratio ,in, for The actual measured value of the interval, For the corresponding Interval. Deviation of the QT interval from the mean heart rate ,in It can be calculated by Bazett formula: , is a constant, approximately 0.37 for men and 0.40 for women.

[0054]

[0055] Among them, Cov(QT,RR) is the covariance of QT and RR, and are the standard deviations of QT and RR, respectively.

[0056] The calculation of these parameters provides comprehensive data support for evaluating heart rate variability.

[0057] In the result evaluation and feedback step S4, the algorithm model is used to perform cluster analysis on the sleep parameters during the sleep period and generate a corresponding report. The clustered feature performance results are compared with the standard, the improvement of sleep characteristics and breathing characteristics is summarized, and feedback results are generated. The sleep data is analyzed using a regression model and a corresponding report is generated. According to the relevant performance results of the regression model, the characteristic curve of heart rate changes is analyzed.

[0058] In this step, the objective change results of the monitoring target will also be compared with the standard value. Specifically: when the calculated average RR interval value is <350ms, and the standard deviation of the RR interval is <50ms, the monitoring target is judged to have a normal sinoatrial node signal. When the average RR interval is >500ms and the standard deviation of the RR interval is >120ms, the monitoring target is judged to have an abnormal sinoatrial node signal. When the average RR interval result is basically stable within the normal value range, and the RR time series variability SD1 is <50ms, and SD2 is <75ms, the conduction function of the sinoatrial node efferent nerve is considered to be within the normal range. When the average RR interval value basically shows a downward trend, and SD1 has an increasing trend, and SD2 increases significantly, the sinoatrial node outflow function is considered to be weakened.

[0059] In addition, heart rate time domain and frequency domain analysis is performed through data extraction:

[0060] If VM>4 in the heart rate variability frequency domain analysis results, it means that the heart rate variability is large. VM<2 indicates that the heart rate variability is low. When the transition characteristics from low frequency to high frequency are HF>50%, the heart rate variability tends to improve. When the high frequency is>80%, the heart rate variability is small.

[0061] In order to show the evaluation results more intuitively, the test data is also visualized to generate characteristic curves of heart rate changes in each sleep period. These characteristic curves show the characteristics of heart rate oscillation, with the horizontal and vertical axes representing the average heart rate per minute, the heart rate oscillation amplitude, and the frequency value of the heart rate RR interval. By observing these curves, the following conclusions can be drawn:

[0062] When the horizontal axis is constant, the increase in heart rate amplitude and frequency proves that the heart rate variability increases, and the stronger the heart rate variability. When the frequency is constant, the increase in amplitude indicates that the degree of heart rate oscillation has increased, and vice versa, it indicates that the degree of heart rate oscillation has weakened. Finally, in the continuous monitoring and optimization step S5, the feedback results are fed back to the equipment manufacturer so that the sleep patch can be adjusted and optimized. This step ensures the closed-loop nature of the evaluation method and can continuously improve the effectiveness of the sleep patch product.

[0063] It is worth noting that the evaluation method of the present invention is not just a one-time process, but an iterative process; two or more rounds of data collection and data analysis are recommended. In the second and subsequent rounds of analysis, a more in-depth evaluation will be conducted.

[0064] The first round of data collection and data analysis process includes: step S1: using a sleep patch to perform a one-week sleep intervention treatment on the human body to obtain physiological signals and respiratory signals during sleep, and the physiological signals include heart rate, respiratory rate, and blood oxygen signals; step S2: converting the acquired physiological signals and respiratory signals during sleep into text files; step S3: using the relevant data functions of the R language to analyze the text files to obtain characteristic parameters that characterize sleep and respiratory characteristics; the specific data processing steps of the data analysis include: step S4: filtering the waveform after the characteristics; calculating the characteristic parameters based on the filtering results; the data processing executes electrocardiogram, heart rate, respiratory and blood oxygen waveforms Shape processing; ECG signal processing, the steps include: selecting ECG signal, filtering, normalizing, then extracting R wave, extracting corresponding waveform Q and T wave, calculating the deviation of QT time relative to the average heart rate, QT interval ratio and QT and RR correlation value, and performing feature analysis; heart rate, respiration, and blood oxygen signal processing, the steps include: heart rate signal denoising, respiration and blood oxygen windowing processing, calculating the mean of heart rate, respiration and blood oxygen, extracting RR interval, then calculating the mean of RR interval and RR variance, and obtaining characteristic values ​​after performing variability of QT interval, RR interval and RR time series; performing statistical analysis, algorithm model fit test and characteristic curve graphical analysis on the characteristic parameters.

[0065] Specifically, the second round of data collection and data analysis process includes the following steps:

[0066] Step S5: After analyzing according to the data formula and algorithm model in step 3, output feature data.

[0067] Step S6: using the average heart rate mean and standard deviation in the next day's heart rate signal characteristic data output from step S5 as parameters to determine the fluctuation of the sleep physiological signal.

[0068] Step S7: Using the output data of step S5 as parameters, determine the variability of the RR time series.

[0069] Step S8: Based on the data output from step S5, determine the variability of the RR interval.

[0070] Step S9: Determine the heart rate variability based on the output data of step S5.

[0071] Step S10: Performing graphical comparative analysis on the changes of each indicator in different sleep periods.

[0072] Preferably, in one embodiment of the present invention, an additional step is also included:

[0073] Step A1: Perform heart rate analysis on the heart rate characteristic curve of the next day output in S7 to obtain a heart rate variability characteristic value.

[0074] Step A2: After obtaining the index results output in step S6, the variability of the RR interval is calculated and compared with the output results in S5 to determine the curve changes of the heart rate and respiratory waveform.

[0075] The data output by S6 is compared with the data output by S5 in a graphical way to complete the RR interval variability analysis. These charts use time as the horizontal axis (in seconds) and the vertical axis as the heart rate fluctuation index / respiratory waveform index. The curve graph is divided into five sleep stages, namely Wake stage, N1 stage (i.e. light sleep stage), N2 stage (i.e. light sleep stage), N3 stage (i.e. deep sleep stage), and R stage (i.e. rapid eye movement stage), so as to display the changes in heart rate, respiratory waveform, and time series characteristics of heart rate.

[0076] When judging the fluctuation of heart rate, it is necessary to take into account the spontaneous cycle characteristics of the sleep cycle. Generally speaking, the cycle from the waking stage to the light sleep stage is about 4 times per hour, and then enters the medium to deep sleep stage, which lasts about 4 to 6 minutes, and then enters the rapid eye movement stage, which lasts for 10 to 20 minutes, and the rapid eye movement stage goes back and forth into the light sleep stage and starts the next light sleep stage. In contrast, the changes in heart rate are relatively constant.

[0077] From waking to deep sleep, the standard deviation of the RR interval is regressed against the variance and mean of the RR. If it shows a downward trend, it means that the variability of the RR interval from the REM period to the wakeful period is small; if it shows an upward trend from the slow wave to the REM period, the variability of the RR interval is small.

[0078] In addition, a regression analysis of the heart rate prediction curve was performed. This chart uses time as the horizontal axis (in seconds) and the spectrum as the vertical axis (in units) to fit the frequency distribution of the heart rate. Usually, the heart rate frequency distribution shows a peak value of 5 to 6 Hz.

[0079] By performing spectrum analysis on the heart rate curve, we can get the changes in heart rate. The characteristics of heart rate oscillation are presented, with the horizontal and vertical coordinates representing the average heart rate per minute, the heart rate oscillation amplitude, and the frequency value of the heart rate RR interval.

[0080] Finally, we focus on two parameters: maximum blood oxygen saturation and minute average blood oxygen saturation. These parameters are important for evaluating sleep quality and breathing status.

[0081] In this process, it is important to note that different groups of people have different sensitivities to sleep intervention products. Therefore, the evaluation indicators of sleep data for different groups of people should be calibrated before implementation. The calibration process includes the following steps:

[0082] Step S11: In the multi-target sleep intervention, the target heart rate, respiratory waveform and blood oxygen signal are collected. This step aims to obtain comprehensive physiological data to provide a basis for subsequent analysis.

[0083] Step S12: De-noise the signal collected in step S11 to extract the R wave peak information. Preferably, in one embodiment of the present invention, a wavelet transform method is used for de-noising, which can effectively retain important features of the signal while removing high-frequency noise.

[0084] Step S13: Gait detection is performed based on the information from step S12 to determine different sleep periods. Specifically, sleep can be divided into different stages such as light sleep, deep sleep, and rapid eye movement according to heart rate variability, changes in respiratory rate, and body movement.

[0085] Step S14: Calculate the mean and standard deviation of each sleep period according to the sleep detection information in step S13. These statistical data can help understand the physiological characteristics of different sleep stages.

[0086] Step S15: The heart rate standard deviation and RR interval fluctuation index in step S13 and step S14 can be used as evaluation indicators for judging the effect of heart rate intervention on different populations, and the indicators can be adjusted according to the fluctuation of the population.

[0087] It should be noted that different groups of people have different sensitivities to heart rate variability. For example, athletes and patients with depression have significant differences in their sensitivities to heart rate variability. Preferably, in one embodiment of the present invention, different evaluation criteria are set for different groups of people:

[0088] For the average adult, the standard assessment methods mentioned above are used.

[0089] For athletes, since their heart rate variability is usually higher, the assessment criteria may be appropriately raised. For example, the range of normal RR interval variability may be expanded to SD1 < 60ms and SD2 < 90ms.

[0090] For depressed patients, given that their heart rate variability is often lower, the criteria may be lowered accordingly. For example, the range of normal RR interval variability may be narrowed to SD1 < 40 ms and SD2 < 60 ms.

[0091] For the elderly, the standard will be adjusted appropriately to take into account the impact of age on heart rate variability. For example, the range of normal RR interval variability may be set as SD1 < 45ms and SD2 < 70ms.

[0092] This personalized evaluation method for different populations can greatly improve the accuracy and applicability of sleep patch product effectiveness evaluation.

[0093] In addition, in order to further improve the accuracy of the assessment, long-term monitoring is also recommended. Preferably, in one embodiment of the present invention, the same test target group is used as the research object, and the evaluation indicators for four consecutive weeks are compared, including various indicators of heart rate variability, electrocardiogram, respiration and blood oxygen signals, and the changes before and after are compared to determine the regulating effect of the sleep patch on the heart rate of a specific population.

[0094] Specifically, the following evaluation indicators can be set:

[0095] The degree of improvement in heart rate variability indicators: For example, if SD1 and SD2 show a gradual increasing trend over a four-week period, it can be considered that the sleep patch has a positive effect on improving heart rate variability.

[0096] Changes in electrocardiogram indicators: For example, if the QT interval gradually tends to the normal range, it can be considered that the sleep patch has a positive effect on improving the electrophysiological activity of the heart.

[0097] Improvement in respiratory indicators: If the respiratory rate tends to stabilize and the number of apnea events decreases, it can be considered that the sleep patch has a positive effect on improving sleep breathing.

[0098] Improvement in blood oxygen saturation: If the average blood oxygen saturation during the night increases and the fluctuation in blood oxygen saturation decreases, it can be considered that the sleep patch has a positive effect on improving nighttime oxygen supply.

[0099] Improvements in subjective sleep quality: Changes in subjective sleep quality can be assessed using standardized sleep quality questionnaires such as the Pittsburgh Sleep Quality Index (PSQI).

[0100] Through this long-term, multi-dimensional evaluation, we can have a more comprehensive understanding of the effectiveness of sleep patch products, evaluating not only their immediate effects, but also their cumulative effects over the long term.

[0101] In summary, the sleep patch product efficacy intelligent evaluation method provided by the present invention provides a comprehensive, objective and reliable method for the efficacy evaluation of sleep patch products through systematic data collection, refined data analysis and scientific result evaluation. This method not only takes into account the individual differences of different populations, but also incorporates the dimension of long-term monitoring, thereby greatly improving the accuracy and practicality of the evaluation.

[0102] We believe that with the widespread application of this evaluation method, the quality of sleep patch products will be significantly improved, thus making an important contribution to improving users' sleep quality and quality of life. At the same time, this method also provides a useful reference for the performance evaluation of other types of health monitoring products.

[0103] It should be noted that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for intelligently evaluating the efficacy of a sleeping patch product, characterized in that include: Equipment selection and preparation: select the appropriate target population, install and connect the corresponding equipment, collect physiological data, and calibrate the equipment; Data collection: a sleep intervention for one week is carried out on the target population through the sleep patch product, physiological signal data and sleep observation data are collected, and the data are preprocessed; data analysis: physiological signals during the sleep period are extracted, and characteristic parameters representing the characteristic values ​​of sleep and breathing are extracted and calculated; result evaluation and feedback: the characteristic parameters obtained in the data analysis step are clustered by using the algorithm model, wherein the characteristic parameters include heart rate variability and heart rate fluctuation index, and a corresponding report is generated. The improvement of sleep characteristics and breathing characteristics is summarized according to the clustered characteristic performance results and the standard, and feedback results are generated; continuous monitoring and optimization: the feedback results are fed back to the device to adjust and optimize the sleep patch; the specific steps of data analysis are as follows: step 1, data extraction, extracting various physiological parameters of the data sample from the data points, and using the clustering algorithm for classification to obtain the characteristic curve; step 2, data preprocessing, denoising the characteristic curve obtained in step 1; step 3, data analysis, regression analysis of the data curve after data preprocessing, calculating the regression equation and the curve fitting goodness, and determining the correlation of the data curve and the change between the characteristic curves according to the fitting goodness; In step 2, the data preprocessing step includes: dividing the data point context, calculating the average heart rate 2 seconds before and 5 seconds after the data point as the background value mean of the current data point, , the data point is included in the snapshot, where Refers to the absolute deviation between the data point and the background value; In the data analysis of step 3, heart rate fluctuation index and heart rate variability analysis are performed; parameters of heart rate, electrocardiogram, pulse and respiratory physiological signals are extracted from each sleep period; HR variability parameters are calculated from the heart rate signal: take the entire electrocardiogram of 1 minute, and the heart rate variability time domain analysis parameters are: number of premature ventricular contractions , maximum degree of premature ventricular contraction ; The frequency domain analysis parameters of heart rate variability are ultra-low frequency , low frequency ,high frequency , , the transition characteristics from low frequency domain to high frequency domain: 。 2. The evaluation method according to claim 1, characterized in that: RR, P-QRS, PR, PP intervals, RR intervals, RR-RR intervals, PR intervals, and RR intervals are calculated from the QT interval signal. The variability of the RR interval is calculated and used as an objective indicator standard. The calculation formula for the RR interval variability indicator is: ； ； ； 。 3. The evaluation method according to claim 1, characterized in that It also includes comparing the objective change results of the monitoring target with the standard value: when the calculated average RR interval value is <350ms and the standard deviation of the RR interval is <50ms, the monitoring target is judged to have a normal sinoatrial node signal; when the average RR interval is >500ms and the standard deviation of the RR interval is >120ms, the monitoring target is judged to have an abnormal sinoatrial node signal; when the average RR interval result is basically stable within the normal value range, and the RR time series variability SD1 is <50ms, SD2 is <75ms, indicating that the conduction function of the sinoatrial node efferent nerve is within the normal range; when the average RR interval value basically shows a downward trend, and SD1 has an increasing trend, and SD2 increases significantly, it indicates that the sinoatrial node efferent function is weakened; perform heart rate time domain and frequency domain analysis through data extraction: if VM>4 in the frequency domain analysis result of heart rate variability, it means that the heart rate variability is large; VM<2 indicates that the heart rate variability is low, and VM is the frequency domain analysis parameter of heart rate variability; when the transition characteristics from low frequency to high frequency are When the high frequency was >50%, the heart rate variability tended to improve, and when the high frequency was >80%, the heart rate variability was smaller.

4. The evaluation method according to claim 1, characterized in that It also includes generating a characteristic curve diagram of heart rate changes in each sleep period by visualizing the heart rate data collected by the monitoring device, the monitoring device including an electrocardiograph, and the characteristic curve diagram presents the characteristics of heart rate oscillations, with the horizontal and vertical axes representing the average heart rate per minute, the heart rate oscillation amplitude, and the frequency value of the heart rate RR interval, respectively. When the horizontal axis is constant, an increase in the heart rate amplitude and frequency proves that the heart rate variability increases, and the stronger the heart rate variability; when the frequency is constant, an increase in the amplitude indicates that the degree of heart rate oscillation increases, and vice versa, the degree of radial heart rate oscillation decreases.

5. The evaluation method according to claim 1, characterized in that Data collection and data analysis steps were performed in two or more rounds.

6. The evaluation method according to claim 5, characterized in that The first round of data collection and data analysis process includes: step S1: using a sleep patch to perform a one-week sleep intervention treatment on the human body to obtain physiological signals and respiratory signals during sleep, and the physiological signals include heart rate, respiratory rate, and blood oxygen signals; step S2: converting the acquired physiological signals and respiratory signals during sleep into text files; step S3: using the relevant data functions of the R language to analyze the text files to obtain characteristic parameters characterizing sleep and respiratory characteristics; the specific data processing steps of the first round of data analysis include: step S4: extracting features and filtering the original waveforms of the electrocardiogram, heart rate, respiration, and blood oxygen; step S5: calculating the characteristic parameters characterizing the electrocardiogram based on the filtering results; the data processing execution of the electrocardiogram, heart rate, respiration, and blood oxygen raw waveforms Processing of heart rate, respiration and blood oxygen waveforms; processing of ECG signals, the steps of which include selecting ECG signals, filtering, normalizing, then extracting R waves, extracting corresponding waveforms Q and T waves, calculating the deviation of QT time relative to the average heart rate, QT interval ratio and QT and RR correlation value as ECG signal characteristic parameters; processing of heart rate, respiration and blood oxygen signals, the steps of which include: denoising of heart rate signals, windowing of respiration and blood oxygen, calculating the mean of heart rate, respiration and blood oxygen, extracting RR interval, then calculating the mean of RR interval and RR variance, and obtaining characteristic values ​​after performing variability of QT interval, RR interval and RR time series; performing statistical analysis, algorithm model fit test and characteristic curve graphical analysis on the characteristic parameters obtained in step S3.

7. The evaluation method according to claim 6, characterized in that The second round of data collection and data analysis process also includes the following detailed steps: Step S6: output feature data according to the data analysis in Step 3; Step S7: use the heart rate mean and its standard deviation of the next day's heart rate signal in the output feature data obtained in Step S6 as parameters to judge the fluctuation of the sleep physiological signal; Step S8: use the output feature data obtained in Step S6 as parameters to judge the variability of the RR time series; Step S9: judge the variability of the RR interval based on the output feature data obtained in Step S6; Step S10: judge the heart rate variability based on the output feature data obtained in Step S6; and Step S11: perform graphical comparative analysis on the changes of each indicator in different sleep periods; Among them, in the step of using the sleep patch to intervene in the human body for one week, the sensitivity of the target population and different populations to the sleep patch is analyzed. The evaluation indicators of sleep data of different groups should be calibrated before execution, and the calibration includes the following steps: Step S12: In multi-target sleep intervention, target heart rate, respiratory waveform and blood oxygen signal are collected; Step S13: Noise reduction is performed on the collected signals to extract R wave peak information; Step S14: Gait detection is performed according to the R wave peak information to determine different periods of sleep; Step S15: According to the different periods of sleep determined in step S14, the mean and standard deviation of the heart rate are calculated for each period; Step S16: The RR interval fluctuation index and the heart rate standard deviation in step S15 are obtained through step S14 as evaluation indicators for judging the effect of heart rate intervention on different groups of people, and the indicators can be adjusted according to the fluctuation of the group; wherein the sensitivity of different groups of people to heart rate variability includes different sensitivity of athletes and depressive patients.

8. The evaluation method according to any one of claims 6 to 7, characterized in that It also includes a long-term evaluation of changes in heart rate variability trends. The steps are: using a group of subjects who receive sleep patch intervention, including any group of people in the general population, athletes, depressed patients or the elderly as research subjects, comparing the evaluation indicators for four consecutive weeks, including various indicators of heart rate variability, electrocardiogram, respiration and blood oxygen signals, comparing the changes before and after, and judging the regulatory effect of the sleep patch on the heart rate of a specific population.

Images