Sleep respiratory muscle state detection quantification and auxiliary adjustment method, system and device

Abstract

According to the sleep respiratory muscle state detection quantification and auxiliary adjustment method, system and device provided by the invention, sleep muscle state signals are collected and analyzed, sleep respiratory muscle state dynamic signals are extracted, and sleep respiratory muscle state event intensity is calculated and extracted in combination with sleep motion body position change characteristics; scientific detection quantification of process characteristics and state events of sleep breathing of the user is realized; by predicting and analyzing the occurrence trend of the sleep breathing muscle state event, a sleep breathing auxiliary adjustment strategy is generated and sent to the sleep breathing adjustment equipment in real time, and dynamic auxiliary adjustment of sleep breathing of the user is achieved; through an integrated framework of innovative evaluation and auxiliary adjustment of the sleep respiratory muscle state, sleep respiration of a user is assisted, and the sleep quality is improved. Scientific detection evaluation and efficient dynamic auxiliary adjustment of sleep breathing can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of sleep breathing detection, evaluation and auxiliary regulation, and in particular to a method, system and device for detecting, quantifying and auxiliary regulation of sleep respiratory muscle states. Background Art

[0002] Respiratory movement is the most basic behavior for human survival, and the driving force comes from the respiratory muscles; the respiratory muscles mainly include intercostal muscles, diaphragm, abdominal wall muscles, sternocleidomastoid muscles, back muscles, chest muscles, etc. Respiratory movement not only brings about changes in lung volume and pressure, but also causes changes in muscle behavior of the thorax, throat, mouth and nose. The physiological state and behavioral changes of respiratory muscles and respiratory-related link organs have very good respiratory dynamics. In addition, events or phenomena such as obstructive sleep apnea, hypopnea, snoring, etc. caused by narrowing and obstruction of the upper airway are accompanied by vibrations of multiple parts such as the oropharynx, soft palate, and mandible; while central sleep apnea is accompanied by abnormal chest and abdominal breathing behavior, and chest and abdominal breathing is also the main breathing mode during sleep; the above events or phenomena not only directly lead to frequent intermittent hypoxia at night, but also damage sleep structure and sleep quality, and indirectly have different degrees of impact on people's quality of life and work, physical health and life safety.

[0003] The prior art solution CN116013470A relates to a method, system and device for dynamically adjusting the level of active sleep behavior: a method for dynamically adjusting the level of active sleep behavior is provided, which collects and processes the physiological state signals and behavioral state signals of the user's sleep process in real time, generates real-time characteristics of physiological and behavioral states, identifies the sleep phase state in real time and generates a sleep phase curve; performs real-time behavioral ability quantitative analysis, baseline change analysis and extreme value harmonic analysis on the real-time characteristics of physiological and behavioral states, generates real-time indexes of muscle tension level, central motor ability and behavioral action level; performs real-time baseline change analysis and extreme value harmonic analysis based on the above indexes, generates a real-time curve of active sleep behavior level, predicts and generates a trend curve of active sleep behavior level, generates a dynamic adjustment strategy for active sleep behavior level in real time and adjusts it dynamically, generates a sleep behavior active level adjustment report and establishes a personalized behavior adjustment long-term database. The invention realizes the scientific detection, evaluation and training of the active or inhibited level of sleep behavior; however, the technical solution does not specifically involve the identification and extraction of muscle state dynamic characteristics, the identification and detection of sleep breathing events, the calculation and extraction of event intensity, and the regulation and control of sleep breathing.

[0004] In general, the detection, analysis and evaluation of sleep breathing in the existing technical solutions mainly rely on polysomnography (PSG), which identifies respiratory event detection and event types through traditional respiratory signal waveform analysis such as oral and nasal pressure, thermal sensitivity and chest and abdominal belt pressure, and further completes simple statistical analysis such as hypopnea index and number of respiratory events. The existing technical solutions lack systematic and scientific detection, analysis and evaluation of the process characteristics, state events and occurrence and development of the sleep respiratory muscle state, especially lacking a clear definition or characteristic quantification of the intensity of sleep respiratory events from the perspective of respiratory dynamics. In addition, in the existing technical solutions and existing application scenarios, sleep breathing control equipment is usually separated from sleep breathing detection equipment. Most sleep breathing control equipment can feedback control parameters online but still use offline preset program control to complete device feedback control, making it difficult to perform intelligent, accurate and personalized dynamic auxiliary adjustment according to the user's real-time status.

[0005] From the above, we can see that how to systematically and scientifically detect, quantify and evaluate the process characteristics, state events and occurrence and development of the sleep respiratory muscle state, and further optimize and improve the efficiency and effectiveness of existing sleep breathing control equipment, so as to assist users' sleep breathing and improve sleep quality, is a problem that needs to be further solved in current domestic and foreign product technical solutions and actual application scenarios. Summary of the invention

[0006] In view of the above defects and improvement needs of the existing methods, the purpose of the present invention is to provide a method for detecting, quantifying and assisting the regulation of the state of sleep respiratory muscles. By collecting and analyzing the sleep muscle state signals and extracting the sleep respiratory muscle state dynamics signals, the intensity of the sleep respiratory muscle state events is calculated and extracted in combination with the characteristics of the changes in sleep movement and body position, thereby realizing the scientific detection and quantification of the process characteristics and state events of the user's sleep breathing; by predicting and analyzing the occurrence trend of sleep respiratory muscle state events, a sleep breathing assistance regulation strategy is generated and sent to the sleep breathing regulation device in real time, thereby realizing the dynamic assistance regulation of the user's sleep breathing; through the integrated architecture of innovative evaluation and assistance regulation of the sleep respiratory muscle state, the user's sleep breathing is assisted and the sleep quality is improved. The present invention also provides a sleep respiratory muscle state detection, quantification and assistance regulation system for realizing the above method. The present invention also provides a sleep respiratory muscle state detection, quantification and assistance regulation device for realizing the above system.

[0007] According to the purpose of the present invention, the present invention proposes a method for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles, comprising the following steps:

[0008] Collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleeping muscle state signals and sleeping movement position signals respectively;

[0009] Performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal;

[0010] Performing event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state event, and further extracting event intensity in combination with the sleep movement position signal;

[0011] Performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain a sleep body position muscle state dynamics feature;

[0012] Performing signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predicting and identifying the sleep respiratory muscle state events and event occurrence trends;

[0013] According to the sleep respiratory muscle state events and the occurrence trends of the events, combined with the sleep respiratory knowledge base and the user sleep respiratory database, a sleep respiratory assistance adjustment strategy is generated and sent to the sleep respiratory adjustment device through a signal interface.

[0014] More preferably, the specific steps of collecting and processing the muscle state signal and limb movement signal of the user during sleep to generate the sleeping muscle state signal and the sleeping movement body position signal respectively further include:

[0015] Collecting and processing the muscle state signal of the user during sleep to generate the sleeping muscle state signal;

[0016] The limb movement signals of the user during sleep are collected and processed to generate the sleep movement position signals.

[0017] More preferably, the acquisition processing at least includes acquisition, analog-to-digital conversion, resampling, re-referencing, artifact removal, signal correction, noise reduction, power frequency notch, bandpass filtering, mean filtering, smoothing and signal time window segmentation; wherein, the signal time window segmentation is specifically to perform continuous time window segmentation on the target signal with a preset length time window and a preset translation time step to obtain a multi-time window signal set.

[0018] More preferably, the sleeping muscle state signal includes at least any one of an electromyographic signal and an electrical skin signal; the sleeping movement position signal includes at least any one of a limb acceleration signal, a video movement position detection signal and a radar movement position detection signal.

[0019] More preferably, the specific step of performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate the sleep respiratory muscle state dynamics signal also includes:

[0020] Decomposing the sleep muscle state signal to obtain a sleep muscle state description signal set;

[0021] Performing signal recognition on the sleep muscle state description signal set to obtain a sleep muscle state dynamics description signal set;

[0022] Signal fusion and / or envelope extraction are performed on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

[0023] More preferably, the signal decomposition is specifically to decompose the target signal into a set of component signals with multiple different modes, different sources and different characteristic attributes; the signal decomposition method includes at least time series decomposition, empirical mode decomposition, variational mode decomposition, local mean decomposition, wavelet transform, wavelet packet transform, time-frequency transform, blind source separation, linear discriminant analysis, detrending analysis, principal component analysis, independent component analysis, waveform analysis and numerical fitting, and their evolutionary variants.

[0024] More preferably, the signal recognition is specifically to identify and label the signal source of the target signal through a waveform determination method or a machine learning model; the signal source includes at least breathing, swallowing, clenching / grinding / chewing, head movement, and body movement.

[0025] More preferably, the signal fusion is specifically to perform synchronous average superposition or delayed average superposition on different component signals from the same signal source to fuse them into one source signal.

[0026] More preferably, the sleep respiratory muscle state dynamics signal at least includes an electromyography envelope signal, a skin electrode SCL level signal, and a skin electrode SCR level signal.

[0027] Preferably, the specific step of performing event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state event, and further extracting the event intensity in combination with the sleep movement position signal also includes:

[0028] Performing event detection and analysis on the sleep respiratory muscle state dynamics signal, extracting basic event information, and obtaining the sleep respiratory muscle state event;

[0029] The event intensity of the sleep respiratory muscle state event is calculated based on the sleep respiratory muscle state dynamics signal, the event basic information and the sleep movement body position signal.

[0030] More preferably, the event detection and analysis is specifically to perform waveform feature recognition and marker extraction on the sleep respiratory muscle state dynamics signal according to a preset sleep respiratory-muscle event knowledge base and / or machine learning model to obtain basic event information of the sleep respiratory muscle state event; the basic event information includes at least event type, start time, end time, duration, peak and valley values, and time at peak and valley values.

[0031] More preferably, the sleep respiratory muscle state event at least includes event type, start time, end time, duration, peak and valley values, time at peak and valley values, and event intensity.

[0032] More preferably, the event intensity is specifically determined by the duration, start time, end time, peak-to-valley values, and relative time between the peak-to-valley values ​​of the sleep respiratory muscle state event, as well as the characteristic amplitude change of the sleep respiratory muscle state.

[0033] More preferably, the method for calculating and generating the event intensity is specifically:

[0034] 1) obtaining the start time, end time, duration, peak-to-valley value and peak-to-valley value time of the sleep respiratory muscle state event, and presetting the sleep respiratory muscle state dynamics signal threshold;

[0035] 2) calculating the relative change of the peak-to-valley value of the sleep respiratory muscle state event and the preset sleep respiratory muscle state dynamics signal threshold to obtain the peak-to-valley relative value;

[0036] 3) calculating the linear slope according to the start time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley front slope;

[0037] 4) calculating the linear slope according to the end time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley trailing edge slope;

[0038] 5) extracting sleeping angle and sleeping direction information or preset sleeping position correction calculation strategy according to the sleeping movement position signal, and extracting sleeping position correction coefficient according to the preset sleeping position correction coefficient comparison table;

[0039] 6) extracting the correction coefficient of the dynamics signal source according to the signal source of the sleep respiratory muscle state dynamics and according to a preset dynamics signal source-correction coefficient comparison table;

[0040] 7) The event intensity is obtained according to the numerical fusion calculation of the duration of the sleep respiratory muscle state event, the peak-to-valley relative value, the peak-to-valley leading slope, the peak-to-valley trailing slope, the sleep position correction coefficient, and the dynamic signal source correction coefficient.

[0041] More preferably, the specific step of performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature also includes:

[0042] Performing characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristics;

[0043] Performing feature analysis on the sleep movement position signal to obtain the sleep movement position feature;

[0044] The sleep respiratory muscle state characteristics and the sleep movement body position characteristics are aggregated to obtain the sleep body position muscle state dynamic characteristics.

[0045] More preferably, the feature analysis includes at least numerical feature analysis, envelope feature analysis, time-frequency feature analysis, and nonlinear feature analysis; wherein the numerical features include at least mean value, root mean square, maximum value, minimum value, variance, standard deviation, coefficient of variation, kurtosis, and skewness; the time-frequency features include at least frequency band power, frequency band power ratio, and frequency band center frequency; and the nonlinear features include at least entropy features, fractal features, and complexity features.

[0046] More preferably, the sleep position muscle state dynamic characteristics at least include the sleep respiratory muscle state characteristics and the sleep movement position characteristics; the sleep respiratory muscle state characteristics at least include any one of respiratory numerical characteristics, electromyographic characteristics, and skin electricity characteristics; the sleep movement position characteristics at least include limb movement characteristics and sleeping posture characteristics.

[0047] More preferably, the specific steps of performing signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predicting and identifying the sleep respiratory muscle state events and event occurrence trends also include:

[0048] Performing signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state dynamics prediction signal;

[0049] Performing signal trend analysis on the sleep movement position signal to obtain a sleep movement position prediction signal;

[0050] Performing signal trend analysis on the sleep postural muscle state dynamic characteristics to obtain sleep postural muscle state dynamic prediction characteristics;

[0051] Performing event detection analysis on the sleep respiratory muscle state dynamics prediction signal, combining the sleep movement body position prediction signal, predicting and identifying the sleep respiratory muscle state event and the event intensity;

[0052] According to the sleep respiratory muscle state event and the event intensity, combined with the sleep postural muscle state dynamic prediction characteristics, the event occurrence trend is extracted and generated.

[0053] More preferably, the signal trend analysis includes at least any one of exponential smoothing, Holt-Winters method, AR, MA, ARMA, ARIMA, SARIMA, SARIMAX, VAR, VARMA, VARMAX, and machine learning.

[0054] More preferably, the event occurrence trend includes at least the event occurrence type, event occurrence probability and event occurrence intensity; the event occurrence trend is obtained by machine learning through learning and training, data modeling and analytical calculation of large-scale sleeping user sample data such as the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamic characteristics related to the sleep respiratory muscle state event.

[0055] Preferably, the specific steps of generating a sleep breathing assistance adjustment strategy based on the sleep breathing muscle state event and the event occurrence trend, combining a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to the sleep breathing adjustment device through a signal interface also include:

[0056] Generate the sleep breathing assistance adjustment strategy according to the sleep breathing muscle state event and the event occurrence trend, combined with the sleep breathing knowledge base and the user sleep breathing database;

[0057] Sending the sleep breathing assistance adjustment strategy to the sleep breathing adjustment device through the signal interface, optimizing the control execution of the sleep breathing adjustment device, so as to achieve dynamic assistance adjustment of the user's sleep breathing;

[0058] According to a preset cycle or strategy, a sleep respiratory muscle state detection and auxiliary adjustment report is generated and output, and the user sleep respiratory database is updated.

[0059] Preferably, the sleep breathing knowledge base mainly comes from the knowledge and experience of health management and clinical medicine related to sleep breathing, and at least includes the laws of sleep breathing-muscle movement, characteristics of common sleep breathing events, commonly used sleep breathing adjustment methods and scenario intervention parameter guidance; the user sleep breathing database at least includes the sleep respiratory muscle state events, the sleep respiratory muscle state dynamic signals, the sleep movement position signals, the sleep position muscle state dynamic characteristics, the event occurrence trends and the sleep breathing assistance adjustment strategies.

[0060] More preferably, the sleep breathing assistance adjustment strategy at least includes a target value for breathing frequency, a target value for breathing depth, an adjustment method, an adjustment timing, a duration, and device control parameters.

[0061] More preferably, the sleep breathing regulation device includes at least any one of a ventilator, a smart mattress, an odor stimulation device, an electrical stimulation device, a tactile stimulation device, an environmental temperature and humidity control device, and an environmental CO2 concentration control device.

[0062] More preferably, the sleep respiratory muscle state detection and auxiliary regulation report at least includes a statistical analysis of the sleep respiratory muscle state events, a sleep respiratory muscle state detection evaluation summary, an auxiliary regulation evaluation summary, and sleep respiratory behavior improvement suggestions.

[0063] According to the purpose of the present invention, the present invention proposes a sleep respiratory muscle state detection quantification and auxiliary regulation system, including the following modules:

[0064] The sleep behavior detection module is used to collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleep muscle state signals and sleep movement body position signals respectively;

[0065] A muscle dynamics analysis module, used for performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal;

[0066] An event detection and analysis module, used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain sleep respiratory muscle state events, and further extract event intensity in combination with the sleep movement position signal;

[0067] A state feature analysis module, used for performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature;

[0068] An event trend prediction module is used to perform signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predict and identify the sleep respiratory muscle state events and event occurrence trends;

[0069] A strategy auxiliary adjustment module, for generating a sleep breathing auxiliary adjustment strategy based on the sleep respiratory muscle state event and the event occurrence trend, in combination with a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to a sleep breathing adjustment device through a signal interface;

[0070] The data operation management module is used for visual management, unified storage and operation management of all process data of the system.

[0071] Preferably, the sleep behavior detection module further includes the following functional units:

[0072] A muscle state detection unit, used to collect and process the muscle state signal of the user during sleep and generate the sleeping muscle state signal;

[0073] The motion state detection unit is used to collect and process the limb motion signals of the user during sleep and generate the sleep motion position signals.

[0074] More preferably, the muscle dynamics analysis module further includes the following functional units:

[0075] A signal decomposition unit, used for performing signal decomposition on the sleep muscle state signal to obtain a sleep muscle state description signal set;

[0076] A signal recognition unit, configured to perform signal recognition on the sleep muscle state description signal set to obtain a sleep muscle state dynamics description signal set;

[0077] The fusion and extraction unit is used to perform signal fusion and / or envelope extraction on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

[0078] Preferably, the event detection and analysis module further includes the following functional units:

[0079] An event detection and analysis unit, used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal, extract basic event information, and obtain the sleep respiratory muscle state event;

[0080] The event intensity calculation unit is used to calculate the event intensity of the sleep respiratory muscle state event according to the sleep respiratory muscle state dynamics signal, the event basic information and the sleep movement position signal.

[0081] Preferably, the state feature analysis module further includes the following functional units:

[0082] A muscle state characteristic analysis unit, used for performing characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristic;

[0083] A motion body position characteristic analysis unit, used for performing characteristic analysis on the sleep motion body position signal to obtain the sleep motion body position characteristic;

[0084] The state feature collection unit is used to collect the sleep respiratory muscle state features and the sleep movement posture features to obtain the sleep posture muscle state dynamic features.

[0085] Preferably, the event trend prediction module further includes the following functional units:

[0086] A muscle state prediction unit, used for performing a signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state dynamics prediction signal;

[0087] A movement position prediction unit, used for performing signal trend analysis on the sleep movement position signal to obtain a sleep movement position prediction signal;

[0088] A dynamic feature prediction unit, used for performing signal trend analysis on the dynamic feature of the sleeping position muscle state to obtain the dynamic feature prediction of the sleeping position muscle state;

[0089] An event detection and extraction unit, configured to perform event detection and analysis on the sleep respiratory muscle state dynamics prediction signal, and predict and identify the sleep respiratory muscle state event and the event intensity in combination with the sleep movement body position prediction signal;

[0090] The event trend recognition unit is used to extract and generate the event occurrence trend according to the sleep respiratory muscle state event and the event intensity in combination with the sleep postural muscle state dynamic prediction feature.

[0091] Preferably, the strategy auxiliary adjustment module further includes the following functional units:

[0092] An auxiliary strategy generating unit, configured to generate the sleep breathing auxiliary adjustment strategy according to the sleep breathing muscle state event and the event occurrence trend, in combination with a sleep breathing knowledge base and a user sleep breathing database;

[0093] A strategy auxiliary control unit, used to send the sleep breathing auxiliary adjustment strategy to the sleep breathing adjustment device through a signal interface, optimize the control execution of the sleep breathing adjustment device, so as to achieve dynamic auxiliary adjustment of the user's sleep breathing;

[0094] The report data management unit is used to generate and output a sleep respiratory muscle state detection and auxiliary adjustment report according to a preset cycle or strategy, and update the user sleep respiratory database.

[0095] Preferably, the data operation management module further includes the following functional units:

[0096] User information management unit, used for registration, input, editing, query, output and deletion of user information;

[0097] A data visualization management unit, used for visual display and management of all data in the system;

[0098] The data operation management unit is used to store, back up, migrate and export all data in the system.

[0099] According to the purpose of the present invention, the present invention proposes a sleep respiratory muscle state detection quantification and auxiliary regulation device, including the following modules:

[0100] The sleep behavior detection module is used to collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleep muscle state signals and sleep movement position signals respectively;

[0101] A muscle dynamics analysis module, used for performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal;

[0102] An event detection and analysis module is used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain sleep respiratory muscle state events, and further extract event intensity in combination with the sleep movement position signal;

[0103] A state feature analysis module, used for performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature;

[0104] An event trend prediction module is used to perform signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predict and identify the sleep respiratory muscle state events and event occurrence trends;

[0105] A strategy-assisted adjustment module, for generating a sleep breathing assistance adjustment strategy according to the sleep breathing muscle state event and the occurrence trend of the event, in combination with a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to a sleep breathing adjustment device through a signal interface;

[0106] The data operation management module is used for visual management, unified storage and operation management of all process data of the system.

[0107] The present invention provides a method, system and device for detecting, quantifying and assisting the regulation of sleep respiratory muscle states. The method collects and analyzes sleep muscle state signals and extracts sleep respiratory muscle state dynamic signals, and calculates and extracts the intensity of sleep respiratory muscle state events in combination with the characteristics of changes in sleep movement and body position. The method generates and sends a sleep respiratory assistance regulation strategy through predictive analysis of the occurrence trend of sleep respiratory muscle state events, and further optimizes and improves the efficiency and effectiveness of sleep respiratory regulation equipment. Finally, through an integrated architecture of innovative sleep respiratory evaluation and assistance regulation, the method assists users in their sleep breathing and improves their sleep quality.

[0108] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0110] Figure 1 It is a schematic diagram of the process steps of a method for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles provided by an embodiment of the present invention;

[0111] Figure 2 It is a schematic diagram of the module composition of a sleep respiratory muscle state detection, quantification and auxiliary regulation system provided by an embodiment of the present invention;

[0112] Figure 3 The present invention is a schematic diagram of a module structure of a device for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0113] In order to more clearly illustrate the purpose and technical solution of the present invention, the present invention will be further described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, rather than all of the embodiments. Without creative work, other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention should all belong to the protection scope of the present invention. It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily.

[0114] The applicant observed and analyzed that breathing behavior is performed by muscle movements and nerve conduction of a series of respiratory physiological organs; the laws and dynamic characteristics of breathing behavior can be identified and extracted from these muscle movements and nerve conduction activities, especially the state of respiratory-related muscles; the muscle nerve behavior activities in specific parts can accurately and timely describe the respiratory dynamic characteristics, and can achieve scientific and comprehensive evaluation and dynamic auxiliary regulation of sleep breathing behavior. The first aspect: First, sleep is a physiological process dominated by basic life functions and a sharp decrease in limb behavior movements, and sleep breathing is one of the most important physiological processes; secondly, the complete respiratory system structure or physiological organ chain is the mouth, nose, pharynx, larynx, main trachea, bronchi and lungs in sequence, and the latter half occurs in the chest and abdomen, and the breathing behavior is completed by the muscle nerve system of these physiological organs; therefore, the dynamic signals of respiratory behavior can be restored from the muscle nerve movement of these physiological organs, especially in the sleep state with low movement interference. On the second aspect, the muscle nerve movement of the respiratory physiological organs is the most direct cause of the formation of respiratory events and the most accurate expression of the formation of respiratory events.

[0115] Therefore, the present invention uses sleep muscle state signals and sleep movement position signals as the basis to achieve scientific and comprehensive detection and quantitative evaluation of sleep respiratory muscle state events, and further completes the real-time generation of sleep breathing regulation strategies through the prediction of sleep respiratory muscle state events and sends them to the sleep breathing regulation equipment in real time, thereby realizing the integration of scientific detection and evaluation of the user's sleep breathing and dynamic auxiliary regulation, and efficiently assisting the user's sleep breathing.

[0116] Combination Figure 1 As shown, a method for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles provided by an embodiment of the present invention comprises the following steps:

[0117] P100: Collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleeping muscle state signals and sleeping movement position signals respectively.

[0118] In this embodiment, the acquisition processing at least includes acquisition, analog-to-digital conversion, resampling, re-referencing, artifact removal, signal correction, noise reduction, power frequency notch, bandpass filtering, mean filtering, smoothing processing and signal time window segmentation; wherein, the signal time window segmentation is specifically to perform continuous time window segmentation on the target signal with a preset length time window and a preset translation time step to obtain a multi-time window signal set.

[0119] In this embodiment, the sleeping muscle state signal includes at least any one of an electromyographic signal and an electrical skin signal; the sleeping movement position signal includes at least any one of a limb acceleration signal, a video movement position detection signal and a radar movement position detection signal.

[0120] The first step is to collect and process the muscle state signals of the user during sleep to generate sleeping muscle state signals.

[0121] In this embodiment, the specific implementation method of the present invention is described by taking the sleeping muscle state signal as an example.

[0122] In this embodiment, myoelectric signals are selected as muscle state signals to illustrate the specific implementation method of the present invention. Two groups use myoelectric sensors with a sampling rate of 2000 Hz: the first group, oral pharynx and throat, four channels, the collection electrodes are located at the left and right mandibles and the left and right pharynx respectively, and the reference electrode is located at the center of the chin; the second group, chest and abdomen, two channels, right chest, horizontally inclined at 45°, the reference electrode is in the middle, and the collection electrodes are on the left and right sides.

[0123] The unified processing of electromyographic signals is artifact removal, signal correction, wavelet noise reduction, and 50Hz and double-frequency power frequency notching and 30-500Hz bandpass filtering are completed through Hamming window and zero-phase FIR digital filter, and multi-signal time window segmentation is performed with 2 seconds as the time window and 1 second as the time step; finally, the sleeping muscle state signal is obtained.

[0124] The second step is to collect and process the user's limb movement signals during sleep to generate sleep movement position signals.

[0125] In this embodiment, the specific implementation method of the present invention is described by taking the sleep movement position signal as an example.

[0126] In this embodiment, the acceleration signal of the forehead center Fpz and the upper left chest is collected by a wireless six-axis gyroscope sensor patch device with a sampling rate of 64 Hz to generate a sleep movement position signal.

[0127] The signal processing of the acceleration signal is resampling (2000Hz), removing artifacts, signal correction, and multi-signal time window segmentation with 2 seconds as the time window and 1 second as the time step; finally, the sleep movement position signal is obtained.

[0128] P200: performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal.

[0129] In this embodiment, signal decomposition specifically decomposes the target signal into a set of component signals with multiple different modes, different sources and different characteristic attributes; the signal decomposition method includes at least time series decomposition, empirical mode decomposition, variational mode decomposition, local mean decomposition, wavelet transform, wavelet packet transform, time-frequency transform, blind source separation, linear discriminant analysis, detrending analysis, principal component analysis, independent component analysis, waveform analysis and numerical fitting, and their evolutionary variants.

[0130] In this embodiment, signal recognition specifically identifies and labels the signal source of the target signal through a waveform determination method or a machine learning model; the signal source includes at least breathing, swallowing, clenching / grinding / chewing, head movement, and body movement.

[0131] In this embodiment, signal fusion specifically involves synchronously averaging or delaying averaging different component signals from the same signal source to fuse them into one source signal.

[0132] In this embodiment, the sleep respiratory muscle state dynamics signal at least includes an electromyography envelope signal, a skin electrical SCL level signal, and a skin electrical SCR level signal.

[0133] The first step is to decompose the sleep muscle state signal to obtain a set of sleep muscle state description signals.

[0134] In this embodiment, a variational mode decomposition method is selected to perform signal decomposition on each signal in the sleeping muscle state signal.

[0135] In actual applications, the combined use of methods such as empirical mode decomposition (EMD), variational mode decomposition (VMD), blind source separation (BSS), independent component analysis (ICA), principal component analysis (PCA) and their evolutionary variants can meet the signal decomposition needs of different user scenarios.

[0136] The second step is to perform signal recognition on the sleep muscle state description signal set to obtain the sleep muscle state dynamics description signal set.

[0137] In this embodiment, first, the scaled electromyographic data of breathing, swallowing, clenching / grinding / chewing, head movement, and body movement are learned and modeled by machine learning to build an electromyographic source identification model. Then, by pre-building a machine learning model-electromyographic source identification model, the signal components of different sources such as breathing, swallowing, clenching / grinding / chewing, head movement, and body movement are identified from the decomposed signal set of each source signal in the sleep muscle state description signal set. Finally, the signal marked with "breathing" is screened to obtain the sleep muscle state dynamics description signal set.

[0138] The third step is to perform signal fusion and / or envelope extraction on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

[0139] In this embodiment, first, the signals of the "breathing" mark of each source signal in the signal set describing the dynamics of the sleep muscle state are fused through signal fusion to obtain the pure breathing signal of each source signal; then, the envelope signal of these pure breathing signals is extracted by Hilbert transform to obtain the breathing envelope signal of each source signal; thirdly, one signal is selected from the signal sources of the oropharynx-four channels and the chest and abdomen-two channels, and one respiratory dynamics signal is obtained through signal fusion (linear addition) to obtain the sleep respiratory muscle state dynamics signal.

[0140] In actual application scenarios, the oropharyngeal-muscle state signal is more likely to detect and describe events or phenomena such as obstructive sleep apnea, hypopnea, and snoring; chest and abdominal breathing is also the main breathing pattern during sleep, and central sleep apnea is more likely to be detected and fully described from the chest and abdomen-muscle state signal and chest and abdomen breathing behavior abnormalities. The fusion analysis of the oropharyngeal-muscle state signal and the chest and abdomen-muscle state signal can more comprehensively describe the sleep breathing behavior process and various sleep breathing events.

[0141] P300: performing event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state event, and further extracting event intensity in combination with the sleep movement position signal.

[0142] The first step is to perform event detection and analysis on the sleep respiratory muscle state dynamics signal, extract the basic event information, and obtain the sleep respiratory muscle state event.

[0143] In this embodiment, the event detection and analysis specifically performs waveform feature recognition and marker extraction on the sleep respiratory muscle state dynamics signal according to a preset sleep respiratory-muscle event knowledge base and / or a machine learning model to obtain the basic event information of the sleep respiratory muscle state event; the basic event information includes at least event type, start time, end time, duration, peak and valley values, and peak and valley time.

[0144] In this embodiment, the sleep respiratory muscle state event at least includes event type, start time, end time, duration, peak and valley values, peak and valley value time, and event intensity.

[0145] In this embodiment, the sleep respiratory muscle state dynamics signal and the sleep respiratory muscle state event scale sample data are trained and modeled by machine learning, and a sleep respiratory muscle state event recognition model is pre-built. In the application process, the sleep respiratory muscle state dynamics signal is subjected to event detection analysis and information annotation by the sleep respiratory muscle state event recognition model, so that the basic event information of the sleep respiratory muscle state event can be easily extracted.

[0146] The second step is to calculate the event intensity of the sleep respiratory muscle state event based on the sleep respiratory muscle state dynamics signal, the basic event information and the sleep movement position signal.

[0147] In this embodiment, the event intensity is specifically determined by the duration, start time, end time, peak and valley values, and relative time between the peak and valley values ​​of the sleep respiratory muscle state event, as well as the characteristic amplitude change of the sleep respiratory muscle state.

[0148] In this embodiment, the calculation and generation method of event intensity is specifically as follows:

[0149] 1) Obtaining the start time, end time, duration, peak-to-valley value, and peak-to-valley value time of the sleep respiratory muscle state event, as well as the preset sleep respiratory muscle state dynamic signal threshold;

[0150] 2) calculating the relative change of the peak-to-valley value of the sleep respiratory muscle state event and the preset sleep respiratory muscle state dynamic signal threshold to obtain the peak-to-valley relative value;

[0151] 3) Calculate the linear slope according to the start time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley front slope;

[0152] 4) Calculate the linear slope according to the end time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley trailing edge slope;

[0153] 5) Extracting the sleeping angle and sleeping direction information according to the sleeping movement position signal, and extracting the sleeping position correction coefficient according to a preset sleeping position correction coefficient comparison table or a preset sleeping position correction calculation strategy;

[0154] 6) according to the signal source of the sleep respiratory muscle state dynamics, according to the preset dynamic signal source-correction coefficient comparison table, extract the dynamic signal source correction coefficient;

[0155] 7) The event intensity is obtained by numerically fusion calculation based on the duration of the sleep respiratory muscle state event, the relative value of the peak and valley, the leading slope of the peak and valley, the trailing slope of the peak and valley, the sleep position correction coefficient, and the correction coefficient of the dynamic signal source.

[0156] In this embodiment, the preset sleep respiratory muscle state dynamics signal threshold is the mean of the sleep respiratory muscle state dynamics signal (electromyographic envelope signal) in the non-sleep respiratory muscle state event interval. In actual use scenarios, the mean of the user's electromyographic envelope signal in a resting state can also be used as the preset sleep respiratory muscle state dynamics signal threshold. In this embodiment, 20%, 50%, 80%, 100%, 150%, 200%, 300%, 500%, 800%, and 1000% exceeding the preset sleep respiratory muscle state dynamics signal threshold are used as grading intervals (preset sleep respiratory muscle state event intensity grading rules) to determine the event level or signal threshold selection.

[0157] In this embodiment, a preset sleeping position correction calculation strategy is used: the coefficient of the supine 0° angle is 0.5, and the coefficient of the prone 180° angle is 1.0 for linear angle calculation; the coefficient of the left side lying is a constant of 0.95, the coefficient of the right side lying is a constant of 0.90, the coefficient of the supine is a constant of 0.85, and the coefficient of the prone is a constant of 1.0. By obtaining the user's different sleeping posture directions and sleeping posture angles, the sleeping position correction coefficient is directly converted and calculated. In actual application scenarios, different sleeping positions (directions and angles) have a very large impact on sleep breathing behavior, directly affecting the local muscle state, especially the respiratory muscles. For example, obstructive sleep apnea is very easy to cause in a supine lying posture, which can be alleviated by sleeping on the side, but sleeping on the side will bring compression to the chest or heart.

[0158] In this embodiment, the preset dynamic signal source-correction coefficient comparison table has a comparison relationship of: oropharyngeal signal -0.80, thoracic and abdominal signal -0.90, oropharyngeal thoracic and abdominal combined signal -1.00.

[0159] In this embodiment, two widely used event intensity calculation formulas are provided to illustrate the construction method of numerical fusion calculation. In the actual application process, the numerical combination calculation method of event intensity is selected according to the actual situation of different user scenarios.

[0160] The first calculation formula method of event intensity is as follows:

[0161]

[0162] The second calculation formula method of event intensity is as follows:

[0163]

[0164] Among them, eti is the event intensity, K sig is the correction factor of the kinetic signal source, K posis the sleep position correction coefficient, Td is the duration of the sleep respiratory muscle state event, Pv is the peak-to-valley relative value, LSp is the peak-to-valley leading edge slope, RSp is the peak-to-valley trailing edge slope, || || is the absolute value operation, and max() is the maximum value operation.

[0165] P400: Performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain a sleep body position muscle state dynamics feature.

[0166] In this embodiment, the feature analysis includes at least numerical feature analysis, envelope feature analysis, time-frequency feature analysis, and nonlinear feature analysis; wherein the numerical features include at least mean value, root mean square, maximum value, minimum value, variance, standard deviation, coefficient of variation, kurtosis, and skewness; the time-frequency features include at least frequency band power, frequency band power ratio, and frequency band center frequency; and the nonlinear features include at least entropy features, fractal features, and complexity features.

[0167] In this embodiment, the dynamic characteristics of the sleep position muscle state include at least the sleep respiratory muscle state characteristics and the sleep movement position characteristics; the sleep respiratory muscle state characteristics include at least any one of the respiratory numerical characteristics, electromyographic characteristics, and skin electricity characteristics; the sleep movement position characteristics include at least the limb movement characteristics and the sleeping posture characteristics.

[0168] The first step is to perform characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristics.

[0169] In this embodiment, electromyographic features are used as sleep respiratory muscle state features, including the envelope root mean square, center frequency and SVD entropy of the sleep respiratory muscle state dynamics signal, and the respiratory frequency of the respiratory numerical feature.

[0170] The second step is to perform feature analysis on the sleep movement position signal to obtain the sleep movement position features.

[0171] In this embodiment, sleeping posture characteristics are used as sleeping movement posture characteristics, which at least include sleeping direction characteristics and sleeping angle characteristics.

[0172] The third step is to aggregate the sleep respiratory muscle state characteristics and sleep movement position characteristics to obtain the sleep position muscle state dynamic characteristics.

[0173] In this embodiment, the sleep postural muscle state dynamic characteristics include the envelope root mean square, center frequency and SVD entropy of the sleep respiratory muscle state dynamic signal, as well as the sleeping posture direction characteristics and the sleeping posture angle characteristics.

[0174] P500: performing signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predicting and identifying the sleep respiratory muscle state events and event occurrence trends.

[0175] In this embodiment, the signal trend analysis includes at least any one of exponential smoothing method, Holt-Winters method, AR, MA, ARMA, ARIMA, SARIMA, SARIMAX, VAR, VARMA, VARMAX, and machine learning.

[0176] The first step is to perform signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state dynamics prediction signal.

[0177] In this embodiment, the AR method is used to perform signal trend analysis on the sleep respiratory muscle state dynamics signal (the fused respiratory signal of the oropharynx and the chest and abdomen) to predict the development and changes of the user's sleep respiratory behavior state.

[0178] The second step is to perform signal trend analysis on the sleep movement position signal to obtain a sleep movement position prediction signal.

[0179] In this embodiment, the AR method is used to perform signal trend analysis on the sleep movement position signal (acceleration signal, including sleeping direction signal and sleeping angle signal) to predict the development and change of the user's sleeping position state.

[0180] The third step is to perform signal trend analysis on the dynamic characteristics of the sleeping postural muscle state to obtain the dynamic prediction characteristics of the sleeping postural muscle state.

[0181] In this embodiment, the AR method is used to perform signal trend analysis on the dynamic characteristics of the sleep postural muscle state (the envelope root mean square, center frequency and SVD entropy of the sleep respiratory muscle state dynamic signal, breathing frequency, and sleeping direction characteristics and sleeping angle characteristics) to predict the development and changes of the user's sleep respiratory state characteristics.

[0182] The fourth step is to perform event detection and analysis on the sleep respiratory muscle state dynamics prediction signal, and combine it with the sleep movement position prediction signal to predict and identify the sleep respiratory muscle state events and event intensity.

[0183] In this embodiment, according to the aforementioned method steps, the sleep respiratory muscle state dynamics prediction signal is firstly subjected to event detection and analysis to predict the basic event information of the "next" sleep respiratory muscle state event; and then combined with the sleep movement position prediction signal, the "next" event intensity is calculated.

[0184] Step 5: According to the sleep respiratory muscle state events and event intensity, combined with the sleep postural muscle state dynamic prediction characteristics, the trend of event occurrence is extracted and generated.

[0185] In this embodiment, the event occurrence trend includes at least the event occurrence type, event occurrence probability and event occurrence intensity; the event occurrence trend is obtained by machine learning through learning and training, data modeling and analysis and calculation of large-scale sleeping user sample data such as sleep respiratory muscle state dynamic signals, sleep movement position signals and sleep position muscle state dynamic characteristics related to sleep respiratory muscle state events.

[0186] In this embodiment, machine learning technology is used to implement the event occurrence trend model. The event occurrence trend recognition model obtained by pre-learning and training can be continuously updated through the accumulation of data from different users to improve the recognition, calculation or decision-making capabilities of the model.

[0187] P600: Generate a sleep breathing assistance adjustment strategy based on the sleep breathing muscle state event and the event occurrence trend, combined with a sleep breathing knowledge base and a user sleep breathing database, and send it to a sleep breathing adjustment device through a signal interface.

[0188] In this embodiment, the sleep breathing knowledge base mainly comes from the knowledge and experience of health management and clinical medicine related to sleep breathing, and at least includes sleep breathing-muscle movement rules, common sleep breathing event characteristics, common sleep breathing adjustment methods and scene intervention parameter guidance. The user sleep breathing database at least includes sleep respiratory muscle state events, sleep respiratory muscle state dynamic signals, sleep movement position signals, sleep position muscle state dynamic characteristics, event occurrence trends and sleep breathing assistance adjustment strategies.

[0189] In this embodiment, the sleep breathing assistance adjustment strategy at least includes a target value for breathing frequency, a target value for breathing depth, an adjustment method, an adjustment time point, a duration, and device control parameters.

[0190] The first step is to generate a sleep breathing assistance adjustment strategy based on sleep respiratory muscle state events and event occurrence trends, combined with a sleep breathing knowledge base and a user sleep breathing database.

[0191] In this embodiment, the respiratory frequency and the target value of the respiratory frequency are taken as the main targets of the sleep breathing assistance adjustment strategy as an example. The current respiratory frequency and the predicted respiratory frequency value can be obtained from the dynamic characteristics of the sleep postural muscle state and the dynamic prediction characteristics of the sleep postural muscle state, respectively. The target value of the respiratory frequency can be obtained through the sleep breathing knowledge base or expert system. It is easy to obtain the numerical path planning from the respiratory frequency, the predicted respiratory frequency value to the target value of the respiratory frequency through the preset path planning strategy or machine learning model. In addition, with the help of the preset sleep breathing knowledge base and the pre-built composite expert system, the sleep respiratory muscle state events and event occurrence trends, and the target value of the respiratory frequency are input into the composite expert system, and the composite expert system can output parameters such as the adjustment method, adjustment time point, duration, and device control parameters. The composite expert system can be a composite knowledge decision system of traditional knowledge retrieval and application systems and machine learning models.

[0192] The second step is to send the sleep breathing assistance adjustment strategy to the sleep breathing adjustment device through the signal interface, optimize the control execution of the sleep breathing adjustment device, so as to realize the dynamic assistance adjustment of the user's sleep breathing.

[0193] In this embodiment, the sleep breathing adjustment device includes at least any one of a ventilator, a smart mattress, an odor stimulation device, an electrical stimulation device, a tactile stimulation device, an environmental temperature and humidity control device, and an environmental CO2 concentration control device.

[0194] In this embodiment, taking people with obstructive sleep apnea or hypopnea as an example, the sleep breathing adjustment equipment and means mainly directly improve the respiratory pressure / power cycle through the intelligent ventilator, adjust the sleeping position through the intelligent mattress, and adjust the muscle state of the respiratory muscles or the obstructed part through the electrical stimulation device of the diaphragm or pharynx. The above products and means can improve the sleep breathing state of this group of people and improve the sleep quality. In actual application scenarios, different sleep breathing adjustment devices can be selected according to the needs of different groups of people or user scenarios, provided that these sleep breathing adjustment devices have basic command reception and analysis and remote wireless control functions.

[0195] The third step is to generate and output the sleep respiratory muscle status detection and auxiliary adjustment report according to the preset cycle or strategy, and update the user's sleep respiratory database.

[0196] In this embodiment, the sleep respiratory muscle state detection and auxiliary regulation report at least includes a statistical analysis of sleep respiratory muscle state events, a sleep respiratory muscle state detection evaluation summary, an auxiliary regulation evaluation summary, and sleep respiratory behavior improvement suggestions.

[0197] In actual application scenarios, different user report generation or output cycles and different user sleep breathing database strategies or mechanisms may be preset.

[0198] Combination Figure 2 As shown, a sleep respiratory muscle state detection, quantification and auxiliary regulation system provided by an embodiment of the present invention includes the following modules:

[0199] The sleep behavior detection module S100 is used to collect and process the muscle state signal and limb movement signal of the user during sleep, and generate a sleep muscle state signal and a sleep movement position signal respectively;

[0200] The muscle dynamics analysis module S200 is used to perform signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal;

[0201] The event detection and analysis module S300 is used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state event, and further extract the event intensity in combination with the sleep movement position signal;

[0202] The state characteristic analysis module S400 is used to perform characteristic analysis on the sleep respiratory muscle state dynamics signal and the sleep movement position signal to obtain the sleep position muscle state dynamics characteristics;

[0203] The event trend prediction module S500 is used to perform signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics characteristics, and predict and identify the sleep respiratory muscle state events and event occurrence trends;

[0204] The strategy auxiliary adjustment module S600 is used to generate a sleep breathing auxiliary adjustment strategy based on sleep respiratory muscle state events and event occurrence trends, combined with a sleep breathing knowledge base and a user sleep breathing database, and send it to a sleep breathing adjustment device through a signal interface;

[0205] The data operation management module S700 is used for visual management, unified storage and operation management of all process data of the system.

[0206] In this embodiment, the sleep behavior detection module S100 also includes the following functional units:

[0207] A muscle state detection unit, used to collect and process muscle state signals during the user's sleep process and generate a sleeping muscle state signal;

[0208] The motion state detection unit is used to collect and process the limb motion signals of the user during sleep and generate sleep motion position signals.

[0209] In this embodiment, the muscle dynamics analysis module S200 also includes the following functional units:

[0210] A signal decomposition unit, used for decomposing the sleep muscle state signal to obtain a sleep muscle state description signal set;

[0211] A signal recognition unit, used for performing signal recognition on a sleep muscle state description signal set to obtain a sleep muscle state dynamics description signal set;

[0212] The fusion and extraction unit is used to perform signal fusion and / or envelope extraction on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

[0213] In this embodiment, the event detection and analysis module S300 also includes the following functional units:

[0214] An event detection and analysis unit, used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal, extract basic event information, and obtain sleep respiratory muscle state events;

[0215] The event intensity calculation unit is used to calculate the event intensity of the sleep respiratory muscle state event according to the sleep respiratory muscle state dynamics signal, the event basic information and the sleep movement position signal.

[0216] In this embodiment, the state feature analysis module S400 further includes the following functional units:

[0217] A muscle state characteristic analysis unit, used for performing characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristics;

[0218] A motion body position characteristic analysis unit, used for performing characteristic analysis on the sleep motion body position signal to obtain the sleep motion body position characteristic;

[0219] The state feature aggregation unit is used to aggregate the sleep respiratory muscle state features and the sleep movement posture features to obtain the sleep posture muscle state dynamic features.

[0220] In this embodiment, the event trend prediction module S500 also includes the following functional units:

[0221] A muscle state prediction unit, used for performing signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state dynamics prediction signal;

[0222] A motion body position prediction unit, used for performing signal trend analysis on the sleep motion body position signal to obtain a sleep motion body position prediction signal;

[0223] A dynamic feature prediction unit, used for performing signal trend analysis on the dynamic features of the sleeping position muscle state to obtain the dynamic feature prediction of the sleeping position muscle state;

[0224] An event detection and extraction unit is used to perform event detection and analysis on the sleep respiratory muscle state dynamics prediction signal, and predict and identify the sleep respiratory muscle state events and event intensity in combination with the sleep movement position prediction signal;

[0225] The event trend recognition unit is used to extract and generate event occurrence trends based on sleep respiratory muscle state events and event intensity combined with sleep postural muscle state dynamic prediction features.

[0226] In this embodiment, the policy auxiliary adjustment module S600 further includes the following functional units:

[0227] An auxiliary strategy generating unit, used to generate a sleep breathing auxiliary adjustment strategy according to sleep breathing muscle state events and event occurrence trends, combined with a sleep breathing knowledge base and a user sleep breathing database;

[0228] A strategy assistance control unit, used to send the sleep breathing assistance adjustment strategy to the sleep breathing adjustment device through a signal interface, optimize the control execution of the sleep breathing adjustment device, so as to realize dynamic assistance adjustment of the user's sleep breathing;

[0229] The report data management unit is used to generate and output sleep respiratory muscle status detection and auxiliary adjustment reports according to preset cycles or strategies, and update the user's sleep respiratory database.

[0230] In this embodiment, the data operation management module S700 also includes the following functional units:

[0231] User information management unit, used for registration, input, editing, query, output and deletion of user information;

[0232] Data visualization management unit, used for visualization and management of all data in the system;

[0233] The data operation management unit is used to store, back up, migrate and export all data in the system.

[0234] The system is configured to execute Figure 1 The various steps in the method will not be repeated here.

[0235] Combination Figure 3 As shown, a sleep respiratory muscle state detection, quantification and auxiliary regulation device provided by an embodiment of the present invention includes the following modules:

[0236] The sleep behavior detection module M100 is used to collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleep muscle state signals and sleep movement position signals respectively;

[0237] The muscle dynamics analysis module M200 is used to perform signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate the sleep respiratory muscle state dynamics signal;

[0238] The event detection and analysis module M300 is used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal, obtain the sleep respiratory muscle state event, and further extract the event intensity in combination with the sleep movement position signal;

[0239] The state feature analysis module M400 is used to perform feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement position signal to obtain the sleep position muscle state dynamics feature;

[0240] Event trend prediction module M500 is used to perform signal trend analysis and event detection analysis on sleep respiratory muscle state dynamics signals, sleep movement position signals and sleep position muscle state dynamic characteristics, and predict and identify sleep respiratory muscle state events and event occurrence trends;

[0241] The strategy auxiliary adjustment module M600 is used to generate a sleep breathing auxiliary adjustment strategy based on sleep respiratory muscle state events and event occurrence trends, combined with a sleep breathing knowledge base and a user sleep breathing database, and send the sleep breathing adjustment device through a signal interface;

[0242] The data operation management module M700 is used for visual management, unified storage and operation management of all process data of the system.

[0243] The device is configured to execute Figure 1 The various steps in the method are not repeated here.

[0244] The present invention also provides various types of programmable processors (FPGA, ASIC or other integrated circuits), which are used to run programs, wherein the steps in the above embodiments are executed when the program is running.

[0245] The present invention also provides a corresponding computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps in the above-mentioned embodiments are implemented when the memory executes the program.

[0246] Although the embodiments disclosed in the present invention are as above, the contents described are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the field to which the present invention belongs can make any modifications and changes, equivalent substitutions, etc. in the form and details of implementation without departing from the spirit and principles disclosed in the present invention, and these all belong to the protection scope of the present invention. Therefore, the patent protection scope of the present invention shall still be subject to the scope defined by the attached claims.

Claims

1. A method for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles, characterized in that: The following steps are involved: Collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleeping muscle state signals and sleeping movement position signals respectively; Performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal; Performing event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state event, and further extracting event intensity in combination with the sleep movement position signal; Performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain a sleep body position muscle state dynamics feature; Performing signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predicting and identifying the sleep respiratory muscle state events and event occurrence trends; According to the sleep respiratory muscle state events and the occurrence trends of the events, combined with the sleep respiratory knowledge base and the user sleep respiratory database, a sleep respiratory assistance adjustment strategy is generated and sent to the sleep respiratory adjustment device through a signal interface.

2. The method according to claim 1, characterized in that The specific steps of collecting and processing the muscle state signal and limb movement signal of the user during sleep to generate the sleeping muscle state signal and the sleeping movement body position signal respectively also include: Collecting and processing the muscle state signal of the user during sleep to generate the sleeping muscle state signal; The limb movement signals of the user during sleep are collected and processed to generate the sleep movement position signals.

3. The method according to claim 2, characterized in that The acquisition processing at least includes acquisition, analog-to-digital conversion, resampling, re-referencing, artifact removal, signal correction, noise reduction, power frequency notch, bandpass filtering, mean filtering, smoothing processing and signal time window segmentation; wherein, the signal time window segmentation is specifically to perform continuous time window segmentation on the target signal with a preset length time window and a preset translation time step to obtain a multi-time window signal set.

4. The method according to claim 2, characterized in that , the sleeping muscle state signal includes at least any one of an electromyographic signal and an electrical skin signal; the sleeping movement position signal includes at least any one of a limb acceleration signal, a video movement position detection signal and a radar movement position detection signal.

5. The method according to claim 1, characterized in that The specific steps of performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate the sleep respiratory muscle state dynamics signal also include: Decomposing the sleep muscle state signal to obtain a sleep muscle state description signal set; Performing signal recognition on the sleep muscle state description signal set to obtain a sleep muscle state dynamics description signal set; Signal fusion and / or envelope extraction are performed on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

6. The method according to claim 5, characterized in that The signal decomposition specifically decomposes the target signal into a set of component signals with multiple different modes, different sources and different characteristic attributes; the signal decomposition method includes at least time series decomposition, empirical mode decomposition, variational mode decomposition, local mean decomposition, wavelet transform, wavelet packet transform, time-frequency transform, blind source separation, linear discriminant analysis, detrending analysis, principal component analysis, independent component analysis, waveform analysis and numerical fitting, and any of their evolutionary variants.

7. The method according to claim 5, characterized in that The signal recognition specifically involves identifying and labeling the signal source of the target signal through a waveform determination method or a machine learning model; the signal source includes at least breathing, swallowing, clenching / grinding / chewing, head movement, and body movement.

8. The method according to claim 5 or 7, characterized in that The signal fusion is specifically to perform synchronous average superposition or delayed average superposition on different component signals from the same signal source to fuse them into one source signal.

9. The method according to claim 1 or 5, characterized in that: The sleep respiratory muscle state dynamics signal includes at least one of an electromyography envelope signal, a skin electrode SCL level signal, and a skin electrode SCR level signal.

10. The method according to claim 1, characterized in that The specific steps of performing event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state event, and further extracting the event intensity in combination with the sleep movement position signal also include: Performing event detection and analysis on the sleep respiratory muscle state dynamics signal, extracting basic event information, and obtaining the sleep respiratory muscle state event; The event intensity of the sleep respiratory muscle state event is calculated based on the sleep respiratory muscle state dynamics signal, the event basic information and the sleep movement body position signal.

11. The method according to claim 10, characterized in that The event detection analysis specifically includes performing waveform feature recognition and label extraction on the sleep respiratory muscle state dynamics signal according to a preset sleep respiratory-muscle event knowledge base and / or a machine learning model to obtain basic event information of the sleep respiratory muscle state event.

12. The method according to claim 11, characterized in that The basic event information includes at least event type, start time, end time, duration, peak and valley values, and time at the peak and valley values.

13. The method according to any one of claims 10 to 12, characterized in that: The sleep respiratory muscle state event at least includes event type, start time, end time, duration, peak and valley values, peak and valley value time, and event intensity.

14. The method according to claim 13, characterized in that The event intensity is specifically determined by the duration, start time, end time, peak-to-valley values, and relative time between the peak-to-valley values ​​of the sleep respiratory muscle state event, as well as the characteristic amplitude change of the sleep respiratory muscle state.

15. The method according to claim 13, characterized in that The method for calculating and generating the event intensity includes: 1) obtaining the start time, end time, duration, peak-to-valley value and peak-to-valley value time of the sleep respiratory muscle state event, and presetting the sleep respiratory muscle state dynamics signal threshold; 2) calculating the relative change of the peak-to-valley value of the sleep respiratory muscle state event and the preset sleep respiratory muscle state dynamic signal threshold to obtain the peak-to-valley relative value; 3) calculating the linear slope according to the start time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley front slope; 4) calculating the linear slope according to the end time, peak-to-valley value and time at the peak-to-valley value of the sleep respiratory muscle state event to obtain the peak-to-valley trailing edge slope; 5) extracting sleeping angle and sleeping direction information or preset sleeping position correction calculation strategy according to the sleeping movement position signal, and extracting sleeping position correction coefficient according to the preset sleeping position correction coefficient comparison table; 6) extracting the correction coefficient of the dynamics signal source according to the signal source of the sleep respiratory muscle state dynamics and according to a preset dynamics signal source-correction coefficient comparison table; 7) The event intensity is obtained according to the numerical fusion calculation of the duration of the sleep respiratory muscle state event, the peak-to-valley relative value, the peak-to-valley leading slope, the peak-to-valley trailing slope, the sleep position correction coefficient, and the dynamic signal source correction coefficient.

16. The method according to claim 1, wherein: The specific step of performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature also includes: Performing characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristics; Performing feature analysis on the sleep movement position signal to obtain the sleep movement position feature; The sleep respiratory muscle state characteristics and the sleep movement body position characteristics are aggregated to obtain the sleep body position muscle state dynamic characteristics.

17. The method according to claim 16, characterized in that The feature analysis includes at least one of numerical feature analysis, envelope feature analysis, time-frequency feature analysis, and nonlinear feature analysis; wherein the numerical feature includes at least one of the mean value, root mean square, maximum value, minimum value, variance, standard deviation, coefficient of variation, kurtosis, and skewness; the time-frequency feature includes at least one of the frequency band power, frequency band power ratio, and frequency band center frequency; and the nonlinear feature includes at least one of the entropy feature, fractal feature, and complexity feature.

18. The method according to claim 16, characterized in that The sleep position muscle state dynamic characteristics at least include the sleep respiratory muscle state characteristics and the sleep movement position characteristics; the sleep respiratory muscle state characteristics at least include any one of respiratory numerical characteristics, electromyographic characteristics, and skin electricity characteristics; the sleep movement position characteristics at least include limb movement characteristics and sleeping posture characteristics.

19. The method of claim 1, wherein: The specific steps of performing signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predicting and identifying the sleep respiratory muscle state event and the event occurrence trend also include: Performing signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state dynamics prediction signal; Performing signal trend analysis on the sleep movement position signal to obtain a sleep movement position prediction signal; Performing signal trend analysis on the sleep postural muscle state dynamic characteristics to obtain sleep postural muscle state dynamic prediction characteristics; Performing event detection analysis on the sleep respiratory muscle state dynamics prediction signal, combining the sleep movement body position prediction signal, predicting and identifying the sleep respiratory muscle state event and the event intensity; According to the sleep respiratory muscle state event and the event intensity, combined with the sleep postural muscle state dynamic prediction characteristics, the event occurrence trend is extracted and generated.

20. The method of claim 19, wherein: The signal trend analysis includes at least any one of exponential smoothing method, Holt-Winters method, AR, MA, ARMA, ARIMA, SARIMA, SARIMAX, VAR, VARMA, VARMAX, and machine learning.

21. The method of claim 19, wherein: The event occurrence trend includes at least the event occurrence type, event occurrence probability and event occurrence intensity; the event occurrence trend is obtained by machine learning through learning and training, data modeling and analytical calculation of large-scale sleeping user sample data such as the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamic characteristics related to the sleep respiratory muscle state event.

22. The method of claim 1, wherein: The specific steps of generating a sleep breathing assistance adjustment strategy based on the sleep breathing muscle state event and the event occurrence trend, combining a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to a sleep breathing adjustment device through a signal interface also include: Generate the sleep breathing assistance adjustment strategy according to the sleep breathing muscle state event and the event occurrence trend, combined with the sleep breathing knowledge base and the user sleep breathing database; Sending the sleep breathing assistance adjustment strategy to the sleep breathing adjustment device through the signal interface, optimizing the control execution of the sleep breathing adjustment device, so as to achieve dynamic assistance adjustment of the user's sleep breathing; According to a preset cycle or strategy, a sleep respiratory muscle state detection and auxiliary adjustment report is generated and output, and the user sleep respiratory database is updated.

23. The method of claim 22, wherein: The sleep breathing knowledge base mainly comes from the knowledge and experience of health management and clinical medicine related to sleep breathing, and at least includes the laws of sleep breathing-muscle movement, common sleep breathing event characteristics, commonly used sleep breathing adjustment methods and scene intervention parameter guidance; the user sleep breathing database at least includes the sleep respiratory muscle state events, the sleep respiratory muscle state dynamic signals, the sleep movement position signals, the sleep position muscle state dynamic characteristics, the event occurrence trends and the sleep breathing assistance adjustment strategies.

24. The method of claim 22, wherein: The sleep breathing assistance adjustment strategy includes at least one of a target value of breathing frequency, a target value of breathing depth, an adjustment method, an adjustment time point, a duration and a device control parameter.

25. The method of claim 22, wherein: The sleep breathing regulation device at least includes a ventilator, an intelligent mattress, an odor stimulation device, an electrical stimulation device, a tactile stimulation device, an environmental temperature and humidity control device and an environmental Any item of concentration control equipment.

26. The method according to any one of claims 22 to 25, characterized in that The sleep respiratory muscle state detection and auxiliary regulation report at least includes a statistical analysis of the sleep respiratory muscle state events, a sleep respiratory muscle state detection evaluation summary, an auxiliary regulation evaluation summary, and sleep respiratory behavior improvement suggestions.

27. A sleep respiratory muscle state detection, quantification and auxiliary regulation system, characterized in that: Includes the following modules: The sleep behavior detection module is used to collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleep muscle state signals and sleep movement body position signals respectively; A muscle dynamics analysis module, used for performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal; An event detection and analysis module, used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain sleep respiratory muscle state events, and further extract event intensity in combination with the sleep movement position signal; A state feature analysis module, used for performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature; An event trend prediction module is used to perform signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predict and identify the sleep respiratory muscle state events and event occurrence trends; A strategy auxiliary adjustment module, for generating a sleep breathing auxiliary adjustment strategy based on the sleep respiratory muscle state event and the event occurrence trend, in combination with a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to a sleep breathing adjustment device through a signal interface; The data operation management module is used for visual management, unified storage and operation management of all process data of the system.

28. The system of claim 27, wherein: The sleep behavior detection module also includes the following functional units: A muscle state detection unit, used to collect and process the muscle state signal of the user during sleep and generate the sleeping muscle state signal; The motion state detection unit is used to collect and process the limb motion signals of the user during sleep and generate the sleep motion position signals.

29. The system according to claim 27 or 28, characterized in that The muscle dynamics analysis module also includes the following functional units: A signal decomposition unit, used for performing signal decomposition on the sleep muscle state signal to obtain a sleep muscle state description signal set; A signal recognition unit, configured to perform signal recognition on the sleep muscle state description signal set to obtain a sleep muscle state dynamics description signal set; The fusion and extraction unit is used to perform signal fusion and / or envelope extraction on the sleep muscle state dynamics description signal set to obtain the sleep respiratory muscle state dynamics signal.

30. The system of claim 29, wherein: The event detection and analysis module also includes the following functional units: An event detection and analysis unit, used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal, extract basic event information, and obtain the sleep respiratory muscle state event; The event intensity calculation unit is used to calculate the event intensity of the sleep respiratory muscle state event according to the sleep respiratory muscle state dynamics signal, the event basic information and the sleep movement position signal.

31. The system of claim 30, wherein: The state feature analysis module also includes the following functional units: A muscle state characteristic analysis unit, used for performing characteristic analysis on the sleep respiratory muscle state dynamics signal to obtain the sleep respiratory muscle state characteristic; A motion body position characteristic analysis unit, used for performing characteristic analysis on the sleep motion body position signal to obtain the sleep motion body position characteristic; The state feature collection unit is used to collect the sleep respiratory muscle state features and the sleep movement posture features to obtain the sleep posture muscle state dynamic features.

32. The system of claim 30, wherein: The event trend prediction module also includes the following functional units: A muscle state prediction unit, used for performing a signal trend analysis on the sleep respiratory muscle state dynamics signal to obtain a sleep respiratory muscle state dynamics prediction signal; A movement position prediction unit, used for performing a signal trend analysis on the sleep movement position signal to obtain a sleep movement position prediction signal; A dynamic feature prediction unit, used for performing signal trend analysis on the dynamic feature of the sleeping position muscle state to obtain the dynamic feature prediction of the sleeping position muscle state; An event detection and extraction unit, configured to perform event detection and analysis on the sleep respiratory muscle state dynamics prediction signal, and predict and identify the sleep respiratory muscle state event and the event intensity in combination with the sleep movement body position prediction signal; The event trend identification unit is used to extract and generate the event occurrence trend according to the sleep respiratory muscle state event and the event intensity in combination with the sleep postural muscle state dynamic prediction feature.

33. The system of claim 31 or 32, wherein: The strategy auxiliary adjustment module also includes the following functional units: An auxiliary strategy generating unit, configured to generate the sleep breathing auxiliary adjustment strategy according to the sleep breathing muscle state event and the event occurrence trend, in combination with a sleep breathing knowledge base and a user sleep breathing database; A strategy auxiliary control unit, used to send the sleep breathing auxiliary adjustment strategy to the sleep breathing adjustment device through a signal interface, optimize the control execution of the sleep breathing adjustment device, so as to achieve dynamic auxiliary adjustment of the user's sleep breathing; The report data management unit is used to generate and output a sleep respiratory muscle state detection and auxiliary adjustment report according to a preset cycle or strategy, and update the user sleep respiratory database.

34. The system of claim 31 or 32, wherein: The data operation management module also includes the following functional units: User information management unit, used for registration, input, editing, query, output and deletion of user information; A data visualization management unit, used for visual display and management of all data in the system; The data operation management unit is used to store, back up, migrate and export all data in the system.

35. A device for detecting, quantifying and assisting in regulating the state of sleeping respiratory muscles, characterized in that: Includes the following modules: The sleep behavior detection module is used to collect and process the muscle state signals and limb movement signals of the user during sleep, and generate sleep muscle state signals and sleep movement position signals respectively; A muscle dynamics analysis module, used for performing signal decomposition, signal recognition, signal fusion and / or envelope extraction on the sleep muscle state signal to generate a sleep respiratory muscle state dynamics signal; An event detection and analysis module is used to perform event detection and analysis on the sleep respiratory muscle state dynamics signal to obtain sleep respiratory muscle state events, and further extract event intensity in combination with the sleep movement position signal; A state feature analysis module, used for performing feature analysis on the sleep respiratory muscle state dynamics signal and the sleep movement body position signal to obtain the sleep body position muscle state dynamics feature; An event trend prediction module is used to perform signal trend analysis and event detection analysis on the sleep respiratory muscle state dynamics signal, the sleep movement position signal and the sleep position muscle state dynamics feature, and predict and identify the sleep respiratory muscle state events and event occurrence trends; A strategy auxiliary adjustment module, for generating a sleep breathing auxiliary adjustment strategy according to the sleep breathing muscle state event and the event occurrence trend, in combination with a sleep breathing knowledge base and a user sleep breathing database, and sending the strategy to a sleep breathing adjustment device through a signal interface; The data operation management module is used for visual management, unified storage and operation management of all process data of the device.

Images