Sleep data analysis and management system based on artificial intelligence

Abstract

The invention relates to the technical field of integrated sleep management, in particular to a sleep data analysis management system based on artificial intelligence. The method comprises the following steps: segmenting a sleep cycle of a user based on brain wave data to obtain N sleep stages of the user; based on the body movement monitoring data, the pulse monitoring data and the respiration monitoring data, calculating to obtain a first sleep quality characteristic value of the sleep stage of the user, and based on the sleep posture category and a sleep quality weight parameter corresponding to the sleep posture category, determining a second sleep quality characteristic value of the sleep stage of the user; the sleep quality analysis index of the sleep cycle of the user is generated based on the sleep quality parameters of the sleep stage of the user, the sleep early warning signal is generated based on the sleep quality analysis index of the sleep cycle of the user, multi-dimensional data such as brain waves, body movement, pulse, breathing and sleeping postures can be combined, the sleep stage can be accurately divided, and the sleep early warning effect is improved. And sleep quality parameters are calculated through multi-modal data fusion, so that the accuracy and practicability of sleep monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated sleep management, and in particular to a sleep data analysis and management system based on artificial intelligence. Background Art

[0002] With the accelerating pace of modern life, sleep health issues are receiving increasing attention. Sleep quality directly affects a person's physical health, psychological state, and work efficiency. Therefore, scientific monitoring and analysis of sleep are of great significance.

[0003] In recent years, with the development of wearable devices and artificial intelligence technology, sleep monitoring solutions based on non-invasive sensors have gradually emerged. For example, smart bracelets or mattress sensors collect heart rate, body movement, breathing, and other data, and combine them with algorithms to analyze sleep quality. However, existing technologies still have the following shortcomings: Insufficient precision in sleep stage classification: Some solutions rely solely on body movement or heart rate data to roughly determine sleep status, making it difficult to accurately distinguish between the various stages of the sleep cycle, affecting the accuracy of sleep quality assessment.

[0004] Insufficient multimodal data fusion: Existing systems often analyze physiological signals (such as breathing and pulse) or behavioral data (such as sleeping posture) separately, lacking collaborative analysis of multi-source data. This leads to a single dimension of sleep quality assessment, further affecting the accuracy of sleep quality assessment.

[0005] Therefore, there is an urgent need for an artificial intelligence-based sleep data analysis and management system to solve the above problems. Summary of the Invention

[0006] The purpose of the present invention is to provide a sleep data analysis and management system based on artificial intelligence to solve the technical problem of low accuracy of sleep quality assessment in existing sleep data analysis systems.

[0007] A sleep data analysis and management system based on artificial intelligence, the system includes: A sleep cycle segmentation unit is used to obtain brain wave data of the user in the sleeping state, and segment the user's sleep cycle based on the brain wave data to obtain the user's N sleep stages; A sleep monitoring data acquisition unit, configured to acquire sleep monitoring data of N sleep stages, wherein the sleep monitoring data includes body movement monitoring data, pulse monitoring data, respiratory monitoring data, and sleeping posture monitoring video data; a sleep data analysis unit, configured to calculate a first sleep quality characteristic value of a user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiratory monitoring data; determine a sleeping posture category of the user's sleep stage based on the sleeping posture monitoring video data; determine a second sleep quality characteristic value of the user's sleep stage based on the sleeping posture category and a sleep quality weight parameter corresponding to the sleeping posture category; and determine a sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value; The sleep warning management unit is used to generate a sleep quality analysis index of the user's sleep cycle based on the sleep quality parameters of the user's sleep stage, and generate a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle.

[0008] Furthermore, segmenting the user's sleep cycle based on brain wave data specifically includes the following steps: Dividing the EEG data into a plurality of EEG segments based on the same time interval, obtaining a first instantaneous average heart rate and a first respiratory rate change trend at a location adjacent to a previous EEG segment, and obtaining a second instantaneous average heart rate and a second respiratory rate change trend at a location adjacent to a next EEG segment, wherein the respiratory rate change trend includes an upward trend and a downward trend; The first instantaneous average heart rate, the first respiratory rate change trend, the second instantaneous average heart rate and the second respiratory rate change trend are recorded as feature combinations, and PCA analysis is performed on the feature combinations to obtain the principal component FI of the feature combinations; Convert FI into a zero-mean unit variance form, and smooth the principal component FI within a small and large frame length based on a smoothing filter, respectively, to obtain the smoothing results SFI and BFI, and determine the offset k of BFI; When SFI>BFI+k, the current brain wave segment is the data of the awake stage or REM stage. The current brain wave segment that is earlier in time sequence is the data of the awake stage, and the current brain wave segment that is later in time sequence is the data of the REM stage. When SFI<BFI+k, the current brain wave segment is the data of the deep sleep stage. When SFI=BFI+k, the current brain wave segment is the data of the light sleep stage.

[0009] Furthermore, calculating the first sleep quality characteristic value of the user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiration monitoring data specifically includes the following process: Based on the body motion monitoring data, the number of tossing and turning, the number of limb twitching, and the number of head rotations during the user's sleep stage are obtained, and the sum of the number of tossing and turning, the number of limb twitching, and the number of head rotations is recorded as the body motion representation value A; Based on the pulse monitoring data, the blood oxygen saturation values ​​of each detection point during the user's sleep stage are collected. A rectangular coordinate system is established with the number of detection time points as the horizontal axis and the blood oxygen saturation value as the vertical axis. A blood oxygen saturation value curve is drawn by plotting points. The difference between the user's average blood oxygen saturation value and the lowest blood oxygen saturation value during the sleep stage is calculated. The integral value of the blood oxygen saturation value curve above the straight line of the average blood oxygen saturation value is calculated. The integral value and the difference are multiplied after unifying the dimensions to obtain the product value B; The number of central apneas C and the total duration D of apneas of the user are collected based on the respiratory monitoring data. If the number of central apneas of the user is 0, the user's respiratory quality parameter R is 1; if the number of central apneas of the user is not 0, the user's respiratory quality parameter R is 0; A first sleep quality characteristic value of the user's sleep stage is obtained based on the body movement representation value A, the product value B, the number of times C, the total duration D and the breathing quality parameter R.

[0010] Furthermore, obtaining the first sleep quality characteristic value of the user's sleep stage based on the body movement representation value A, the product value B, the number of times C, the total duration D, and the breathing quality parameter R specifically includes the following process: Substitute the body movement representation value A, the product value B, the number of times C, the total duration D, and the breathing quality parameter R into the correlation formula to calculate the first sleep quality characteristic value LMS of the user's sleep stage. The correlation formula is as follows: ; in, 、 and is the weight, whose value is set by the system, The value of is 2.72.

[0011] Furthermore, determining the sleeping posture category of the user's sleep stage based on the sleeping posture monitoring video data specifically includes the following process: Extracting user sleeping posture video frames based on sleeping posture monitoring video data; A deep convolutional neural network model is used to realize the judgment of sleeping posture category. The deep convolutional neural network model consists of several cells, each cell contains N1 nodes. Represents the node parameters of each Cell, 、 Represents a node, and its value is ; Given a set of user sleeping posture video frames X with multiple categories, the process of training a deep convolutional neural network model is equivalent to determining each neural network operation The process: make For the and Between nodes, The probability distribution of a parameterized neural network operation, Represents the joint distribution, each Cell is composed of operations, for a given user sleeping posture video frame set H and a candidate operation set , calculate an operation distribution Conditional entropy : ; in, is the probability distribution of the data set X, is a set of neural network operation operations; Conditional entropy of all operations in the entire neural network Cell for: ; Calculate the log-likelihood function based on conditional entropy : ; in, ; is a discrete characteristic indicator function, if If the value of is less than the preset threshold, is 1, otherwise, is 0, For nodes and nodes The Euclidean distance between in, ; By maximizing Add maximum entropy constraints to update the category index of the user's sleeping posture ; Category Index Perform analysis to obtain the category corresponding to the user's sleeping posture.

[0012] Furthermore, the category index The analysis to determine the category corresponding to the user's sleeping posture includes the following steps: Get the category index corresponding to all cells , and class index based on the order of Cell operation Numbering to get category index The number value of The horizontal axis of the rectangular coordinate system with numbered values, category index The vertical axis of the rectangular coordinate system is the vertical axis of the rectangular coordinate system. Mark them as points, connect adjacent points in the rectangular coordinate system to generate a category index curve, draw perpendicular lines from the two ends of the category index curve to the horizontal axis to obtain two start and end line segments, and form a closed figure with the designed category index curve, the two start and end line segments, and the horizontal axis. Calculate the total area of ​​the closed figure; set a constant straight line, obtain the angle at which the constant straight line intersects the category index curve, and record the largest angle as the reference angle; multiply the reference angle by the total area of ​​the closed figure to obtain the category feature value of the user's sleeping posture; The category corresponding to the user's sleeping posture is queried based on the category characteristic value of the user's sleeping posture, wherein each user's sleeping posture category corresponds to a preset category characteristic value interval.

[0013] Furthermore, obtaining the sleep quality weight parameter corresponding to the sleeping posture category specifically includes the following process: Determine the judgment matrix for each sleeping posture category: Determine the number of factors affecting sleep quality weight parameters , construct a set of weighted influencing factors , ,in, is the subset corresponding to the average heart rate of the user when sleeping in the supine position, is the subset corresponding to the average heart rate when the user sleeps in the left side position, is the subset corresponding to the average heart rate of the user when sleeping in the right side, from the set Take out two subsets in the same level and compare them, and use Represents the ratio of importance, and assigns the corresponding importance according to the preset ratio, combines the importance of each layer to form a judgment matrix; calculates the maximum eigenvalue of the judgment matrix : ; in, It is the matrix obtained by normalizing each column vector of the judgment matrix. The value of is 1,2...m, is a matrix The elements of are added row by row to obtain the vector and then normalized into a matrix. is a matrix Add the elements of column by column to obtain the vector and then normalize the matrix; Calculate the sleep quality weight parameter corresponding to the sleeping posture category : , where b is the order of the judgment matrix of the sleeping posture category.

[0014] Furthermore, determining the sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value specifically includes the following process: adding the first sleep quality characteristic value and the second sleep quality characteristic value after dimension normalization to obtain the sleep quality parameter of the user's sleep stage.

[0015] Furthermore, generating a sleep quality analysis index for a user's sleep cycle based on the sleep quality parameters of the user's sleep stages specifically includes the following processes: obtaining the sleep quality parameters for all the user's sleep stages, calculating the average of the sleep quality parameters for all the user's sleep stages, and recording the average as the sleep quality analysis index for the user's sleep cycle.

[0016] Furthermore, generating a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle specifically includes the following process: A sleep quality analysis index threshold is loaded, the value of which is set by the system and stored in the system, and it is determined whether the sleep quality analysis index of the user's sleep cycle exceeds the sleep quality analysis index threshold. If so, no warning signal for low sleep quality is generated; otherwise, a warning signal for low sleep quality is generated.

[0017] Compared with the existing solutions, the present invention achieves the following beneficial effects: Improve the accuracy of sleep stage classification The system accurately segments sleep cycles using brainwave data and automatically identifies the user's N sleep stages using artificial intelligence algorithms. Compared to traditional monitoring methods that rely on body movement or heart rate data, it significantly improves the accuracy of sleep staging and provides a reliable basis for subsequent sleep quality assessment. Enhance the comprehensiveness of sleep quality assessment The system simultaneously collects body movement, pulse, respiration, and sleeping posture monitoring video data, combines the first sleep quality characteristic value (physiological indicator) and the second sleep quality characteristic value (sleeping posture influence), and comprehensively calculates the sleep quality parameters, avoiding the limitations of a single data source and making the evaluation results more in line with the user's actual sleep state; The sleeping posture category (such as supine, side, and prone) is identified through sleeping posture monitoring video data, and the second sleep quality characteristic value is dynamically adjusted based on the weight of the impact of different sleeping postures on sleep quality, making sleep quality assessment more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0019] Figure 1This is a system block diagram of a sleep data analysis and management system based on artificial intelligence according to an embodiment of the present invention; Figure 2 This is a workflow diagram of the first artificial intelligence-based sleep data analysis and management system according to an embodiment of the present invention; Figure 3 This is a workflow diagram of a second artificial intelligence-based sleep data analysis and management system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] In addition, the described features, structures or characteristics can be combined in any suitable manner in one or more example embodiments. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure can be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. can be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0022] This embodiment provides a sleep data analysis and management system based on artificial intelligence. Figure 1 This is a system block diagram of a sleep data analysis and management system based on artificial intelligence according to an embodiment of the present invention. Figure 1 As shown, the system includes: A sleep cycle segmentation unit is used to obtain brain wave data of the user in the sleeping state, and segment the user's sleep cycle based on the brain wave data to obtain the user's N sleep stages; A sleep monitoring data acquisition unit, configured to acquire sleep monitoring data of N sleep stages, wherein the sleep monitoring data includes body movement monitoring data, pulse monitoring data, respiratory monitoring data, and sleeping posture monitoring video data; a sleep data analysis unit, configured to calculate a first sleep quality characteristic value of a user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiratory monitoring data; determine a sleeping posture category of the user's sleep stage based on the sleeping posture monitoring video data; determine a second sleep quality characteristic value of the user's sleep stage based on the sleeping posture category and a sleep quality weight parameter corresponding to the sleeping posture category; and determine a sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value; The sleep warning management unit is used to generate a sleep quality analysis index of the user's sleep cycle based on the sleep quality parameters of the user's sleep stage, and generate a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle.

[0023] In summary, the present invention obtains brain wave data of a user in a sleeping state, segments the user's sleep cycle based on the brain wave data, and obtains N sleep stages of the user; collects sleep monitoring data of N sleep stages, calculates a first sleep quality characteristic value of the user's sleep stage based on body movement monitoring data, pulse monitoring data, and breathing monitoring data, determines the sleeping posture category of the user's sleep stage based on sleeping posture monitoring video data, and determines a second sleep quality characteristic value of the user's sleep stage based on the sleeping posture category and the sleep quality weight parameter corresponding to the sleeping posture category; determines the sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value; generates a sleep quality analysis index of the user's sleep cycle based on the sleep quality parameter of the user's sleep stage, and generates a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle. It can combine multi-dimensional data such as brain waves, body movement, pulse, breathing, and sleeping posture to accurately divide the sleep stages, and calculate the sleep quality parameters through multimodal data fusion, and finally generate a personalized sleep warning signal to improve the accuracy and practicality of sleep monitoring.

[0024] In some embodiments, Figure 2 This is a workflow diagram of the first artificial intelligence-based sleep data analysis and management system according to an embodiment of the present invention. Figure 2 As shown, the segmentation of the user's sleep cycle based on brain wave data specifically includes the following steps: Step S201: Dividing the EEG data into a plurality of EEG segments based on the same time interval, obtaining a first instantaneous average heart rate and a first respiratory rate change trend at a location adjacent to a previous EEG segment, and obtaining a second instantaneous average heart rate and a second respiratory rate change trend at a location adjacent to a next EEG segment; Among them, the respiratory rate change trend includes an upward trend and a downward trend.

[0025] Step S202: Recording the first instantaneous average heart rate, the first respiratory rate change trend, the second instantaneous average heart rate, and the second respiratory rate change trend as a feature combination, performing PCA analysis on the feature combination to obtain a principal component FI of the feature combination; Specifically, PCA analysis can include: (1) Combine the features to form a matrix and calculate the covariance matrix of the matrix samples; (2) Calculate the eigenvectors of the covariance matrix , ,…, , p is the number of characteristic eigenvectors; (3) Project the data into the feature vector space, using the formula: ; Among them, the BV value is the value of the corresponding dimension in the original sample.

[0026] right Feature vectors greater than a preset threshold are extracted to obtain the principal component FI of the feature combination.

[0027] Step S203: Convert FI into a zero-mean unit variance form, and perform smoothing processing on the principal component FI within the small and large frame lengths respectively based on a smoothing filter to obtain smoothing results SFI and BFI, and determine the offset k of BFI; Step S204: When SFI>BFI+k, the current brain wave segment is data of the awake stage or REM stage. The current brain wave segment that is earlier in time sequence is data of the awake stage, and the current brain wave segment that is later in time sequence is data of the REM stage. When SFI<BFI+k, the current brain wave segment is data of the deep sleep stage. When SFI=BFI+k, the current brain wave segment is data of the light sleep stage.

[0028] In some embodiments, Figure 3 This is a workflow diagram of the second artificial intelligence-based sleep data analysis and management system according to an embodiment of the present invention. Figure 3 As shown, calculating the first sleep quality characteristic value of the user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiration monitoring data specifically includes the following process: Step S301: obtaining the number of tossing and turning, limb twitching, and head rotations of the user during sleep based on the body motion monitoring data, and recording the sum of the number of tossing and turning, limb twitching, and head rotation as the body motion representation value A; Step S302: Based on the pulse monitoring data, the blood oxygen saturation values ​​of each detection point during the user's sleep stage are collected. A rectangular coordinate system is established with the number of detection time points as the horizontal axis and the blood oxygen saturation value as the vertical axis. A blood oxygen saturation value curve is drawn by plotting points. The difference between the user's average blood oxygen saturation value and the lowest blood oxygen saturation value during the sleep stage is calculated. The integral value of the blood oxygen saturation value curve above the straight line of the average blood oxygen saturation value is calculated. The integral value and the difference are multiplied after being dimensionalized to obtain a product value B. Step S303: Based on the respiratory monitoring data, the number of central apneas C and the total duration of apnea D of the user are collected. If the number of central apneas is 0, the user's respiratory quality parameter R is 1; if the number of central apneas is not 0, the user's respiratory quality parameter R is 0. Step S304: obtaining a first sleep quality characteristic value of the user's sleep stage based on the body movement representation value A, the product value B, the number of times C, the total duration D and the breathing quality parameter R.

[0029] Furthermore, the body motion representation value A, the product value B, the number of times C, the total duration D, and the breathing quality parameter R are substituted into the correlation formula to calculate the first sleep quality characteristic value LMS of the user's sleep stage. The correlation formula is as follows: ; in, 、 and is the weight, whose value is set by the system, The value of is 2.72.

[0030] In some embodiments, determining the sleeping posture category of the user's sleep stage based on the sleeping posture monitoring video data specifically includes the following process: Extracting user sleeping posture video frames based on sleeping posture monitoring video data; A deep convolutional neural network model is used to realize the judgment of sleeping posture category. The deep convolutional neural network model consists of several cells, each cell contains N1 nodes. Represents the node parameters of each Cell, 、 Represents a node, and its value is ; Given a set of user sleeping posture video frames X with multiple categories, the process of training a deep convolutional neural network model is equivalent to determining each neural network operation The process: make For the and Between nodes, The probability distribution of a parameterized neural network operation, Represents the joint distribution, each Cell is composed of operations, for a given user sleeping posture video frame set H and a candidate operation set , calculate an operation distribution Conditional entropy : ; in, is the probability distribution of the data set X, is a set of neural network operation operations; Conditional entropy of all operations in the entire neural network Cell for: ; Calculate the log-likelihood function based on conditional entropy : ; in, ; is a discrete characteristic indicator function, if If the value of is less than the preset threshold, is 1, otherwise, is 0, For nodes and nodes The Euclidean distance between in, ; By maximizing Add maximum entropy constraints to update the category index of the user's sleeping posture ; Category Index Perform analysis to obtain the category corresponding to the user's sleeping posture.

[0031] In some embodiments, the category index The analysis to determine the category corresponding to the user's sleeping posture includes the following steps: Get the category index corresponding to all cells , and class index based on the order of Cell operation Numbering to get category index The number value of The horizontal axis of the rectangular coordinate system with numbered values, category index The vertical axis of the rectangular coordinate system is the vertical axis of the rectangular coordinate system. Mark them as points, connect adjacent points in the rectangular coordinate system to generate a category index curve, draw perpendicular lines from the two ends of the category index curve to the horizontal axis to obtain two start and end line segments, and form a closed figure with the designed category index curve, the two start and end line segments, and the horizontal axis. Calculate the total area of ​​the closed figure; set a constant straight line, obtain the angle at which the constant straight line intersects the category index curve, and record the largest angle as the reference angle; multiply the reference angle by the total area of ​​the closed figure to obtain the category feature value of the user's sleeping posture; The category corresponding to the user's sleeping posture is queried based on the category characteristic value of the user's sleeping posture, wherein each user's sleeping posture category corresponds to a preset category characteristic value interval.

[0032] In some embodiments, obtaining the sleep quality weight parameter corresponding to the sleeping posture category specifically includes the following process: Determine the judgment matrix for each sleeping posture category: Determine the number of factors affecting sleep quality weight parameters , construct a set of weighted influencing factors , ,in, is the subset corresponding to the average heart rate of the user when sleeping in the supine position, is the subset corresponding to the average heart rate when the user sleeps in the left side position, is the subset corresponding to the average heart rate of the user when sleeping in the right side, from the set Take out two subsets in the same level and compare them, and use Represents the ratio of importance, and assigns the corresponding importance according to the preset ratio, combines the importance of each layer to form a judgment matrix; calculates the maximum eigenvalue of the judgment matrix : ; in, It is the matrix obtained by normalizing each column vector of the judgment matrix. The value of is 1,2...m, is a matrix The elements of are added row by row to obtain the vector and then normalized into a matrix. is a matrix Add the elements of column by column to obtain the vector and then normalize the matrix; Calculate the sleep quality weight parameter corresponding to the sleeping posture category : , where b is the order of the judgment matrix of the sleeping posture category.

[0033] In some embodiments, determining the sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value specifically includes the following process: adding the first sleep quality characteristic value and the second sleep quality characteristic value after dimension normalization to obtain the sleep quality parameter of the user's sleep stage.

[0034] Furthermore, generating a sleep quality analysis index for a user's sleep cycle based on the sleep quality parameters of the user's sleep stages specifically includes the following processes: obtaining the sleep quality parameters for all the user's sleep stages, calculating the average of the sleep quality parameters for all the user's sleep stages, and recording the average as the sleep quality analysis index for the user's sleep cycle.

[0035] Furthermore, generating a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle specifically includes the following process: A sleep quality analysis index threshold is loaded, the value of which is set by the system and stored in the system, and it is determined whether the sleep quality analysis index of the user's sleep cycle exceeds the sleep quality analysis index threshold. If so, no warning signal for low sleep quality is generated; otherwise, a warning signal for low sleep quality is generated.

[0036] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0037] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0038] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0039] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is only for some logical functions. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0040] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0041] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A sleep data analysis and management system based on artificial intelligence, characterized in that: The system includes: A sleep cycle segmentation unit is used to obtain brain wave data of the user in the sleeping state, and segment the user's sleep cycle based on the brain wave data to obtain the user's N sleep stages; A sleep monitoring data acquisition unit, configured to acquire sleep monitoring data of N sleep stages, wherein the sleep monitoring data includes body movement monitoring data, pulse monitoring data, respiratory monitoring data, and sleeping posture monitoring video data; a sleep data analysis unit, configured to calculate a first sleep quality characteristic value of a user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiratory monitoring data; determine a sleeping posture category of the user's sleep stage based on the sleeping posture monitoring video data; determine a second sleep quality characteristic value of the user's sleep stage based on the sleeping posture category and a sleep quality weight parameter corresponding to the sleeping posture category; and determine a sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value; The sleep warning management unit is used to generate a sleep quality analysis index of the user's sleep cycle based on the sleep quality parameters of the user's sleep stage, and generate a sleep warning signal based on the sleep quality analysis index of the user's sleep cycle.

2. The sleep data analysis and management system based on artificial intelligence according to claim 1, characterized in that: Segmenting the user's sleep cycle based on brainwave data specifically involves the following steps: Dividing the EEG data into a plurality of EEG segments based on the same time interval, obtaining a first instantaneous average heart rate and a first respiratory rate change trend at a location adjacent to a previous EEG segment, and obtaining a second instantaneous average heart rate and a second respiratory rate change trend at a location adjacent to a next EEG segment, wherein the respiratory rate change trend includes an upward trend and a downward trend; The first instantaneous average heart rate, the first respiratory rate change trend, the second instantaneous average heart rate and the second respiratory rate change trend are recorded as feature combinations, and PCA analysis is performed on the feature combinations to obtain the principal component FI of the feature combinations; Convert FI into a zero-mean unit variance form, and smooth the principal component FI within a small and large frame length based on a smoothing filter, respectively, to obtain the smoothing results SFI and BFI, and determine the offset k of BFI; When SFI>BFI+k, the current brain wave segment is the data of the awake stage or REM stage. The current brain wave segment that is earlier in time sequence is the data of the awake stage, and the current brain wave segment that is later in time sequence is the data of the REM stage. When SFI<BFI+k, the current brain wave segment is the data of the deep sleep stage. When SFI=BFI+k, the current brain wave segment is the data of the light sleep stage.

3. The sleep data analysis and management system based on artificial intelligence according to claim 1, characterized in that: Calculating the first sleep quality characteristic value of the user's sleep stage based on the body movement monitoring data, the pulse monitoring data, and the respiration monitoring data specifically includes the following process: Based on the body motion monitoring data, the number of tossing and turning, the number of limb twitching, and the number of head rotations during the user's sleep stage are obtained, and the sum of the number of tossing and turning, the number of limb twitching, and the number of head rotations is recorded as the body motion representation value A; Based on the pulse monitoring data, the blood oxygen saturation values ​​of each detection point during the user's sleep stage are collected. A rectangular coordinate system is established with the number of detection time points as the horizontal axis and the blood oxygen saturation value as the vertical axis. A blood oxygen saturation value curve is drawn by plotting points. The difference between the user's average blood oxygen saturation value and the lowest blood oxygen saturation value during the sleep stage is calculated. The integral value of the blood oxygen saturation value curve above the straight line of the average blood oxygen saturation value is calculated. The integral value and the difference are multiplied after unifying the dimensions to obtain the product value B; The number of central apneas C and the total duration D of apneas of the user are collected based on the respiratory monitoring data. If the number of central apneas of the user is 0, the user's respiratory quality parameter R is 1; if the number of central apneas of the user is not 0, the user's respiratory quality parameter R is 0; A first sleep quality characteristic value of the user's sleep stage is obtained based on the body movement representation value A, the product value B, the number of times C, the total duration D and the breathing quality parameter R.

4. The artificial intelligence-based sleep data analysis and management system according to claim 3, characterized in that: Obtaining the first sleep quality characteristic value of the user's sleep stage based on the body movement representation value A, the product value B, the number of times C, the total duration D, and the breathing quality parameter R specifically includes the following process: Substitute the body movement representation value A, the product value B, the number of times C, the total duration D, and the breathing quality parameter R into the correlation formula to calculate the first sleep quality characteristic value LMS of the user's sleep stage. The correlation formula is as follows: ； in, 、 and is the weight, whose value is set by the system, The value of is 2.

72.

5. The sleep data analysis and management system based on artificial intelligence according to claim 1, characterized in that: Determining the user's sleep posture category based on sleep posture monitoring video data specifically includes the following process: Extracting user sleeping posture video frames based on sleeping posture monitoring video data; A deep convolutional neural network model is used to realize the judgment of sleeping posture category. The deep convolutional neural network model consists of several cells, each cell contains N1 nodes. Represents the node parameters of each Cell, 、 Represents a node, and its value is ; Given a set of user sleeping posture video frames X with multiple categories, the process of training a deep convolutional neural network model is equivalent to determining each neural network operation The process: make For the and Between nodes, The probability distribution of a parameterized neural network operation, Represents the joint distribution, each Cell is composed of operations, for a given user sleeping posture video frame set H and a candidate operation set , calculate an operation distribution Conditional entropy : ； in, is the probability distribution of the data set X, is a set of neural network operation operations; Conditional entropy of all operations in the entire neural network Cell for: ； Calculate the log-likelihood function based on conditional entropy : ； in, ; is a discrete characteristic indicator function, if If the value of is less than the preset threshold, is 1, otherwise, is 0, For nodes and nodes The Euclidean distance between in, ; By maximizing Add maximum entropy constraints to update the category index of the user's sleeping posture ; Category Index Perform analysis to obtain the category corresponding to the user's sleeping posture.

6. The artificial intelligence-based sleep data analysis and management system according to claim 5, characterized in that: Category Index The analysis to determine the category corresponding to the user's sleeping posture includes the following steps: Get the category index corresponding to all cells , and class index based on the order of Cell operation Numbering to get category index The number value of The horizontal axis of the rectangular coordinate system with numbered values, category index The vertical axis of the rectangular coordinate system is the vertical axis of the rectangular coordinate system. Mark them as points, connect adjacent points in the rectangular coordinate system to generate a category index curve, draw perpendicular lines from the two ends of the category index curve to the horizontal axis to obtain two start and end line segments, and form a closed figure with the designed category index curve, the two start and end line segments, and the horizontal axis. Calculate the total area of ​​the closed figure; set a constant straight line, obtain the angle at which the constant straight line intersects the category index curve, and record the largest angle as the reference angle; multiply the reference angle by the total area of ​​the closed figure to obtain the category feature value of the user's sleeping posture; The category corresponding to the user's sleeping posture is queried based on the category characteristic value of the user's sleeping posture, wherein each user's sleeping posture category corresponds to a preset category characteristic value interval.

7. The artificial intelligence-based sleep data analysis and management system according to claim 1, characterized in that: Obtaining the sleep quality weight parameter corresponding to the sleeping posture category specifically includes the following process: Determine the judgment matrix for each sleeping posture category: Determine the number of factors affecting sleep quality weight parameters , construct a set of weighted influencing factors , ,in, is the subset corresponding to the average heart rate of the user when sleeping in the supine position, is the subset corresponding to the average heart rate when the user sleeps in the left side position, is the subset corresponding to the average heart rate of the user when sleeping in the right side, from the set Take out two subsets in the same level and compare them, and use Represents the ratio of importance, and assigns the corresponding importance according to the preset ratio, combines the importance of each layer to form a judgment matrix; calculates the maximum eigenvalue of the judgment matrix : ; in, It is the matrix obtained by normalizing each column vector of the judgment matrix. The value of is 1,2...m, is a matrix The elements of are added row by row to obtain the vector and then normalized into a matrix. is a matrix Add the elements of column by column to obtain the vector and then normalize the matrix; Calculate the sleep quality weight parameter corresponding to the sleeping posture category : , where b is the order of the judgment matrix of the sleeping posture category.

8. The artificial intelligence-based sleep data analysis and management system according to claim 1, characterized in that: Determining the sleep quality parameter of the user's sleep stage based on the first sleep quality characteristic value and the second sleep quality characteristic value specifically includes the following process: adding the first sleep quality characteristic value and the second sleep quality characteristic value after dimension normalization to obtain the sleep quality parameter of the user's sleep stage.

9. The artificial intelligence-based sleep data analysis and management system according to claim 1, characterized in that: Generating a sleep quality analysis index for a user's sleep cycle based on the sleep quality parameters of the user's sleep stages specifically includes the following processes: obtaining the sleep quality parameters for all the user's sleep stages, calculating the average of the sleep quality parameters for all the user's sleep stages, and recording the average as the sleep quality analysis index for the user's sleep cycle.

10. The artificial intelligence-based sleep data analysis and management system according to claim 1, characterized in that: Generate sleep warning signals based on the sleep quality analysis index of the user's sleep cycle The following processes are included: A sleep quality analysis index threshold is loaded, the value of which is set by the system and stored in the system, and it is determined whether the sleep quality analysis index of the user's sleep cycle exceeds the sleep quality analysis index threshold. If so, no warning signal for low sleep quality is generated; otherwise, a warning signal for low sleep quality is generated.

Images