Sleep quality quantification method, system, device, and medium

Abstract

The application provides a sleep quality quantification method, system, device and medium. The method comprises a multi-dimensional sleep data acquisition module for acquiring multi-dimensional sleep data; a multi-dimensional sleep feature extraction module for extracting features from the sleep data to obtain multi-dimensional sleep features; a multi-dimensional sleep feature vector determination module for inputting the multi-dimensional sleep features into a multi-dimensional sleep feature quantification model to obtain a multi-dimensional sleep feature vector; wherein the multi-dimensional sleep feature quantification model comprises a basic quantification model and an optimized quantification model; and a sleep quality quantification result determination module for calculating a sleep quality quantification result based on the multi-dimensional sleep feature vector. By extracting features from the multi-dimensional sleep data and obtaining a multi-dimensional sleep feature vector corresponding to the multi-dimensional sleep data, the sleep quality is quantified based on the multi-dimensional sleep feature vector, the accuracy of sleep quality quantification is improved, and the corresponding disease is associated, so that accurate risk prompt of early disease can be realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of sleep monitoring, and relates to a sleep quality quantification method, system, device and medium. BACKGROUND

[0002] There is a complex and significant correlation between current sleep data of different dimensions and disease risks. This not only refers to the length of sleep time, but also covers multiple aspects such as sleep structure (such as the proportion of rapid eye movement sleep and non-rapid eye movement sleep), sleep continuity (such as the number of awakenings and the duration of awakenings), sleep timing (such as sleep time and wake-up time), sleep quality subjective experience, and daytime functional status (such as daytime sleepiness and fatigue). However, although these correlations have been initially revealed, how to accurately and individually correlate these scattered and multi-dimensional sleep characteristic data with specific disease risks remains a great challenge.

[0003] Therefore, how to correlate multi-dimensional sleep characteristics with diseases, and how to achieve accurate disease early warning based on multi-dimensional sleep characteristics, has become one of the technical problems to be solved. SUMMARY

[0004] The application provides a sleep quality quantification method, system, device and medium, which is used to improve the accuracy of disease early warning based on multi-dimensional sleep characteristics.

[0005] In a first aspect, the application provides a sleep quality quantification method. The method comprises: a multi-dimensional sleep data acquisition module for acquiring multi-dimensional sleep data, the multi-dimensional sleep data comprising rhythm early or late data, sleep duration data, sleep stability data, sleep debt data and / or work-rest regularity data; a multi-dimensional sleep feature extraction module for extracting features from the sleep data to obtain multi-dimensional sleep features; a multi-dimensional sleep feature vector determination module for inputting the multi-dimensional sleep features into a multi-dimensional sleep feature quantification model to obtain a multi-dimensional sleep feature vector; wherein the multi-dimensional sleep feature quantification model comprises a basic quantification model and an optimized quantification model; and a sleep quality quantification result determination module for calculating a sleep quality quantification result based on the multi-dimensional sleep feature vector.

[0006] In the sleep quality quantification method of the application, the sleep data is subjected to feature extraction to obtain multi-dimensional sleep features, a multi-dimensional sleep feature vector corresponding to the multi-dimensional sleep features is obtained based on a multi-dimensional sleep feature quantification model, an association matrix of the multi-dimensional sleep feature vector and disease risks is established, an association model is established based on the association matrix and disease categories, and a disease early warning result is output based on the association model, thereby greatly improving the accuracy of disease early warning based on multi-dimensional sleep features.

[0007] In an implementation form of the first aspect, the multi-dimensional sleep feature is input into a base quantization model to obtain a score quantization effect of each dimension of the sleep feature, including: determining, by the multi-dimensional sleep feature quantization model, a score quantization effect of each dimension of the sleep feature according to a base quantization model. The expression of the score quantization effect of each dimension of the sleep feature is: wherein, represents the score quantization effect of each dimension of the sleep feature, represents the original score of the i th data in the j th dimension of the sleep data, represents the weight of the i th data in the j th dimension of the sleep data, represents the weight of the i th data in the j th dimension of the sleep data, represents the number of elements of the j th dimension of the sleep data.

[0008] In an implementation form of the first aspect, the multi-dimensional sleep feature is input into a base quantization model to obtain a score quantization effect of each dimension of the sleep feature, including: determining, by the multi-dimensional sleep feature quantization model, a score quantization effect of each dimension of the sleep feature according to a base quantization model.

[0009] The expression of the optimal quantization effect is: wherein, represents the optimal quantization effect, represents whether the i th optimization condition exists, when the i th optimization condition does not exist, ; otherwise, , represents the optimization objective function, represents the weight set of all problems in the j th dimension; if the optimal quantization effect is less than a preset quantization effect, the weight of at least one parameter of the clustering / classification model, the subjective reference correlation, the physiological index correlation and the classical scale correlation is updated to update the optimal quantization effect; if the optimal quantization effect is greater than or equal to the preset quantization effect, the optimal quantization effect is standardized to obtain the multi-dimensional sleep feature vector output by the multi-dimensional sleep feature quantization model. In an implementation form of the first aspect, the method further includes: obtaining a first quantization effect of the score quantization effect based on a clustering / classification model; obtaining a second quantization effect of the score quantization effect based on a subjective reference correlation; obtaining a third quantization effect of the score quantization effect based on a physiological index correlation; obtaining a fourth quantization effect of the optimal quantization effect based on a classical scale correlation; and performing weighted processing on the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect to obtain the optimal quantization effect.

[0010] In an implementation form of the first aspect, the method further includes: obtaining a first quantization effect of the score quantization effect based on a clustering / classification model; obtaining a second quantization effect of the score quantization effect based on a subjective reference correlation; obtaining a third quantization effect of the score quantization effect based on a physiological index correlation; obtaining a fourth quantization effect of the optimal quantization effect based on a classical scale correlation; and performing weighted processing on the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect to obtain the optimal quantization effect.

[0011] ​​​​​​​In one implementation of the first aspect, the step of extracting features from the sleep data to obtain multidimensional sleep features includes: obtaining circadian rhythm features based on circadian rhythm data; obtaining sleep duration features based on sleep duration data; obtaining sleep stability features based on sleep stability data; obtaining sleep debt features based on sleep debt data; and obtaining sleep pattern features based on sleep schedule data. The multidimensional sleep features include circadian rhythm features, sleep duration features, sleep stability features, sleep debt features, and sleep schedule features.

[0012] In one implementation of the first aspect, the method further includes: using a multidimensional classification model to determine the sleep feature score distribution of each dimension in the multidimensional sleep feature vector; the multidimensional classification model determining a classification threshold based on the distribution characteristics of the sleep feature score distribution of each dimension; and inputting the optimal quantization effect corresponding to the multidimensional sleep feature vector into the multidimensional classification model to obtain the sleep quality classification label output by the multidimensional classification model.

[0013] In one implementation of the first aspect, the step of calculating the quantitative result of sleep quality based on the multidimensional sleep feature vector includes: determining the correlation matrix between the multidimensional sleep feature vector and disease risk; if the matrix elements in the correlation matrix If it is greater than the preset threshold, then the first Increased sleep characteristics and diseases Risks;

[0014] If the matrix elements in the correlation matrix If the value is less than or equal to the preset threshold, then the first... Sleep characteristics reduce disease The risks.

[0015] In one implementation of the first aspect, determining the correlation matrix between the multidimensional sleep feature vector and disease risk includes: extracting the ratio of each sleep feature dimension to its corresponding disease risk based on a database; and then... The logarithmic ratio of the sleep feature is logarithmically processed to obtain the corresponding logarithmic ratio; the variance value corresponding to the logarithmic ratio is determined; the random effect weight is determined based on the variance value; the pooled effect size is determined based on the random effect weight and the logarithmic ratio; and the correlation matrix is ​​determined based on the pooled effect size.

[0016] In a second aspect, the present application provides a sleep quality quantification system, comprising: a multi-dimensional sleep feature acquisition module, configured to perform feature extraction on the sleep data to obtain multi-dimensional sleep features; a multi-dimensional sleep feature vector determination module, configured to input the multi-dimensional sleep features into a multi-dimensional sleep feature quantification model to obtain a multi-dimensional sleep feature vector; an association matrix establishment module, configured to establish an association matrix between the multi-dimensional sleep feature vector and disease risks; an association model establishment module, configured to establish an association model based on the association matrix and disease categories; and a disease early warning result determination module, configured to output a disease early warning result based on the association model.

[0017] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the sleep quality quantification method of any one of the first aspect of the present application.

[0018] In a fourth aspect, the present application provides an electronic device, comprising: a memory storing a computer program; and a processor communicatively connected to the memory, configured to invoke the computer program to execute the sleep quality quantification method of any one of the first aspect of the present application.

[0019] As described above, the sleep quality quantification method, system, device and medium of the present application have the following beneficial effects: in the sleep quality quantification method of the present application, multi-dimensional sleep features are obtained by performing feature extraction on the sleep data, a multi-dimensional sleep feature vector corresponding to the multi-dimensional sleep features is obtained based on a multi-dimensional sleep feature quantification model, an association matrix between the multi-dimensional sleep feature vector and disease risks is established, an association model is established based on the association matrix and disease categories, and a disease early warning result is output based on the association model, thereby greatly improving the accuracy of disease early warning based on multi-dimensional sleep features.

[0020] In the present application, different sleep features corresponding to different sleep data are obtained by means of deep neural networks and the like, thereby providing sufficient sleep feature basis for subsequent establishment of an association matrix between the multi-dimensional sleep feature vector and disease risks.

[0021] The application determines the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect by a clustering / classification model, a subjective reference correlation, a physiological index correlation and a classical scale correlation respectively when determining the optimal quantization effect, and performs weighted processing on the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect to obtain the optimal quantization effect, instead of directly determining the optimal quantization effect based on a single method. In the method, the multi-dimensional weighted quantization method improves the comprehensiveness, robustness, adaptability and objectivity of quantization, and also enhances the explainability and decision support capability of the optimal quantization effect. The method is especially suitable for comprehensive quantization problems in complex scenarios, and the quantization result is more accurate.

[0022] In the application, the classification threshold corresponding to the distribution of different sleep feature scores can be determined respectively, and the classification threshold can be adaptively determined, which can significantly improve the accuracy of the sleep quality classification label of each sleep feature vector.

[0023] In the application, the association matrix of the multi-dimensional sleep feature vector and the disease risk is determined, and the corresponding disease risk is determined based on the relationship between each matrix element in the association matrix and the preset threshold. The preset threshold corresponding to each matrix element can be the same or different, and the specific value of the preset threshold corresponding to each matrix element can be determined based on the actual situation. The method is suitable for determining multi-dimensional and multi-level disease risks, and can significantly improve the accuracy of determining disease risks.

[0024] In the application, the log ratio of each dimension sleep feature and the corresponding disease risk is processed, the variance value corresponding to the log ratio is determined, the random effect weight is determined based on the variance value, the merging effect quantity is determined based on the random effect weight and the log ratio, and the association matrix is determined based on the merging effect quantity. By performing the above multiple data processing on the ratio of each dimension sleep feature and the corresponding disease risk, the accuracy of the association matrix is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1A A hardware application scenario diagram of a sleep quality quantization method provided by an embodiment of the application is shown.

[0026] Figure 1B A flowchart of a sleep quality quantization method provided by an embodiment of the application is shown.

[0027] Figure 2 A flowchart of feature extraction on sleep data to obtain multi-dimensional sleep features provided by an embodiment of the application is shown.

[0028] Figure 3 A flowchart of determining an optimal quantization effect provided by an embodiment of the application is shown.

[0029] Figure 4 A flowchart of determining a sleep quality classification label is shown as an embodiment of the present application.

[0030] Figure 5 A flowchart of determining a sleep quality quantification result is shown as an embodiment of the present application.

[0031] Figure 6 A flowchart of determining a correlation matrix between a multi-dimensional sleep feature vector and a disease risk is shown as an embodiment of the present application.

[0032] Figure 7 A flowchart of another sleep quality quantification method is shown as an embodiment of the present application.

[0033] Figure 8 A structural diagram of a sleep quality quantification system is shown as an embodiment of the present application.

[0034] Figure 9 A structural diagram of an electronic device is shown as an embodiment of the present application. DETAILED DESCRIPTION

[0035] The present application will be described in detail by the following specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the disclosure. The present application can also be implemented or applied by other different specific embodiments, and various modifications or changes can be made to the details in the specification without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0036] It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present application in a schematic manner, and only the components related to the present application are shown in the diagrams, not the number, shape and size of the components when actually implemented. The actual implementation of each component may be a random change in type, number and proportion, and the layout type of the components may also be more complex.

[0037] The following embodiments of the present application provide sleep quality quantification methods, systems, devices and media, including but not limited to the hardware application scenarios listed in the embodiments.

[0038] As Figure 1AAs shown in the diagram, this embodiment provides a hardware application scenario diagram for a sleep quality quantification method, specifically including: a user to be tested, a sleep data acquisition device, and an electronic device. The sleep data acquisition device is used to acquire the sleep data of the user to be tested and send the sleep data to the electronic device. The electronic device processes the received sleep data to obtain multidimensional sleep features, inputs the multidimensional sleep features into a multidimensional sleep feature quantification model to obtain a multidimensional sleep feature vector, and calculates the sleep quality quantification result based on the multidimensional sleep feature vector.

[0039] The technical solutions in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0040] like Figure 1B As shown in the flowchart, this application provides a method for quantifying sleep quality. Figure 1B As shown, the sleep quality quantification method provided in this application includes the following steps S11 to S14.

[0041] S11, Multidimensional sleep data acquisition module, used to extract features from the sleep data to obtain multidimensional sleep features.

[0042] Among them, multidimensional sleep characteristics include circadian rhythm characteristics, sleep duration characteristics, sleep stability characteristics, sleep debt characteristics, and sleep-wake cycle characteristics.

[0043] S12, Multidimensional sleep feature extraction module, is used to extract features from the sleep data to obtain multidimensional sleep features.

[0044] S13, Multidimensional sleep feature vector determination module, is used to input the multidimensional sleep features into the multidimensional sleep feature quantization model to obtain the multidimensional sleep feature vector.

[0045] The multidimensional sleep feature quantification model includes a basic quantification model and an optimized quantification model.

[0046] For example, multidimensional sleep feature vectors can be used express.

[0047] In some embodiments, the multidimensional sleep features are input into a basic quantization model to obtain the score quantization effect of each dimension of the sleep feature, including: the multidimensional sleep feature quantization model determines the score quantization effect of the first dimension of the sleep feature. The expression for the scoring quantification effect of the sleep feature is: in, Indicates the first The scoring and quantification effect of sleep characteristics. Indicates the first The first in the sleep data The raw scores of each data point Indicates the first The first in the sleep data The weight of each data point Indicates the first The number of elements in the sleep data.

[0048] In some embodiments, multiple score results are input into an optimized quantization model to obtain a multidimensional sleep feature vector, including:

[0049] The expression for determining the optimal quantization effect is: in, This indicates the optimal quantization effect. Indicates whether the first [element] exists. The optimization condition is not met when the first condition is not met. When optimizing the conditions of a task, ;otherwise, , This represents the objective function to be optimized. Indicates the first The set of weights for all questions under the dimension; if the optimal quantization effect is less than the preset quantization effect, the weights of at least one parameter among the clustering / classification model, the subjective reference relevance, the physiological index relevance, and the classic scale relevance are adjusted to update the optimal quantization effect; if the optimal quantization effect is greater than or equal to the preset quantization effect, the optimal quantization effect is standardized to obtain the multidimensional sleep feature vector output by the multidimensional sleep feature quantization model.

[0050] S14, Sleep quality quantification result determination module, used to calculate and obtain sleep quality quantification result based on the multidimensional sleep feature vector.

[0051] Among them, the quantitative results of sleep quality can be expressed as scores and / or classifications.

[0052] For example, vectors can be used Predicting diseases probability .

[0053] in, The covariate vector represents a set of confounding factors that require statistical adjustment. These can be used as predictor variables to isolate the independent effects of sleep characteristics on disease risk, such as age, sex, race, smoking, alcohol consumption, exercise habits, education level, and income. Their core function is to control for confounding bias and ensure that the sleep characteristic effects estimated by the model are accurately represented. It is closer to the true causal relationship.

[0054] For example, sleep quality quantification results can be obtained based on a logistic regression model with covariates. Specifically:

[0055] (1) Feature Standardization where, is the standardized score of the i-th sleep feature, quantifies the effect of the i-th sleep feature, is the mean of the i-th sleep feature, is the standard deviation of the i-th sleep feature, quantifies the effect of the i-th sleep feature.

[0056] (2) Model Construction where, is the intercept from the maximum likelihood estimation, is the coefficient of the i-th sleep feature (comparable after standardization), is the coefficient of the covariate (e.g., age, gender), is the covariate vector, , is the number of elements in the covariate vector . (3) Risk Probability Prediction where,

[0057] is the intercept from the maximum likelihood estimation, is the coefficient of the i-th sleep feature (comparable after standardization), is the coefficient of the covariate (e.g., age, gender), is the covariate vector, , is the number of elements in the covariate vector . For example, when predicting the risk of coronary heart disease for a user based on the above logistic regression model, if the input is: (where, Circadian rhythm 54, sleep duration 12, etc.), the output is: If the preset threshold of coronary heart disease is 0.3, the user is judged to be high risk.

[0058]

[0059] ​​​​​​​The embodiment of the present application provides a sleep quality quantification method, which comprises a multi-dimensional sleep data acquisition module, which is used for acquiring multi-dimensional sleep data, wherein the multi-dimensional sleep data comprises rhythm early and late data, sleep duration data, sleep stability data, sleep debt data and / or work and rest regularity data; a multi-dimensional sleep feature extraction module, which is used for extracting features of the sleep data to obtain multi-dimensional sleep features; a multi-dimensional sleep feature vector determination module, which is used for inputting the multi-dimensional sleep features into a multi-dimensional sleep feature quantification model to obtain a multi-dimensional sleep feature vector; wherein the multi-dimensional sleep feature quantification model comprises a basic quantification model and an optimized quantification model; and a sleep quality quantification result determination module, which is used for calculating a sleep quality quantification result based on the multi-dimensional sleep feature vector. By extracting features of multi-dimensional sleep data and obtaining a multi-dimensional sleep feature vector based on the corresponding multi-dimensional sleep data, the sleep quality is quantified based on the multi-dimensional sleep feature vector, the accuracy of sleep quality quantification is improved, and the sleep quality is associated with corresponding diseases, so that accurate risk prompt of early diseases can be realized.

[0060] As shown in Figure 2 The embodiment of the present application provides a flowchart for extracting features of sleep data to obtain multi-dimensional sleep features, as shown in Figure 2 The embodiment of the present application provides a method for extracting features of sleep data to obtain multi-dimensional sleep features, which comprises the following steps S21-S25.

[0061] S21, rhythm early and late features are acquired based on rhythm early and late data.

[0062] For example, the rhythm early and late data comprises wake-up or sleep time on weekdays, wake-up or sleep time on rest days, wake-up time, best state time, and heart rate, blood pressure, core body temperature and the like in different time periods.

[0063] For example, the rhythm early and late features are acquired by acquiring heart rate, blood pressure, core body temperature and the like of a to-be-detected person in different time periods and based on the change rule of the heart rate, blood pressure and core body temperature of the to-be-detected person in different time periods.

[0064] For another example, the rhythm early and late data can be input into a deep neural network (DNN) to obtain rhythm early and late features output by the deep neural network.

[0065] Specifically, the deep neural network can efficiently identify the time sequence pattern of gene expression by training a large amount of rhythm early and late data, and then extract rhythm early and late features.

[0066] It should be noted that the rhythm early-late feature based on the rhythm early-late data is only used for illustrative description, and the rhythm early-late feature corresponding to the rhythm early-late data can also be determined by any other suitable method in actual application, which is not limited in the present application.

[0067] S22, obtaining a sleep duration feature based on the sleep duration data.

[0068] For example, the sleep duration feature includes short sleep time, medium sleep time and long sleep time.

[0069] For example, if the sleep duration data is less than the first sleep duration, the sleep duration feature is short sleep time; if the sleep duration data is greater than or equal to the first sleep duration and less than the second sleep duration, the sleep duration feature is medium sleep time; if the sleep duration data is greater than or equal to the second sleep duration, the sleep duration feature is long sleep time.

[0070] It should be noted that the specific values of the first sleep duration and the second sleep duration can be determined based on specific application scenarios, which are not limited in the present application.

[0071] It should be noted that the sleep duration feature and the determination method of the sleep duration feature are only used for illustrative description, and any other suitable sleep duration data can also be included in actual application, which is not limited in the present application.

[0072] S23, obtaining a sleep stability feature based on the sleep stability data.

[0073] For example, the sleep stability data includes the time length from starting to sleep to entering deep sleep, the frequency of waking up at night, the frequency of waking up at midnight and being difficult to sleep, the frequency of nightmares, etc.

[0074] Specifically, the sleep stability feature includes sleep stability and sleep instability.

[0075] For example, if the time length from starting to sleep to entering deep sleep is less than the preset first time length, the frequency of waking up at night is less than the preset first frequency, the frequency of waking up at midnight and being difficult to sleep is less than the preset second frequency, and the frequency of nightmares is less than the preset third frequency, the corresponding sleep stability feature is sleep stability. Conversely, if the time length from starting to sleep to entering deep sleep is greater than or equal to the preset first time length, or the frequency of waking up at night is greater than or equal to the preset first frequency, or the frequency of waking up at midnight and being difficult to sleep is greater than or equal to the preset second frequency, or the frequency of nightmares is greater than or equal to the preset third frequency, the corresponding sleep stability feature is sleep instability.

[0076] It should be noted that the above-mentioned way of obtaining the sleep stability feature based on the sleep stability data is only used for illustrative description, and in actual application, other any suitable method can be selected to determine the sleep stability feature corresponding to the sleep stability data according to the specific application scenario, and the present application does not make any limitation.

[0077] S24, obtaining a sleep debt feature based on sleep debt data.

[0078] Specifically, the sleep debt data includes the difference between the ideal sleep and the actual sleep duration, the probability of daytime sleepiness, etc.

[0079] The sleep debt feature includes sleep debt and sleep debt-free.

[0080] For example, if the difference between the ideal sleep and the actual sleep duration is greater than the preset sleep difference, and the probability of daytime sleepiness is greater than the preset probability, then the sleep debt feature corresponding to the sleep debt data is sleep debt. On the contrary, if the difference between the ideal sleep and the actual sleep duration is less than or equal to the preset sleep difference, or the probability of daytime sleepiness is less than or equal to the preset probability, then the sleep debt feature corresponding to the sleep debt data is sleep debt-free.

[0081] S25, obtaining a sleep-wake regularity feature based on sleep-wake regularity data.

[0082] The sleep-wake regularity data includes the wake-up and sleep time on weekdays, the wake-up and sleep time on weekends, the number of midnight awakenings, etc.

[0083] The sleep-wake regularity feature includes sleep-wake regularity and sleep-wake irregularity.

[0084] For example, if the sleep time every night is within the first preset time range, the wake-up time in the morning is within the second preset time range, and the number of midnight awakenings is within the preset number range, then the sleep-wake regularity feature corresponding to the sleep-wake regularity data is sleep-wake regularity. On the contrary, if the sleep time every night is not within the first preset time range, or the wake-up time in the morning is not within the second preset time range, or the number of midnight awakenings is within the preset number range, then the sleep-wake regularity feature corresponding to the sleep-wake regularity data is sleep-wake irregularity.

[0085] It should be noted that the present application does not make any limitation on the specific values of the above-mentioned first preset time range, second preset time range and preset number range.

[0086] The sleep data includes rhythm early or late data, sleep duration data, sleep stability data, sleep debt data, and sleep-wake regularity data, and the multi-dimensional sleep feature includes rhythm early or late feature, sleep duration feature, sleep stability feature, sleep debt feature, and sleep-wake regularity feature.

[0087] This application provides a method for extracting features from sleep data to obtain multidimensional sleep features. In this method, different sleep features corresponding to different sleep data are obtained through deep neural networks and other means, providing a sufficient sleep feature foundation for subsequently establishing a correlation matrix between multidimensional sleep feature vectors and disease risk.

[0088] like Figure 3 As shown in the figure, this application provides a flowchart for determining the optimal quantization effect, as follows: Figure 3 As shown, the method for determining the optimal quantization effect provided in this application embodiment includes the following steps S31 to S35.

[0089] S31, Based on the clustering / classification model, obtain the first quantification effect of the score quantification effect.

[0090] Clustering / classification models include SVM, K-means, etc.

[0091] For example, the expression for the first quantification effect of the score quantification effect obtained based on the clustering / classification model is: in, This indicates the first quantification effect. Indicates the first The scoring and quantification effect of sleep characteristics. Indicates the distance between sample groups. This represents the distance within a sample group, and this quantification effect maximizes the classification effect of this feature in the population.

[0092] S32, based on subjective reference relevance, obtain the second quantification effect of the score quantification effect.

[0093] For example, the expression for the second quantification effect of the score quantification effect based on subjective reference relevance is as follows: in, This indicates the second quantification effect. This represents the Pearson correlation coefficient. This represents the subject's subjective evaluation score of a certain dimension of sleep characteristics. For example: "How would you rate the regularity of your sleep over the past month: 1. Extremely regular; 2. Somewhat regular; 3. Uncertain; 4. Somewhat irregular; 5. Extremely irregular."

[0094] S33, based on the correlation of physiological indicators, the third quantitative effect of the score quantification effect is obtained.

[0095] For example, the expression for the third quantitative effect of the score quantification effect obtained based on the correlation of physiological indicators is as follows: in, This indicates the third quantification effect. denotes Pearson correlation coefficient, denotes physiological indicators measured by devices (e.g. actiwatch). For example, Pearson correlation coefficient is calculated between the average number of awakenings in the recent one month measured by actiwatch and the third feature, sleep stability.

[0096] S34, the fourth quantization effect based on the classical scale correlation degree is obtained.

[0097] For example, the expression of the fourth quantization effect based on the classical scale correlation degree is: wherein, denotes the fourth quantization effect, denotes Pearson correlation coefficient, denotes the score based on the classical single-dimension sleep evaluation scale. For example, Pearson correlation coefficient is calculated between the score of the Morning and Evening Questionnaire-19 (MEQ-19) scale and the second feature, chronotype.

[0098] S35, the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect are weighted to obtain the optimal quantization effect.

[0099] In some embodiments, the first quantization effect, the second quantization effect, the third quantization effect and the fourth quantization effect are weighted to obtain the weight under the optimal quantization effect: wherein, denotes the optimal quantization effect, denotes whether the first item optimization condition, when the first item optimization task condition is not met, ; otherwise, , denotes the optimization objective function, denotes the weight set of all problems in the first dimension.

[0100] The embodiment of the present application provides a method for determining an optimal quantification effect, in which a first quantification effect, a second quantification effect, a third quantification effect and a fourth quantification effect are obtained through a clustering / classification model, a subjective reference correlation degree, a physiological index correlation degree and a classical scale correlation degree respectively, and the first quantification effect, the second quantification effect, the third quantification effect and the fourth quantification effect are subjected to weighted processing to obtain the optimal quantification effect, instead of directly determining the optimal quantification effect based on a single method, the multi-dimensional weighted quantification method in the method improves the comprehensiveness, robustness, adaptability and objectivity of quantification, and also enhances the explainability and decision support capability of the optimal quantification effect. The method is especially suitable for comprehensive quantification problems in complex scenarios, and the quantification result is more accurate.

[0101] As shown in Figure 4 The embodiment of the present application provides a flowchart for determining a sleep quality classification label, as shown in Figure 4 The method for determining a sleep quality classification label provided by the embodiment of the present application includes the following steps S41 to S43.

[0102] S41, determining a score distribution of each dimension of the multi-dimensional sleep feature vector by using a multi-dimensional typing model.

[0103] The multi-dimensional typing model includes a rhythm-early-late typing model, a sleep duration typing model, a sleep stability typing model, a sleep debt typing model and a work-rest regularity typing model.

[0104] S42, determining a classification threshold based on a distribution feature of the score distribution of each dimension of the multi-dimensional sleep feature vector by using the multi-dimensional typing model.

[0105] Specifically, if the score distribution of one dimension of the multi-dimensional sleep feature vector is unimodal or uniform distribution, the classification threshold of the optimal quantification effect of the sleep data is determined based on the score ratio.

[0106] For example, the score ratio can be 0.5, 0.6 or any suitable value, and the present application does not limit the specific value of the score ratio.

[0107] For example, if the score distribution of the rhythm-early-late sleep feature is unimodal or uniform distribution, and the score ratio is 0.5, the optimal quantification effect of the rhythm-early-late sleep data between 0 and 0.5 is determined as “early type”, and the optimal quantification effect of the rhythm-early-late sleep data between 0.5 and 1 is determined as “late type”.

[0108] It should be noted that the determination method of the classification threshold corresponding to the sleep duration feature score distribution, the sleep stability feature score distribution, and the sleep debt feature score distribution in the multi-dimensional sleep feature is similar to the determination method of the classification threshold corresponding to the rhythm early or late vector, and the present application will not repeat it.

[0109] Specifically, if the multi-dimensional sleep feature vector is a multi-peak distribution, the classification threshold of the sleep data is determined based on the peak and valley of the multi-peak distribution.

[0110] S43, input the optimal quantization effect corresponding to the multi-dimensional sleep feature vector into the multi-dimensional typing model to obtain a sleep quality classification label output by the multi-dimensional typing model.

[0111] The embodiment of the present application provides a method for determining a sleep quality classification label, in which each dimensional sleep feature score distribution in a multi-dimensional sleep feature vector is determined, a classification threshold is determined based on the distribution characteristics of each dimensional sleep feature score distribution, the optimal quantization effect corresponding to the multi-dimensional sleep feature vector is input into the multi-dimensional typing model, and a sleep quality classification label output by the multi-dimensional typing model is obtained; different sleep feature score distributions are fully considered, and the classification threshold corresponding to each sleep feature score distribution is determined, which can be adaptively determined, and the accuracy of the sleep quality classification label of each sleep feature vector can be significantly improved.

[0112] As shown in Figure 5 , the embodiment of the present application provides a flowchart for determining a sleep quality quantization result, as shown in Figure 5 , the method for determining a sleep quality quantization result provided by the embodiment of the present application includes the following steps S51 to S53.

[0113] S51, determine the association matrix of the multi-dimensional sleep feature vector and the disease risk.

[0114] S52, if the matrix element in the association matrix is greater than a preset threshold, the first dimensional sleep feature increases the risk of the disease. For example, the preset threshold can be set to 1, 0.8, or other suitable values.

[0115] It should be noted that the specific values of the preset threshold listed above are only used for illustrative purposes, and other arbitrary suitable values can also be selected as the preset threshold in actual applications, and the present application does not limit this.

[0116] It should be noted that the preset threshold corresponding to each matrix element in the association matrix can be the same or different.

[0117] It should be noted that the preset threshold corresponding to each matrix element in the association matrix can be the same or different.

[0118] ​S53, if the matrix elements in the correlation matrix If the value is less than or equal to the preset threshold, then the first... Sleep characteristics reduce disease The risks.

[0119] This application provides a method for determining the quantitative results of sleep quality. In this method, a correlation matrix between a multidimensional sleep feature vector and disease risk is determined, and the corresponding disease risk is determined based on the relationship between each matrix element and a preset threshold. The preset thresholds corresponding to each matrix element can be the same or different, and the specific values ​​of the preset thresholds corresponding to each matrix element can be determined based on the actual situation. This method is suitable for determining multidimensional and multi-level disease risks and can significantly improve the accuracy of determining disease risks.

[0120] like Figure 6 As shown in the figure, this application embodiment provides a flowchart for determining the correlation matrix between multidimensional sleep feature vectors and disease risk, as follows: Figure 6 As shown in the embodiments of this application, the method for determining the correlation matrix between multidimensional sleep feature vectors and disease risk includes the following steps S61 to S66.

[0121] S61, based on the database, extract the ratio of each dimension of sleep feature to the corresponding disease risk.

[0122] Among them, the odds ratio of disease risk can be used OR It means, and OR This represents the morbidity rate in the exposed group versus the morbidity rate in the control group.

[0123] S62, for the first The logarithmic ratio of the ratios of the sleep characteristics is obtained by taking the logarithm of the ratios.

[0124] For example, for the first The expression for the ratio of the sleep characteristics to the logarithm is as follows: in, Indicates the first The first disease study Log-ratio ratio of sleep characteristics Indicates the first The first disease study The ratio of sleep characteristics.

[0125] in, , This indicates the total number of disease studies.

[0126] For example, if the third study reports "short sleep duration vs. normal sleep"... OR =1.85, then: ,in, =1 indicates sleep duration.

[0127] S63, determine the variance value corresponding to the logarithmic ratio.

[0128] For example, the expression for determining the variance value corresponding to the logarithmic ratio is: in, , They represent the first The first study Sleep characteristics OR The lower and upper limits of the 95% confidence interval for the value; 3.92 = 2 × 1.96 (97.5% quantile of the standard normal distribution × 2).

[0129] S64, determine the random effect weights based on the variance value.

[0130] For example, the expression for determining the weights of random effects based on the variance value is as follows: in, Indicates the first Heterogeneous variance among studies of sleep characteristics (e.g., the DerSimonian-Laird estimator).

[0131] S65, determine the pooled effect size based on the random effect weights and the log ratio.

[0132] For example, the expression for determining the pooled effect size based on the random effect weights and the log ratio is as follows: in, This indicates the corresponding disease. Calculation of the combined effect size of sleep characteristics.

[0133] S66, Determine the correlation matrix based on the combined effect size.

[0134] For example, for M The correlation matrix between the disease and the 5-dimensional sleep feature vector is as follows:

[0135] This application provides a method for determining the correlation matrix between multidimensional sleep feature vectors and disease risk. In this method, the variance value corresponding to the logarithmic ratio of each sleep feature to its corresponding disease risk is determined by performing logarithmic processing on the ratio; random effect weights are determined based on the variance value; a pooled effect size is determined based on the random effect weights and the logarithmic ratio; and the correlation matrix is ​​determined based on the pooled effect size. By performing the above-mentioned multiple data processing steps on the ratio of each sleep feature to its corresponding disease risk, the accuracy of the correlation matrix is ​​greatly improved.

[0136] The protection scope of the sleep quality quantification method described in the embodiments of the present application is not limited to the execution order of the steps listed in the embodiments, and any scheme realized by increasing, decreasing or replacing the steps of the prior art according to the principles of the present application is included in the protection scope of the present application.

[0137] Please refer to Figure 7 , Figure 7 the flowchart of another sleep quality quantification method provided by an embodiment of the present application, Figure 7 each step in the flowchart has been described in detail in the above Figures 1B to 6 , and the present application will not repeat the description here.

[0138] Please refer to Figure 8 In an embodiment, the sleep quality quantification system 80 of the present application includes a multi-dimensional sleep feature acquisition module 81, a multi-dimensional sleep feature vector determination module 82, an association matrix establishment module 83, an association model establishment module 84, and a disease early warning result determination module 85.

[0139] The multi-dimensional sleep feature acquisition module 81 is configured to extract features from the sleep data and obtain multi-dimensional sleep features.

[0140] The multi-dimensional sleep feature vector determination module 82 is configured to input the multi-dimensional sleep features into a multi-dimensional sleep feature quantification model to obtain a multi-dimensional sleep feature vector.

[0141] The association matrix establishment module 83 is configured to establish an association matrix between the multi-dimensional sleep feature vector and disease risk.

[0142] The association model establishment module 84 is configured to establish an association model based on the association matrix and disease categories.

[0143] The disease early warning result determination module 85 is configured to output a disease early warning result based on the association model.

[0144] The structure and principles of the multi-dimensional sleep feature acquisition module 81, the multi-dimensional sleep feature vector determination module 82, the association matrix establishment module 83, the association model establishment module 84, and the disease early warning result determination module 85 correspond one-to-one to the steps in the sleep quality quantification method described above, and therefore will not be repeated here.

[0145] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, or methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or modules or units may be electrical, mechanical, or other forms.

[0146] The modules / units described as separate components may or may not be physically separate. The components shown as modules / units may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules / units can be selected to achieve the objectives of the embodiments of this application, depending on actual needs. For example, the functional modules / units in the various embodiments of this application may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0147] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art 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.

[0148] This application also provides an electronic device. Figure 9 The diagram shown is a structural schematic of an electronic device 90 in one embodiment of this application. The sleep quality quantification method provided in this embodiment can be applied to… Figure 9 The electronic device shown is 90, but it is not limited to this. For example... Figure 9 As shown, the electronic device 90 includes a processor 91, a memory, a system bus 93, and a network interface 95. The memory may include a non-volatile storage medium 92 and internal memory 94.

[0149] The non-volatile storage medium 92 can store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the sleep quality quantification methods provided in the embodiments of this application.

[0150] The processor is configured to provide computing and control capabilities to support the operation of the entire computer device.

[0151] The internal storage 94 provides an environment for the running of a computer program in a non-volatile storage medium, which, when executed by the processor, can cause the processor to execute any one of the sleep quality quantification methods provided by the embodiments of the present application.

[0152] The network interface 95 is configured to perform network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that, Figure 9 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0153] It should be understood that the processor 91 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0154] The electronic device 90 of the embodiments of the present application can be applied to terminal devices such as tablet computers, notebook computers, mobile phones, supercomputers, smart wearable devices, etc., and can also be applied to databases, servers and terminal artificial intelligence-based service response systems. The specific type of electronic device is not limited by the embodiments of the present application.

[0155] For example, an electronic device can be a station (STATION, STA) in a WLAN, can be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a handheld device having wireless communication function, a computing device or other processing device connected to a wireless modem, a computer, a laptop, a handheld communication device, a handheld computing device, and / or other devices used for communication over a wireless system and next generation communication system, for example, a mobile terminal in a 5G network, a mobile terminal in a future evolved Public Land Mobile Network (PLMN), or a mobile terminal in a future evolved Non-Terrestrial Network (NTN), etc.

[0156] The embodiments of the present application further provide a computer readable storage medium. Those skilled in the art can understand that all or part of the steps of the methods described above can be instructed by a program to complete the processor, and the program can be stored in a computer readable storage medium. The storage medium is a non-transitory medium, for example, a random access memory, a read-only memory, a flash memory, a hard disk, a solid state disk, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available medium sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a digital video disc (DVD)), or a semiconductor medium (for example, a solid state disk (SSD)) and the like.

[0157] The embodiments of the present application can also provide a computer program product, which includes one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in the embodiments of the present application are generated. 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 or data center to another website, computer or data center through wired (for example, coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) mode.

[0158] The computer program product is executed by a computer, and the computer executes the method of the method embodiment. The computer program product can be a software installation package. When the method is needed, the computer program product can be downloaded and executed on the computer.

[0159] The descriptions of the corresponding processes or structures of the various drawings are each focused on. The parts not described in detail in a certain process or structure can be referred to the related description of other processes or structures.

[0160] The above embodiments only exemplarily illustrate the principles and effects of the present application, and are not used to limit the present application. Any person skilled in the art can modify or change the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes completed by those skilled in the art without departing from the spirit and technical idea of the present application should be covered by the claims of the present application.

Claims

1. A method for quantifying sleep quality, characterized in that, The method includes: A multidimensional sleep data acquisition module is used to acquire multidimensional sleep data, which includes circadian rhythm data, sleep duration data, sleep stability data, sleep debt data, and / or sleep pattern data. A multidimensional sleep feature extraction module is used to extract features from the sleep data to obtain multidimensional sleep features; A multidimensional sleep feature vector determination module is used to input the multidimensional sleep features into a multidimensional sleep feature quantization model to obtain a multidimensional sleep feature vector; wherein, the multidimensional sleep feature quantization model includes a basic quantization model and an optimized quantization model; The multidimensional sleep features are input into a basic quantization model to obtain the score quantization effect of each sleep feature, including: the multidimensional sleep feature quantization model determines the score of the first sleep feature. The expression for the scoring quantification effect of the sleep feature is: in, Indicates the first The scoring and quantification effect of sleep characteristics. Indicates the first The first in the sleep data The raw scores of each data point Indicates the first The first in the sleep data The weight of each data point Indicates the first The number of elements in the dimensional sleep data; The multiple quantization results of the scores are input into the optimization quantization model to obtain a multidimensional sleep feature vector, including: the expression for determining the optimal quantization result is: in, This indicates the optimal quantization effect. Indicates whether the first exists. The optimization condition is not met when the first condition is not met. When optimizing the conditions of a task, ;otherwise, , This represents the objective function to be optimized. Indicates the first The set of weights for all questions under the dimension; if the optimal quantization effect is less than the preset quantization effect, the weights of at least one parameter among the clustering / classification model, subjective reference relevance, physiological index relevance, and classic scale relevance are adjusted to update the optimal quantization effect; if the optimal quantization effect is greater than or equal to the preset quantization effect, the optimal quantization effect is standardized to obtain the multidimensional sleep feature vector output by the multidimensional sleep feature quantization model. The sleep quality quantification result determination module is used to calculate and obtain the sleep quality quantification result based on the multidimensional sleep feature vector.

2. The method for quantifying sleep quality according to claim 1, characterized in that, The method further includes: The first quantification effect of the score quantification effect is obtained based on the clustering / classification model; The second quantification effect of the score quantification effect is obtained based on subjective reference relevance; The third quantitative effect of the score quantification effect is obtained based on the correlation of physiological indicators; The fourth quantification effect is obtained based on the correlation of the classic scale; The first quantization effect, the second quantization effect, the third quantization effect, and the fourth quantization effect are weighted to obtain the optimal quantization effect.

3. The method for quantifying sleep quality according to claim 1, characterized in that, The step of extracting features from the sleep data to obtain multidimensional sleep features includes: Obtaining rhythmic morning and evening characteristics based on rhythmic morning and evening data; Obtain sleep duration characteristics based on sleep duration data; Sleep stability characteristics are obtained based on sleep stability data; Characteristics of sleep debt are obtained based on sleep debt data; Identify the characteristics of daily routines based on routine data; The multidimensional sleep characteristics include the circadian rhythm characteristics, the sleep duration characteristics, the sleep stability characteristics, the sleep debt characteristics, and the sleep-wake cycle characteristics.

4. The method for quantifying sleep quality according to claim 1, characterized in that, The method further includes: The score distribution of each dimension of the sleep feature vector is determined using a multidimensional classification model. The multidimensional classification model determines the classification threshold based on the distribution characteristics of the sleep feature score distribution in each dimension; The optimal quantization effect corresponding to the multidimensional sleep feature vector is input into the multidimensional classification model to obtain the sleep quality classification label output by the multidimensional classification model.

5. The method for quantifying sleep quality according to claim 4, characterized in that, The calculation of sleep quality quantification results based on the multidimensional sleep feature vector includes: Determine the correlation matrix between the multidimensional sleep feature vector and disease risk; If the matrix elements in the correlation matrix If it is greater than the preset threshold, then the first Increased sleep characteristics and diseases Risks; If the matrix elements in the correlation matrix If the value is less than or equal to the preset threshold, then the first... Sleep characteristics reduce disease The risks.

6. The method for quantifying sleep quality according to claim 5, characterized in that, Determining the correlation matrix between the multidimensional sleep feature vector and disease risk includes: Based on the database, the ratio of each sleep feature to the corresponding disease risk was extracted; For the The logarithmic ratio of the ratios of the sleep characteristics is obtained by taking the logarithm of the ratios. Determine the variance value corresponding to the logarithmic ratio; The weights of random effects are determined based on the variance values. The pooled effect size is determined based on the random effect weights and the log ratio; The correlation matrix is ​​determined based on the combined effect size.

7. A sleep quality quantification system based on the sleep quality quantification method of claim 1, the system comprising: A multidimensional sleep feature acquisition module is used to extract features from the sleep data to obtain multidimensional sleep features; A multidimensional sleep feature vector determination module is used to input the multidimensional sleep features into a multidimensional sleep feature quantization model to obtain a multidimensional sleep feature vector. The correlation matrix building module is used to build a correlation matrix between the multidimensional sleep feature vector and disease risk; The association model building module is used to build an association model based on the association matrix and disease types. The disease early warning result determination module is used to output disease early warning results based on the association model.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the sleep quality quantification method according to any one of claims 1 to 6.

9. An electronic device, characterized in that, The electronic device includes: A memory that stores a computer program; The processor, which is communicatively connected to the memory, executes the sleep quality quantification method according to any one of claims 1 to 6 when calling the computer program.

Images