CN104706318A - Sleep analysis method and device - Google Patents

Abstract

The invention discloses a sleep analysis method and device, comprising: obtaining multi-axis acceleration data of a person to be monitored collected in each sub-monitoring time period according to a preset sampling frequency; The multi-axis acceleration data of the person to be monitored determines the amount of activity of the person to be monitored in each sub-monitoring time period; and for each sub-monitoring time period, according to the sub-monitoring time included in the time period window corresponding to the sub-monitoring time period The activity amount in each sub-monitoring time period is determined, and the activity characteristic value of the person to be detected is determined in each sub-monitoring time period, and the dynamic threshold of the activity characteristic value in the monitoring time period is determined; the activity amount in each sub-monitoring time period is respectively The feature value is compared with the dynamic threshold to obtain a first sleep analysis result indicating that the person to be monitored is in a sleep state or awake state within each sub-monitoring time period. By adopting the method provided by the embodiment of the present invention, the accuracy rate of wake-sleep classification is improved.

Description

A sleep analysis method and device

technical field

The invention relates to the field of signal analysis, in particular to a sleep analysis method and device.

Background technique

Sleep research is an important part of sleep science and electroencephalography, and it is also one of the hot spots of scientific research in the world today. Polysomnography monitoring is currently the internationally recognized "gold standard" for sleep monitoring. Through electrodes attached to the body of the person to be monitored, the blood oxygen, ECG, eye movement, leg movement, EEG and other indicators are recorded to monitor Judge the sleep condition of the person to be monitored. However, polysomnography monitoring equipment is expensive. With the advancement of science and technology, the research on sleep monitoring is gradually developing in the direction of miniaturization and familyization. By collecting multi-axis acceleration data during the user's sleep, the user's acceleration during sleep is smaller than that of waking to perform wake-sleep analysis.

In the existing technical solution, the multi-axis acceleration data of the person to be monitored is collected by the acceleration sensor, and the multi-axis acceleration data is divided into multiple sub-data segments, and the data obtained by judging the multi-axis acceleration data corresponding to the sub-data segments The number of activities in the time period is used as the amount of activity in the time period, and when the amount of activity in the time period is greater than a fixed threshold, it is determined that the person to be monitored is in the awake state in the time period, otherwise, it is in the sleep state, Analyze each sub-data segment, and finally obtain the sleep analysis result of the person to be monitored during the entire monitoring period. It is also possible to carry out follow-up processing on the sleep analysis results. For example, when the person to be monitored has a short-term wakefulness during a long sleep, the short-term wake-up state is judged as sleep. If there is a short sleep during the process, the short sleep state is judged as awake.

However, for different people to be monitored, sleep habits are different. Some people to be monitored may be relatively quiet during sleep, while others may be more active during sleep. Low.

Contents of the invention

Embodiments of the present invention provide a sleep analysis method and device to solve the problem in the prior art that the accuracy of wake-sleep classification based on fixed threshold judgment is low.

An embodiment of the present invention provides a method for obtaining multi-axis acceleration data of a person to be monitored collected in each sub-monitoring time period according to a preset sampling frequency. The multi-axis acceleration data includes multiple multi-axis accelerations, wherein one monitoring time period Include multiple sub-monitoring time periods;

Determining the amount of activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period;

For each sub-monitoring time period, according to the activity in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, determine the characteristic value of the activity of the person to be detected in each sub-monitoring time period, wherein , the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it;

According to the activity characteristic values in the plurality of sub-monitoring time periods in the monitoring period, determine the dynamic threshold of the activity characteristic value in the monitoring period;

The characteristic value of the amount of activity in each sub-monitoring time period is compared with the dynamic threshold to obtain a first sleep analysis result indicating that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period.

Using the method provided by the embodiment of the present invention, based on the activity of the sub-monitoring time period, and the activity of other sub-monitoring time periods included in the time period window corresponding to the sub-monitoring time period, determine the activity characteristics of the sub-monitoring time period value; according to the activity characteristic value of the sub-monitoring time period in the whole monitoring period, determine the dynamic threshold value of the activity quantity characteristic value in the whole monitoring period; judge the person to be monitored according to the dynamic threshold in each sub-monitoring period sleeping or awake. Compared with the prior art, the accuracy of wake-sleep classification is improved.

The embodiment of the present invention also provides a sleep analysis device, including:

The data acquisition unit is used to acquire the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency, the multi-axis acceleration data includes a plurality of multi-axis accelerations, wherein one monitoring time period includes Multiple sub-monitoring time periods;

An activity determination unit, configured to determine the activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period;

The activity characteristic value determining unit is used to determine, for each sub-monitoring time period, according to the activity amount in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, to determine that the person to be detected is in each sub-monitoring time period. The characteristic value of the amount of activity in the segment, wherein the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it;

A dynamic threshold determining unit, configured to determine the dynamic threshold of the characteristic value of the activity in the monitoring period according to the characteristic values of the activity in the plurality of sub-monitoring periods in the monitoring period;

A processing unit, configured to respectively compare the characteristic value of the amount of activity in each sub-monitoring time period with the dynamic threshold to obtain a first sleep analysis indicating that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period result.

Additional features and advantages of the application will be set forth in the description which follows, and, in part, will be obvious from the description, or may be learned by practice of the application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the description, and are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the attached picture:

Fig. 1 is one of the flowcharts of the sleep analysis method provided by the embodiment of the present invention;

Fig. 2 is the second flow chart of the sleep analysis method provided by the embodiment of the present invention;

FIG. 3 is a flow chart of performing sleep analysis on low-frequency multi-axis acceleration data provided by an embodiment of the present invention;

Fig. 4 is a schematic structural diagram of a sleep analysis device provided by an embodiment of the present invention.

Detailed ways

In order to provide an implementation plan for improving the accuracy of waking-sleep classification of the subject to be monitored, the embodiment of the present invention provides a sleep analysis method and device. The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings. It should be understood that the The described preferred embodiments are only used to illustrate and explain the present invention, not to limit the present invention. And in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

An embodiment of the present invention provides a sleep analysis method, the specific process is shown in Figure 1, including:

Step 101. Obtain the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency. The multi-axis acceleration data includes multiple multi-axis accelerations, wherein one monitoring time period includes multiple sub-monitoring times part.

Step 102: Determine the activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period.

Step 103: For each sub-monitoring time period, according to the activity amount in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, determine the activity characteristic value of the person to be tested in each sub-monitoring time period , wherein the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it.

Step 104, according to the activity characteristic values in the plurality of sub-monitoring time periods in the monitoring period, determine the dynamic threshold of the activity characteristic value in the monitoring period.

Step 105: Comparing the characteristic value of the amount of activity in each sub-monitoring time period with the dynamic threshold, and obtaining the first sleep analysis result indicating that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period.

In the embodiment of the present invention, the multi-axis acceleration data can be collected by the acceleration sensor, and the sleep analysis of the person to be monitored can be regarded as a monitoring time period, and the monitoring time period can be divided into multiple sub-monitoring time periods, and each sub-monitoring time period The multi-axis acceleration data of the monitoring time period is analyzed.

Using the method provided by the embodiment of the present invention, based on the activity of the sub-monitoring time period, and the activity of other sub-monitoring time periods included in the time period window corresponding to the sub-monitoring time period, determine the activity characteristics of the sub-monitoring time period value; according to the activity characteristic value of the sub-monitoring time period in the whole monitoring period, determine the dynamic threshold value of the activity quantity characteristic value in the whole monitoring period; judge the person to be monitored according to the dynamic threshold in each sub-monitoring period sleeping or awake. Compared with the prior art, the accuracy of wake-sleep classification is improved.

The method, device and corresponding system provided by the present invention will be described in detail below with specific embodiments in conjunction with the accompanying drawings. The detailed steps of the method are shown in Figure 2, including:

Step 201, perform band-pass filtering on the collected multi-axis acceleration data of the person to be monitored within the monitoring time period. The multi-axis acceleration generated by the user's human body activity has a frequency range, and the multi-axis acceleration is band-pass filtered, mainly to remove interference.

Step 202 , sampling the filtered data to the multi-axis acceleration data in each sub-monitoring time period according to a preset sampling frequency, wherein the sub-monitoring time period can be set to 1 minute.

Step 203, based on the sampled multi-axis acceleration data, determine the amount of activity of the person to be monitored in each sub-monitoring time period.

Wherein, when the multi-axis acceleration of the sampling point in the sub-monitoring time period is greater than the preset acceleration threshold, it is determined that the person to be monitored is active at the moment corresponding to the sampling point, and the sampling point is determined according to the preset sampling frequency; The total number of activities of the person to be monitored determined within the sub-monitoring time period is determined as the amount of activity of the person to be monitored in the sub-monitoring time period.

There are many methods for calculating the amount of activity, such as the threshold method, the zero-crossing method, and the area method. In this embodiment, the threshold method is used to determine the amount of activity.

Step 204, determine the mean value and variance of the activity amount in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, respectively as the mean value and variance of the activity amount in the time period window corresponding to the sub-monitoring time period, and The number of sub-monitoring time periods in which the activity amount is greater than the preset activity amount in the time period window is determined.

The time period window can be set to 5 minutes, and the sub-monitoring time period is 1 minute, then the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period is the sub-monitoring corresponding to 2 minutes before and after the current sub-monitoring time period time period and the current sub-monitoring time period, and determine the mean value and variance of the activity volume of the five sub-monitoring time periods.

Step 205: Carry out weighted summation of the mean value and variance of the activity amount in the time period window and the number of sub-monitoring time periods in which the activity amount in the time period window is greater than the preset activity amount, and obtain the activity amount in the sub-monitoring time period The characteristic value of the amount of activity is referred to as the PS value.

The determination of the PS value can also be based on the activity of the current sub-monitoring time period in the time period window, the logarithm of the activity in the current sub-monitoring time period, and the maximum value and variation of the activity in other sub-monitoring time periods in the time period window etc. to determine, the weighting coefficient is an empirical value.

Step 206: Determine the mean value and variance of multiple PS values in multiple sub-monitoring time periods in the monitoring time period, and use them as the mean value and variance of PS values in the monitoring time period respectively.

Step 207, based on the mean value and variance of the PS value within the monitoring period, determine the dynamic threshold of the PS value.

When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is greater than the first preset threshold value of the characteristic value of the activity amount, the first preset characteristic value threshold value is determined as the dynamic value of the characteristic value of the amount of activity in the monitoring time period threshold;

When the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring time period is not greater than the first preset threshold value of the characteristic value of the activity amount, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is less than the second preset activity When the threshold of the characteristic value of the activity is determined, the second preset threshold of the characteristic value of the activity is determined as the dynamic threshold of the characteristic value of the activity within the monitoring period, wherein the second preset threshold of the characteristic of the activity is smaller than the first preset threshold Activity characteristic value threshold;

When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not greater than the first preset threshold value of the characteristic value of the activity amount, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not less than the second preset When the characteristic value threshold is determined, the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring period is determined as the dynamic threshold value of the characteristic value of the activity quantity in the monitoring period.

Step 208: Comparing the PS value in the sub-monitoring time period with the dynamic threshold of the PS value in the monitoring time period, and obtaining the first sleep analysis result of the person to be monitored in the sub-monitoring time period.

Wherein, when the activity characteristic value of the sub-monitoring period is greater than the dynamic threshold of the activity characteristic value in the monitoring period, it is determined that the person to be monitored is awake in the sub-monitoring period;

When the characteristic value of the activity amount in the sub-monitoring time period is not greater than the dynamic threshold of the characteristic value of the activity amount in the monitoring time period, it is determined that the person to be monitored is awake in the sub-monitoring time period.

In the methods provided in the above-mentioned embodiments, other methods can also be used to determine the dynamic threshold, such as the maximum inter-class variance algorithm, which is an adaptive threshold method. Based on multiple thresholds, for each threshold, the awake and sleep As two categories, the inter-class variance is calculated, and the threshold that maximizes the inter-class variance of the two categories is used as the final threshold; the entropy threshold method determines the probability of each sub-monitoring time period being awake or sleeping according to different thresholds, and the corresponding entropy Value, determine the threshold that can maximize the entropy value; the minimum error method, this method comes from the Bayes minimum error classification method, Eb(T) is the probability of misclassifying the target class (awake) to the background class (sleep), Eo(T) It is the probability that the background class (sleep) is misclassified to the target class (awake), the total error probability E(T)=Eb(T)+Eo(T), and the minimum value of E(T) is the optimal classification method .

In addition, some users may have the habit of reading books and playing with mobile phones before going to bed. The frequency of multi-axis acceleration data brought by such activities is relatively low, and it may be judged to be in a sleep state only by using the above processing process. Therefore, the embodiment of the present invention Also provided is a method for sleep analysis of low-frequency multi-axis acceleration data, the specific steps are shown in Figure 3, including:

Step 301 , perform low-pass filtering on the multi-axis acceleration data of the person to be monitored in the monitoring period to obtain low-frequency multi-axis acceleration data in each sub-monitoring period.

Step 302. Determine the complexity of the low-frequency multi-axis acceleration data in each sub-monitoring time period respectively.

Among them, there are many ways to determine the complexity. In this scheme, the number of extreme values of the low-frequency multi-axis acceleration data in the sub-monitoring period and the difference between the adjacent maximum and minimum values can be determined first, and the two The parameters are weighted and summed to determine the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period.

Step 303 , according to the mean value and variance of the complexity of low-frequency multi-axis acceleration data in multiple sub-monitoring time periods in the monitoring time period, determine the dynamic threshold of the complexity in the monitoring time period. The method for determining the dynamic threshold of the complexity may be the same as the method for determining the dynamic threshold of the PS value described above, which will not be repeated here.

Step 304: According to whether the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period is greater than the dynamic threshold of the complexity, determine the second sleep analysis result that the person to be monitored is in the sleep state or awake state in the sub-monitoring time period.

When the sum of the mean and variance of the complexity within the monitoring period is greater than a first preset complexity threshold, determine the first preset complexity threshold as the dynamic threshold of complexity within the monitoring period;

When the sum of the mean and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean and variance of the complexity within the monitoring period is less than the second preset complexity threshold, the A second preset complexity threshold is determined as a dynamic threshold of complexity within the monitoring period, wherein the second preset complexity threshold is smaller than the first preset complexity threshold;

When the sum of the mean and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean and variance of the complexity within the monitoring period is not less than the second preset complexity threshold, the The sum of the mean value and the variance of the complexity in the monitoring time period is determined as the dynamic threshold of the complexity in the monitoring time period.

Step 305 , for the sub-monitoring time period in which the first sleep analysis result is a sleep state and the second sleep analysis result is an awake state, determine that the third sleep analysis result of the sub-monitoring time period is an awake state. Wherein, the third sleep analysis result is taken as the final sleep analysis result of the person to be monitored within the monitoring period. Further analysis can also be performed on the sleep status of the person to be monitored based on the third sleep analysis result.

Based on the same inventive concept, according to the sleep analysis method provided by the above-mentioned embodiments of the present invention, correspondingly, another embodiment of the present invention also provides a sleep analysis device. The schematic diagram of the device structure is shown in Figure 4, specifically including:

The data acquisition unit 401 is configured to acquire the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency, the multi-axis acceleration data includes multiple multi-axis accelerations, wherein one monitoring time period includes Multiple sub-monitoring time periods;

An activity determination unit 402, configured to determine the activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period;

The activity characteristic value determining unit 403 is used to determine the amount of activity of the person to be detected during each sub-monitoring time period according to the activity amount in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period for each sub-monitoring time period. The characteristic value of the amount of activity in the segment, wherein the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it;

A dynamic threshold determination unit 404, configured to determine the dynamic threshold of the activity characteristic value in the monitoring period according to the activity characteristic values in the plurality of sub-monitoring periods in the monitoring period;

The processing unit 405 is configured to respectively compare the characteristic value of the amount of activity in each sub-monitoring time period with the dynamic threshold, and obtain the first sleep analysis result that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period .

Further, the activity determination unit 402 is specifically used to determine that the person to be monitored is active at the moment corresponding to the sampling point when the multi-axis acceleration of the sampling point within the sub-monitoring time period is greater than the preset acceleration threshold. The point is determined according to the preset sampling frequency; and the total number of activities of the person to be monitored is determined within the sub-monitoring time period as the activity amount of the person to be monitored in the sub-monitoring time period.

Further, the activity characteristic value determination unit 403 is specifically used to determine the mean value and variance of the activity in the sub-monitoring time period included in the sub-monitoring time period corresponding to the time period window corresponding to the sub-monitoring time period, respectively as the time corresponding to the sub-monitoring time period The mean and variance of the activity in the segment window;

Determine the number of sub-monitoring time periods in which the activity amount is greater than the preset activity amount in the time period window;

Carry out weighted summation of the mean value and variance of the activity amount in the time period window and the number of sub-monitoring time periods in which the activity amount in the time period window is greater than the preset activity amount, and obtain the activity quantity characteristics in the sub-monitoring time period value;

The dynamic threshold determination unit 404 is specifically used to: determine the mean value and variance of the multiple activity characteristic values in the multiple sub-monitoring time periods in the monitoring period, and use them as the mean and variance of the activity characteristic values in the monitoring period;

When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is greater than the first preset threshold value of the characteristic value of the activity amount, the first preset characteristic value threshold value is determined as the dynamic value of the characteristic value of the amount of activity in the monitoring time period threshold;

When the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring time period is not greater than the first preset threshold value of the characteristic value of the activity amount, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is less than the second preset activity When the threshold of the characteristic value of the activity is determined, the second preset threshold of the characteristic value of the activity is determined as the dynamic threshold of the characteristic value of the activity within the monitoring period, wherein the second preset threshold of the characteristic of the activity is smaller than the first preset threshold Activity characteristic value threshold;

When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not greater than the first preset threshold value of the characteristic value of the activity amount, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not less than the second preset When the characteristic value threshold is determined, the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring period is determined as the dynamic threshold value of the characteristic value of the activity quantity in the monitoring period.

Further, the processing unit 405 is specifically configured to determine that the person to be monitored is in an awake state during the sub-monitoring time period when the characteristic value of the activity level in the sub-monitoring time period is greater than the dynamic threshold of the characteristic value of the activity level in the monitoring time period; And when the activity characteristic value of the sub-monitoring period is not greater than the dynamic threshold of the activity characteristic value in the monitoring period, it is determined that the person to be monitored is awake in the sub-monitoring period.

Further, the above-mentioned device also includes: a low-frequency data processing unit 406, configured to perform low-pass filtering on the multi-axis acceleration data in the multiple sub-monitoring time periods according to a preset frequency to obtain low-frequency multi-axis acceleration data in each sub-monitoring time period. acceleration data;

Determine the complexity of the low-frequency multi-axis acceleration data in each sub-monitoring time period separately;

Determine the dynamic threshold of the complexity in the monitoring time period according to the mean value and variance of the complexity of the low-frequency multi-axis acceleration data in the plurality of sub-monitoring time periods in the monitoring time period;

According to whether the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period is greater than the dynamic threshold of the complexity, determine the second sleep analysis result that the person to be monitored is in a sleep state or awake state in the sub-monitoring time period;

For the sub-monitoring time period in which the first sleep analysis result is a sleep state and the second sleep analysis result is an awake state, it is determined that the third sleep analysis result of the sub-monitoring time period is an awake state.

Further, the low-frequency data processing unit 406 determines the complexity of the low-frequency multi-axis acceleration data in a sub-monitoring time period, and is specifically used to determine the number of extreme values of the low-frequency multi-axis acceleration data in the sub-monitoring time period and the number of adjacent poles. The difference between the maximum value and the minimum value; and the number of extreme values of the low-frequency multi-axis acceleration data within the sub-monitoring time period and the difference between the adjacent maximum value and the minimum value are weighted and summed to determine the sub-monitoring time Complexity of low-frequency multi-axis acceleration data within a segment.

Further, the low-frequency data processing unit 406 determines the dynamic threshold of the complexity within the monitoring time period, specifically for when the sum of the mean and variance of the complexity within the monitoring time period is greater than the first preset complexity threshold, the The first preset complexity threshold is determined as a dynamic threshold of complexity within the monitoring period;

When the sum of the mean and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean and variance of the complexity within the monitoring period is less than the second preset complexity threshold, the A second preset complexity threshold is determined as a dynamic threshold of complexity within the monitoring period, wherein the second preset complexity threshold is smaller than the first preset complexity threshold;

When the sum of the mean and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean and variance of the complexity within the monitoring period is not less than the second preset complexity threshold, the The sum of the mean value and the variance of the complexity in the monitoring time period is determined as the dynamic threshold of the complexity in the monitoring time period.

The functions of the above units may correspond to the corresponding processing steps in the flow shown in FIG. 1 to FIG. 3 , and will not be repeated here.

To sum up, the solution provided by the embodiment of the present invention obtains the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency; According to the multi-axis acceleration data of the person to be monitored, the activity amount of the person to be monitored in each sub-monitoring period is determined; and for each sub-monitoring period, according to the sub-monitoring period included in the time period window corresponding to the sub-monitoring period Determine the activity characteristic value of the person to be detected in each sub-monitoring time period; and determine the activity characteristic in the monitoring time period according to the activity characteristic values in the multiple sub-monitoring time periods in the monitoring time period The dynamic threshold of the value; the characteristic value of the amount of activity in each sub-monitoring time period is compared with the dynamic threshold, and the first sleep analysis result that the person to be monitored is sleeping or awake in each sub-monitoring time period is obtained. Compared with the prior art, the method provided by the embodiment of the present invention improves the accuracy of wake-sleep classification.

The sleep analysis device provided by the embodiments of the present application can be realized by a computer program. Those skilled in the art should be able to understand that the above-mentioned module division method is only one of many module division methods. If it is divided into other modules or not divided into modules, as long as the sleep analysis device has the above functions, it should be within the protection scope of this application within.

The present application is described with reference to flowcharts and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and combinations of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a Means for realizing the functions specified in one or more steps of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart flow or flows and/or block diagram block or blocks.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (14)

1. A sleep analysis method, characterized in that, comprising: Obtain the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency, the multi-axis acceleration data includes multiple multi-axis accelerations, wherein one monitoring time period includes multiple sub-monitoring time periods; Determining the amount of activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period; For each sub-monitoring time period, according to the activity in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, determine the characteristic value of the activity of the person to be detected in each sub-monitoring time period, wherein , the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it; According to the activity characteristic values in the plurality of sub-monitoring time periods in the monitoring period, determine the dynamic threshold of the activity characteristic value in the monitoring period; The characteristic value of the amount of activity in each sub-monitoring time period is compared with the dynamic threshold to obtain a first sleep analysis result indicating that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period. 2. The method according to claim 1, wherein, based on the multi-axis acceleration data of the person to be monitored collected in the sub-monitoring time period, the activity level of the person to be monitored in the sub-monitoring time period is determined , including: When the multi-axis acceleration of the sampling point in the sub-monitoring time period is greater than the preset acceleration threshold, it is determined that the person to be monitored is active at the moment corresponding to the sampling point, and the sampling point is determined according to the preset sampling frequency; The total number of activities of the person to be monitored within the sub-monitoring time period is determined as the activity amount of the person to be monitored in the sub-monitoring time period. 3. The method according to claim 1, wherein, according to the amount of activity in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, it is determined that the person to be tested is within each sub-monitoring time period The characteristic values of the activity volume, specifically include: Determine the mean value and variance of the amount of activity in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, as the mean value and variance of the activity amount in the time period window corresponding to the sub-monitoring time period; Determine the number of sub-monitoring time periods in which the activity amount is greater than the preset activity amount in the time period window; Carry out weighted summation of the mean value and variance of the activity amount in the time period window and the number of sub-monitoring time periods in which the activity amount in the time period window is greater than the preset activity amount, and obtain the activity quantity characteristics in the sub-monitoring time period value; According to the activity characteristic values in the plurality of sub-monitoring time periods in the monitoring period, determine the dynamic threshold of the activity characteristic value in the monitoring period, specifically including: Determine the mean value and variance of the multiple activity characteristic values in the plurality of sub-monitoring time periods in the monitoring time period, and use them as the mean value and variance of the activity characteristic values in the monitoring time period respectively; When the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring period is greater than the first preset threshold value of the characteristic value of the activity amount, the first preset characteristic value threshold value is determined as the characteristic value of the amount of activity in the monitoring period dynamic threshold of value; When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is not greater than the first preset threshold value of the characteristic value of the amount of activity, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is less than the second preset When setting the threshold of the characteristic value of activity, the second preset threshold of characteristic value of activity is determined as the dynamic threshold of the characteristic value of activity in the monitoring time period, wherein the second preset threshold of characteristic value of activity is less than The first preset activity level feature value threshold; When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not greater than the first preset threshold value of the characteristic value of the amount of activity, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not less than the second When the feature value threshold is preset, the sum of the mean value and the variance of the feature value of the activity amount in the monitoring period is determined as the dynamic threshold of the feature value of the activity mass in the monitoring period. 4. The method according to claim 1, characterized in that, the activity characteristic value in the sub-monitoring time period is compared with the dynamic threshold of the activity characteristic value in the monitoring time period, and the person to be monitored is obtained in the sub-monitoring period. The first sleep analysis results for the time period, including: When the activity characteristic value of the sub-monitoring period is greater than the dynamic threshold of the activity characteristic value in the monitoring period, it is determined that the person to be monitored is awake in the sub-monitoring period; When the activity characteristic value of the sub-monitoring period is not greater than the dynamic threshold of the activity characteristic value in the monitoring period, it is determined that the person to be monitored is awake in the sub-monitoring period. 5. The method according to claim 1, wherein, after obtaining the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency, further comprising: performing low-pass filtering on the multi-axis acceleration data in the plurality of sub-monitoring time periods according to a preset frequency to obtain low-frequency multi-axis acceleration data in each sub-monitoring time period; Determine the complexity of the low-frequency multi-axis acceleration data in each sub-monitoring time period separately; Determine the dynamic threshold of the complexity in the monitoring time period according to the mean value and variance of the complexity of the low-frequency multi-axis acceleration data in the plurality of sub-monitoring time periods in the monitoring time period; According to whether the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period is greater than the dynamic threshold of the complexity, determine the second sleep analysis result that the person to be monitored is in a sleep state or awake state in the sub-monitoring time period; For the sub-monitoring time period in which the first sleep analysis result is a sleep state and the second sleep analysis result is an awake state, it is determined that the third sleep analysis result of the sub-monitoring time period is an awake state. 6. The method according to claim 5, wherein determining the complexity of the low-frequency multi-axis acceleration data in a sub-monitoring time period specifically includes: Determine the number of extreme values of the low-frequency multi-axis acceleration data within the sub-monitoring time period and the difference between adjacent maximum values and minimum values; Perform weighted summation on the number of extreme values of the low-frequency multi-axis acceleration data in the sub-monitoring time period and the difference between the adjacent maximum value and the minimum value, and determine the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period Spend. 7. The method according to claim 5, wherein determining the dynamic threshold of complexity within the monitoring time period specifically comprises: When the sum of the mean value and the variance of complexity within the monitoring period is greater than a first preset complexity threshold, determining the first preset complexity threshold as a dynamic threshold of complexity within the monitoring period; When the sum of the mean value and the variance of the complexity within the monitoring time period is not greater than the first preset complexity threshold, and the sum of the mean value and the variance of the complexity within the monitoring time period is less than the second preset complexity threshold, determining the second preset complexity threshold as a dynamic threshold of complexity within the monitoring period, wherein the second preset complexity threshold is smaller than the first preset complexity threshold; When the sum of the mean value and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean value and variance of the complexity within the monitoring period is not less than the second preset complexity threshold , determining the sum of the mean value and the variance of the complexity within the monitoring time period as the dynamic threshold of the complexity within the monitoring time period. 8. A sleep analysis device, characterized in that, comprising: The data acquisition unit is used to acquire the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period according to the preset sampling frequency, the multi-axis acceleration data includes a plurality of multi-axis accelerations, wherein one monitoring time period includes Multiple sub-monitoring time periods; An activity determination unit, configured to determine the activity of the person to be monitored in each sub-monitoring time period based on the multi-axis acceleration data of the person to be monitored collected in each sub-monitoring time period; The activity characteristic value determining unit is used to determine, for each sub-monitoring time period, according to the activity amount in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period, to determine that the person to be detected is in each sub-monitoring time period. The characteristic value of the amount of activity in the segment, wherein the time period window corresponding to the sub-monitoring time period includes the sub-monitoring time period and several sub-monitoring time periods before and after it; A dynamic threshold determining unit, configured to determine the dynamic threshold of the characteristic value of the activity in the monitoring period according to the characteristic values of the activity in the plurality of sub-monitoring periods in the monitoring period; A processing unit, configured to respectively compare the characteristic value of the amount of activity in each sub-monitoring time period with the dynamic threshold to obtain a first sleep analysis indicating that the person to be monitored is in a sleep state or awake state in each sub-monitoring time period result. 9. The device according to claim 8, wherein the activity determination unit is specifically configured to determine that the to-be-monitored The person is active at the moment corresponding to the sampling point, and the sampling point is determined according to the preset sampling frequency; The amount of activity of the person during the sub-monitoring time period. 10. The device according to claim 8, wherein the activity characteristic value determining unit is specifically configured to determine the mean value of the activity in the sub-monitoring time period included in the time period window corresponding to the sub-monitoring time period , variance, respectively as the mean value and variance of the amount of activity in the time period window corresponding to the sub-monitoring time period; Determine the number of sub-monitoring time periods in which the activity amount is greater than the preset activity amount in the time period window; Carry out weighted summation of the mean value and variance of the activity amount in the time period window and the number of sub-monitoring time periods in which the activity amount in the time period window is greater than the preset activity amount, and obtain the activity quantity characteristics in the sub-monitoring time period value; The dynamic threshold determination unit is specifically used for: Determine the mean value and variance of the multiple activity characteristic values in the plurality of sub-monitoring time periods in the monitoring time period, and use them as the mean value and variance of the activity characteristic values in the monitoring time period respectively; When the sum of the mean value and the variance of the characteristic value of the activity amount in the monitoring period is greater than the first preset threshold value of the characteristic value of the activity amount, the first preset characteristic value threshold value is determined as the characteristic value of the amount of activity in the monitoring period dynamic threshold of value; When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is not greater than the first preset threshold value of the characteristic value of the amount of activity, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring time period is less than the second preset When setting the threshold of the characteristic value of activity, the second preset threshold of characteristic value of activity is determined as the dynamic threshold of the characteristic value of activity in the monitoring time period, wherein the second preset threshold of characteristic value of activity is less than The first preset activity level feature value threshold; When the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not greater than the first preset threshold value of the characteristic value of the amount of activity, and the sum of the mean value and the variance of the characteristic value of the amount of activity in the monitoring period is not less than the second When the feature value threshold is preset, the sum of the mean value and the variance of the feature value of the activity amount in the monitoring period is determined as the dynamic threshold of the feature value of the activity mass in the monitoring period. 11. The device according to claim 8, wherein the processing unit is specifically configured to determine that when the characteristic value of the activity amount in the sub-monitoring period is greater than the dynamic threshold of the characteristic value of the activity amount in the monitoring period The person to be monitored is awake in the sub-monitoring time period; Awakened state during the time period. 12. The device according to claim 8, further comprising: a low-frequency data processing unit, configured to low-pass the multi-axis acceleration data in the plurality of sub-monitoring time periods according to a preset frequency Filtering to obtain the low-frequency multi-axis acceleration data in each sub-monitoring time period; Determine the complexity of the low-frequency multi-axis acceleration data in each sub-monitoring time period separately; Determine the dynamic threshold of the complexity in the monitoring time period according to the mean value and variance of the complexity of the low-frequency multi-axis acceleration data in the plurality of sub-monitoring time periods in the monitoring time period; According to whether the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period is greater than the dynamic threshold of the complexity, determine the second sleep analysis result that the person to be monitored is in a sleep state or awake state in the sub-monitoring time period; For the sub-monitoring time period in which the first sleep analysis result is a sleep state and the second sleep analysis result is an awake state, it is determined that the third sleep analysis result of the sub-monitoring time period is an awake state. 13. The device according to claim 12, wherein the low-frequency data processing unit determines the complexity of the low-frequency multi-axis acceleration data in a sub-monitoring time period, and is specifically used to determine the low-frequency acceleration data in the sub-monitoring time period. The number of extreme values of the multi-axis acceleration data and the difference between the adjacent maximum value and the minimum value; and the number of extreme values of the low-frequency multi-axis acceleration data and the adjacent maximum value and The difference between the minimum values is weighted and summed to determine the complexity of the low-frequency multi-axis acceleration data in the sub-monitoring time period. 14. The device according to claim 12, wherein the low-frequency data processing unit determines a dynamic threshold of complexity within the monitoring time period, specifically for when the mean value of the complexity within the monitoring time period is equal to When the sum of the variances is greater than a first preset complexity threshold, the first preset complexity threshold is determined as a dynamic threshold of complexity within the monitoring period; When the sum of the mean value and the variance of the complexity within the monitoring time period is not greater than the first preset complexity threshold, and the sum of the mean value and the variance of the complexity within the monitoring time period is less than the second preset complexity threshold, determining the second preset complexity threshold as a dynamic threshold of complexity within the monitoring period, wherein the second preset complexity threshold is smaller than the first preset complexity threshold; When the sum of the mean value and variance of the complexity within the monitoring period is not greater than the first preset complexity threshold, and the sum of the mean value and variance of the complexity within the monitoring period is not less than the second preset complexity threshold , determining the sum of the mean value and the variance of the complexity within the monitoring time period as the dynamic threshold of the complexity within the monitoring time period.