A multi-parameter based sleep apnea frequency quantification method

Abstract

The application discloses a sleep apnea frequency measurement method based on multiple parameters, records a sleep time, and respectively acquires a heart rate, a blood oxygen concentration and a micro-motion acceleration of a human body during sleep, divides and analyzes time periods based on the blood oxygen concentration, gives different weight values of the heart rate, the blood oxygen concentration and the micro-motion acceleration in the analysis time periods according to states of the human body to form a heart rate weight, a blood oxygen weight and an acceleration weight, obtains a sleep stage weight according to the sleep time, and regards the analysis time period as one apnea event and marks the analysis time period when a comprehensive index of the heart rate weight, the blood oxygen weight, the acceleration weight and the sleep stage weight exceeds a set calculation threshold value, and finally counts a total number of marks in the sleep time. The application provides a sleep apnea frequency measurement method based on multiple parameters, which can obtain an index of a short calculation process, an easy calculation mode and a balanced judgment of the apnea frequency without affecting the sleep quality of a patient.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical intelligence, more specifically, to a sleep apnea frequency measurement method based on multiple parameters. BACKGROUND

[0002] Sleep breathing disorder refers to a phenomenon that various reasons cause breathing interruption during sleep, thereby affecting sleep quality, and sleep apnea is the most common type of sleep breathing disorder, and sleep apnea syndrome (SAS) is manifested as respiratory obstruction of the nose-mouth-lung. People with sleep apnea syndrome have a monitoring demand for sleep quality.

[0003] In the prior art, a polysomnography (PSG) is often used to monitor patients. Since the PSG needs to collect multiple body data, mostly heart rate data, a plurality of electrodes need to be connected to the human body during use, and the electrode patches are all concentrated on the chest cavity of the human body, which greatly affects the normal rest of the patient and also leads to low objectivity and accuracy of the collected data.

[0004] In addition, if the breathing is interrupted during sleep, the flow of air into the lungs is interrupted, which causes the human body to be in a state of hypoxia, and the blood oxygen concentration is reduced. Therefore, the blood oxygen concentration is also an important parameter for monitoring sleep quality. On this basis, the prior art uses an oximeter to monitor the blood oxygen concentration in order to obtain judgment data of sleep quality. However, the decrease in blood oxygen concentration may also be caused by other reasons such as heart failure. Therefore, monitoring the blood oxygen concentration alone cannot obtain accurate data about sleep quality.

[0005] In view of the above, the prior art combines the above two parameters to monitor sleep quality, such as patent document 202010913921.2 “Method, system, terminal and storage medium for monitoring sleep apnea event”. This patent analyzes the changes in the heart rate and respiratory rate of the monitored object by detecting the ballistocardiogram signal and the chest and abdominal breathing signal, and simultaneously monitors the blood oxygen concentration. According to the heart rate and respiratory rate, the blood oxygen concentration is optimized, and the optimized blood oxygen concentration data is used to analyze the sleep apnea event. This technical solution avoids directly measuring the heart rate using a polysomnography, but instead obtains the heart rate data through a ballistocardiogram signal, and then measures the respiratory rate and blood oxygen concentration. The heart rate data and respiratory rate data are used to optimize the blood oxygen concentration data. This technical solution has the following defects:

[0006] 1. This technical solution uses blood oxygen concentration as the only index for judging sleep quality and analyzing sleep apnea events. The judgment index is single, the deviation value tends to one direction, the data comprehensive balance is poor, and the analysis of the monitoring personnel is prone to deviation, resulting in incorrect judgment.

[0007] 2. The optimization of blood oxygen concentration by heart rate and respiratory rate is quite complex, involving multiple functions, covariance formulas, integral calculations and other mathematical processing methods, which are difficult for people to understand and have strong professional nature, and are long in calculation process and large in data cache for intelligent devices, which is not conducive to the popularization, improvement and server laying of subsequent products. SUMMARY

[0008] In order to solve the problems of complex optimization process of blood oxygen concentration by heart rate and respiratory rate and deviation of single index judgment in the prior art, the present application provides a sleep apnea frequency measurement method based on multiple parameters, which obtains an index that can balance the judgment of apnea frequency without affecting the sleep quality of the patient, and has a short calculation process and easy calculation method.

[0009] The technical scheme of the present application is as follows:

[0010] A sleep apnea frequency measurement method based on multiple parameters, records the sleep time, and obtains the heart rate, blood oxygen concentration and micro-motion acceleration of the human body during sleep, divides the analysis time period based on the blood oxygen concentration, gives different weight values to the heart rate, blood oxygen concentration and micro-motion acceleration in the analysis time period according to the state of the human body to form heart rate weight, blood oxygen weight and acceleration weight, and obtains the sleep stage weight according to the sleep time, when the comprehensive index of the heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight exceeds the set calculation threshold, the analysis time period is regarded as an apnea event and marked, and finally the total number of marks in the sleep time is counted.

[0011] The present application takes blood oxygen concentration as the analysis starting point, extracts the sleep time period with low blood oxygen concentration, analyzes the apnea relationship and correlation of each parameter index corresponding to the heart rate, blood oxygen concentration, micro-motion acceleration and sleep stage in the time period, respectively gives different weight values, and then comprehensively considers the weights of each parameter index to finally obtain the comprehensive index of the sleep time period, compares the comprehensive index with the set calculation threshold, and exceeds the set calculation threshold to be regarded as an apnea event. The calculation index of the measurement method includes heart rate, blood oxygen concentration, micro-motion acceleration and sleep stage, avoiding the problem of single index measurement, multiple parameter indexes, and different weights of each parameter index according to different states of the human body and different positions of the analysis time period, balancing each parameter index, preventing data deviation of the parameter index in a certain direction, and affecting the measurement result.

[0012] The sleep apnea frequency measurement method based on multiple parameters sets a low blood oxygen threshold value, collects the blood oxygen saturation of the human body, extracts and marks a first instantaneous time point at which the blood oxygen saturation is lower than the low blood oxygen threshold value and a second instantaneous time point at which the blood oxygen saturation adjacent to the first instantaneous time point is higher than the low blood oxygen threshold value, and a time period between the first instantaneous time point and the second instantaneous time point is an analysis time period.

[0013] When the apnea event occurs, the air flow into the lung is interrupted due to the blockage of the nasal-oral-lung respiratory tract, the air intake is reduced, the human body is in a hypoxic state, the blood oxygen concentration is reduced, and therefore the blood oxygen concentration can be considered as a sufficient condition of the apnea event. When the apnea event occurs, the phenomenon of the reduction of the blood oxygen concentration will occur. Meanwhile, the apnea event itself is a short event, and the value of the reduction of the blood oxygen concentration will be restored in a short time. Or in another case, the apnea event occurs continuously, the value of the blood oxygen concentration remains low for a long time, but it will eventually return to normal. Therefore, the blood oxygen concentration is used as the trigger condition for each analysis, the data of each reduction and recovery process of the blood oxygen concentration is intercepted, a short time period is obtained, the time period is considered as a low blood oxygen event, and the data monitored in the time period is extracted as the basis for analysis and calculation.

[0014] The sleep apnea frequency measurement method based on multiple parameters sets a low blood oxygen threshold value, collects the blood oxygen saturation of the human body, extracts and marks a first instantaneous time point at which the blood oxygen saturation is lower than the low blood oxygen threshold value and a second instantaneous time point at which the blood oxygen saturation adjacent to the first instantaneous time point is higher than the low blood oxygen threshold value, and a time period between the first instantaneous time point and the second instantaneous time point is an analysis time period.

[0015] The minimum blood oxygen concentration in the analysis time period is recorded, the low blood oxygen range interval in which the analysis time period is located is determined according to the minimum blood oxygen concentration, the weight value corresponding to the current analysis time period is determined, and the blood oxygen weight is obtained by combining the minimum blood oxygen concentration and the weight value of the corresponding low blood oxygen range interval.

[0016] The low blood oxygen state caused by apnea has different phenomena, and the amplitude and rate of the decrease of the blood oxygen saturation of patients with different degrees of sleep apnea syndrome are not the same, and are positively correlated with the degree of sleep apnea syndrome. The more serious the symptoms of apnea, the faster the amplitude and rate of the decrease of the blood oxygen concentration, and therefore, when the blood oxygen concentration is used as an index for analysis, different concentration ranges should be divided for hierarchical processing. The division method can adopt average division, that is, a blood oxygen concentration interval is set at the same blood oxygen concentration interval, or the blood oxygen concentration interval is set according to the positive correlation between the symptoms of apnea and the amplitude and rate of the decrease of the blood oxygen concentration.

[0017] Further, the length of the analysis time period is recorded, the analysis time period is given a weight value to form a low blood oxygen time weight in the calculation of the blood oxygen weight, and the blood oxygen weight is obtained by superimposing the low blood oxygen time weight and the weight corresponding to the minimum blood oxygen concentration.

[0018] The low oxygen state caused by apnea is caused by the blockage of the human respiratory tract. If the apnea is caused by a pathological problem, the duration of the continuous low oxygen state is relatively short and regular. However, if the apnea is caused by a non-pathological problem, the duration of the continuous low oxygen state is longer and the symptoms are relatively severe. Therefore, the duration of the low oxygen state is also one of the important indicators for judging the symptoms of apnea. In addition, the possibility of a short low oxygen state caused by other factors cannot be completely ruled out, and it cannot be too certain. Therefore, the weight value of the time length is set to distinguish the difference and possibility between short and long low oxygen states, and to improve the accuracy.

[0019] The above-mentioned sleep apnea frequency measurement method based on multiple parameters uses a PPG device to obtain human heart rate data, judges and analyzes whether there is a process of heart rate reduction and recovery in the analysis period. If there is, the maximum heart rate and the minimum heart rate in the analysis period are extracted, and the heart rate weight is obtained by combining the maximum heart rate, the minimum heart rate and the corresponding weight value.

[0020] In order to reduce the impact on the patient, the prior art changes the polysomnography to a portable wearable device, which uses the PPG (Photoplethysmography) principle to collect data on the patient's heart rate. The PPG device uses the fact that different tissues of the human body absorb different amounts of detection light. After the detection light emitted by the PPG device is transmitted through the human body or reflected on the surface of the human body, the light intensity will change regularly. Therefore, the detection light can be shone on certain tissues of the human body, and the changed detection light or the image of the human tissue during the process can be received by the sensor. When using the PPG device to collect human heart rate data, the detection light is mostly transmitted through the blood vessels. The blood vessels change regularly due to the contraction and relaxation of the heart, which causes the transmitted detection light to change regularly, and thus the data of the heart rate is obtained.

[0021] The correlation between heart-lung function, blood oxygen concentration and heart rate, etc. determines that the heart rate is directly affected during apnea. According to medical common sense, when the human body function decreases or the oxygen and energy supply is insufficient, the heart movement is limited and the movement intensity is reduced. Therefore, the heart rate in the apnea state is lower than the heart rate in the normal state. When the breathing recovers, the heart rate also recovers. Therefore, the heart rate has a downward trend and then rises during the apnea process. This process is also a sufficient condition for apnea. Therefore, the heart rate drop caused by apnea is also related to the symptoms of apnea, and needs to be considered in the process of calculating the weight of the parameter index.

[0022] The multi-parameter-based sleep apnea frequency measurement method sets an acceleration threshold, obtains human micro-motion acceleration data, and when the micro-motion acceleration is less than the acceleration threshold, combines the acceleration threshold, the minimum blood oxygen concentration in the analysis time period, and the corresponding weight value to form an acceleration weight.

[0023] When apnea occurs, the blood oxygen saturation is low to a certain threshold, and the human body is in a low blood oxygen state. During the apnea recovery period, body movement is highly probable, that is, the sleeping human body experiences micro-waking or conscious situations, which is a stress response of the human body in a hypoxic state. Since the apnea recovery period is a period of time away from the decrease in blood oxygen, body movement is likely to occur during the apnea recovery period. Therefore, micro-motion acceleration can be used as a parameter index of sleep apnea.

[0024] The existence of acceleration collected by the acceleration sensor is not the acceleration data occurring in a low blood oxygen state, but may be normal sleep movement, not body movement caused by sleep apnea. Therefore, it needs to correspond to a low blood oxygen state. Since the cause of such body movement is mainly caused by low blood oxygen, an acceleration threshold is set to exclude large-scale body movement. On the other hand, this value is associated with a high blood oxygen concentration. Therefore, when calculating the weight, the blood oxygen concentration needs to be considered to form the acceleration weight.

[0025] The multi-parameter-based sleep apnea frequency measurement method determines the sleep stage in which the analysis time period is located and obtains the weight value corresponding to the sleep stage according to the sleep time, and obtains the sleep stage weight by combining the number of marked apnea events and the weight value.

[0026] As known, sleep has multiple sleep stages, and according to different standards, there are multiple division methods for sleep stages, one of which includes four sleep stages, initial sleep stage, light sleep stage, moderate sleep stage, and deep sleep stage, etc. The present application divides the sleep stage into three stages of rapid eye movement, light sleep, and deep sleep by using pulse rate variability. Regardless of the division method of the sleep stage, in each sleep stage, the physiological indicators and parameter indicators of the human body all have different standards and normal activity values, and therefore need to be treated differently.

[0027] Similarly, in the present application, the sleep activity in different stages corresponds to different reference standards and relative normal values of apnea symptoms, blood oxygen concentration, heart rate, micro-motion acceleration, etc., so the state of sleep stage can be used as one of the reference indexes for counting the number of apnea. With the deepening of sleep state, the human activity frequency decreases, and various physiological indexes decrease, but due to the existence of apnea, the sleep quality of the human body is affected, resulting in the indexes of people with apnea in each sleep stage being slightly higher than normal. The more serious the apnea symptoms, the higher the abnormal indexes of the corresponding sleep stage, and the deeper the sleep stage, the greater the impact of apnea, and the two are positively correlated.

[0028] The above-mentioned sleep apnea number counting method based on multiple parameters includes the following steps:

[0029] Step S1. Determine whether the detected object is asleep. If it is asleep, mark the sleep time point, start counting the sleep time, and start the data collection device to record the human body data. If it is in a wake-up state, mark the wake-up time point, calculate the sleep duration, and enter step S2.

[0030] Step S2. Obtain all parameter index data for counting apnea during sleep, including micro-motion acceleration, heart rate, and blood oxygen saturation.

[0031] Step S3. Collect and process data.

[0032] Step S4. Extract the time point when the blood oxygen saturation is lower than the low blood oxygen threshold, and record it as a low blood oxygen event. Determine whether this low blood oxygen event is caused by non-pathological causes. If yes, continue to extract the next time point when the blood oxygen saturation is lower than the low blood oxygen threshold for analysis. If no, extract the collected data in the analysis period corresponding to this low blood oxygen event.

[0033] Step S5. Determine whether this low blood oxygen event is an apnea event. If yes, mark this low blood oxygen event as an apnea event. If no, continue to analyze the next low blood oxygen event. After all low blood oxygen events are analyzed, count the number of marked apnea events, and record it as the total number of apnea events in this sleep time.

[0034] Further, in step S4, determine whether the blood oxygen saturation in this low blood oxygen event has recovered to normal level after the low peak. If it has recovered to normal level, the time period from when the blood oxygen saturation is lower than the low blood oxygen threshold to when the blood oxygen saturation recovers to normal level is defined as the analysis period, and the low blood oxygen concentration range interval to which the low peak value of blood oxygen saturation in the analysis period belongs is determined.

[0035] If the low peak of blood oxygen saturation is from the third blood oxygen concentration interval, and the length of the analysis time period is higher than the average length of a normal apnea event, the cause of the low blood oxygen event is determined to be non-pathological;

[0036] If the low peak of blood oxygen saturation is not from the third blood oxygen concentration interval, the cause of the low blood oxygen event is determined to be pathological;

[0037] If the length of the analysis time period is lower than or equal to the average length of a normal apnea event, the cause of the low blood oxygen event is determined to be pathological;

[0038] If the blood oxygen concentration does not recover to a normal level after the low peak, the cause of the low blood oxygen event is determined to be non-pathological.

[0039] Further, in step S5, the heart rate weight, the blood oxygen weight, the acceleration weight, and the sleep stage weight are obtained, the sum of the heart rate weight, the blood oxygen weight, the acceleration weight, and the sleep stage weight is calculated, the calculation threshold is set to 1, and when the sum of the heart rate weight, the blood oxygen weight, the acceleration weight, and the sleep stage weight is greater than 1, the low blood oxygen event is determined to be an apnea event.

[0040] According to the application of the above scheme, the beneficial effects are that,

[0041] 1. The PPG wearable device collects human heart rate data, compared with a polysomnogram, a portable wearable device based on PPG principle is more convenient, can also make the sleep state of the patient more natural, the data collected is closer to the true value, the timeliness is high, the professional requirement is low, the use cost is relatively low, the patient is also more easily accepted, and it is beneficial to popularization and application.

[0042] 2. The application adopts multiple parameter indexes to judge and measure sleep apnea events, compared with judging apnea events by only a single parameter index, the judgment standard is more detailed and detailed, the balance is good, the accuracy is higher, and the medical result judged is more reliable.

[0043] 3. The calculation formula for judging the apnea event in the sleep state is based on weight, the overall calculation method is simple, and it is basically common four arithmetic operations without covariance, calculus and other calculation processes, the calculation process is simple, the understanding difficulty is low, the threshold of practitioners is reduced, the popularization and science popularization are easier, the popularization in non-patient groups is facilitated, and the medical health consciousness of the whole people is improved. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0045] Figure 1 Flow chart for the process of measuring the number of sleep apnea in the present application. DETAILED DESCRIPTION

[0046] In order to make the technical problems to be solved by the present application, technical solutions and beneficial effects more clearly, the following will further describe the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0047] A sleep apnea number measurement method based on multiple parameters, records the sleep time, and respectively obtains the heart rate, blood oxygen concentration and micro-movement acceleration of the human body during sleep, intercepts the time period when the blood oxygen concentration is below the threshold and then returns to normal as an analysis time period, according to the state of the human body, gives different weight values to the heart rate, blood oxygen concentration and micro-movement acceleration in the analysis time period to form heart rate weight, blood oxygen weight and acceleration weight, and obtains sleep stage weight according to sleep time, when the comprehensive index of heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight exceeds the set calculation threshold, the analysis time period is regarded as one apnea event and marked, and finally the total number of marks in the sleep time is counted.

[0048] 1. Blood oxygen concentration:

[0049] In the present application, blood oxygen concentration can be detected by PPG (Photoplethysmography) equipment. PPG equipment usually obtains information about blood oxygen concentration by observing the light after the detection light passes through the blood vessels. Blood oxygen concentration is reflected in the oxygen content of hemoglobin in blood, and the oxygen content of hemoglobin affects the light transmittance of hemoglobin, or blood, resulting in different feedback of blood with different blood oxygen concentrations to the detection light, thereby obtaining the blood oxygen concentration.

[0050] According to clinical experience and standards, the normal value of blood oxygen saturation (SpO2) should not be lower than 94%, and if it is lower than that, it means insufficient supply. When the blood oxygen saturation is lower than 90%, it can be determined as a low blood oxygen state. In actual monitoring, apnea will cause the blood oxygen saturation to decrease to below 90%, and non-pathological reasons will cause the blood oxygen saturation to decrease more greatly, and the difference between the two cases is obvious.

[0051] According to the clinical observation value of low blood oxygen concentration and the practical application factors, the blood oxygen concentration value triggering analysis is set to 95% in the application. When the blood oxygen concentration is lower than 95%, it is considered to enter the low blood oxygen state, and the time point is marked as the starting point of the analysis time period. When the blood oxygen concentration is recovered and equal to 95% again, the time point is marked as the end point of the analysis time period.

[0052] After determining the position of the analysis time period, the blood oxygen concentration is divided into a plurality of low blood oxygen concentration range intervals. In the embodiment, the low blood oxygen concentration range intervals are divided in a uniform manner, and the specific division intervals are as follows:

[0053] The first blood oxygen concentration interval: blood oxygen saturation ∈ [85, 95).

[0054] The second blood oxygen concentration interval: blood oxygen saturation ∈ [80, 85).

[0055] The third blood oxygen concentration interval: blood oxygen saturation ∈ (-∞, 80).

[0056] Since the blood oxygen concentration is a continuous change, when calculating the blood oxygen weight, the blood oxygen saturation used for calculation and the blood oxygen saturation for dividing the low blood oxygen concentration range interval are all the minimum values in the analysis time period, so as to accurately represent the decline amplitude of the blood oxygen concentration when the low blood oxygen event occurs.

[0057] Each low blood oxygen concentration range interval corresponds to a weight value. When calculating the blood oxygen weight only by using the blood oxygen saturation, the decline amplitude of the blood oxygen saturation is directly multiplied by the weight value of the low blood oxygen concentration range interval corresponding to the minimum blood oxygen concentration.

[0058] In the symptoms of apnea, the duration of the low blood oxygen state is positively correlated with the time of the blocked respiratory tract, so the length of the low blood oxygen state can also reflect the symptoms of apnea. The low blood oxygen duration threshold is set in advance. When the low blood oxygen state, that is, the length of the analysis time period, exceeds the set low blood oxygen duration threshold, it means that the human body is in a long-term low blood oxygen state, which is harmful to the human body. Therefore, no matter which low blood oxygen concentration range interval the blood oxygen saturation is in, the weight value is calculated in the third blood oxygen concentration interval, that is, the low blood oxygen concentration range interval corresponding to the minimum blood oxygen concentration of the analysis time period is the third blood oxygen concentration interval when the length of the analysis time period exceeds the set low blood oxygen duration threshold.

[0059] When calculating the blood oxygen weight, the low blood oxygen time weight formed by the low blood oxygen state is superimposed with the weight corresponding to the blood oxygen concentration, so as to obtain the blood oxygen weight. The specific calculation formula of the blood oxygen weight is:

[0060] SPO2_Weight = (95 - SP02_min) x k1 + t x k2 + b,

[0061] wherein SPO2_Weight is blood oxygen weight, SP02_min is the minimum blood oxygen saturation in the analysis time period, k1 is the weight value of the blood oxygen concentration interval corresponding to the minimum blood oxygen saturation, t is the length of the analysis time period, k2 is the analysis time period weight value, and b is an arbitrary correction value.

[0062] The above blood oxygen weight calculation formula is a two-parameter calculation method of superimposing blood oxygen concentration and low blood oxygen state duration. Any one of the related contents can be deleted according to requirements to change to a single parameter calculation method. For example, the blood oxygen weight calculation formula is SPO2_Weight = (95 - SP02_min) x k1 + b, which only judges the blood oxygen weight according to the blood oxygen concentration, and takes the drop amplitude of the blood oxygen concentration in the current low blood oxygen event as the parameter index of the blood oxygen weight; the blood oxygen weight calculation formula is SPO2_Weight = t x k2 + b, which only judges the blood oxygen weight according to the low blood oxygen state time, and takes the low blood oxygen state duration in the current low blood oxygen event as the parameter index of the blood oxygen weight. It can be arbitrarily split according to requirements.

[0063] 2. Heart rate:

[0064] The embodiment adopts a PPG device to obtain human heart rate data. The detection light of the PPG is mostly infrared light. Compared with other wavelengths of light, infrared light has better penetration in muscles and can detect deeper muscle tissue, that is, it is easier to penetrate the sensor to capture. In addition, infrared light can be emitted through an LED lamp, and the entire system can be powered by a low voltage through a polymer lithium battery, combined with a DC-DC boost circuit to realize stable output and ensure the stability of the detection light of the light source. The sensor of the PPG device exists in two forms of contact type and non-contact type. The contact type sensor is mostly loaded on a wristband device such as a smart watch. Such a device is usually also provided with a light source, which is attached to the human skin, and then the sensor receives the light transmitted or reflected by the human body, which is captured by the contact type sensor, so as to obtain the related heart rate data. The non-contact sensor is commonly a high-definition camera. The high-definition camera captures the light intensity change of parts such as face, ear, finger, etc. with dense blood vessels and thin skin, so as to obtain the corresponding heart rate data.

[0065] According to clinical experience and standards, and experimental results data analysis of human night apnea, in the analysis period, the minimum heart rate, the maximum heart rate and the average heart rate during apnea are less than the values during apnea recovery, and the heart rate drop amplitude is positively correlated with the severity of apnea symptoms. After determining the process of heart rate drop and recovery in the analysis period, the minimum heart rate and the maximum heart rate in the analysis period are extracted, and the heart rate weight is calculated. The specific heart rate weight calculation formula is:

[0066] HR_Weight=(HR_END-HR_START)×k3+b,

[0067] Wherein, HR_Weight is the heart rate weight, HR_END is the maximum heart rate, HR_START is the minimum heart rate, k3 is the weight value corresponding to the heart rate, and b is an arbitrary correction value.

[0068] 3. Micro-motion acceleration:

[0069] Micro-motion acceleration is mainly obtained through an acceleration sensor. In this embodiment, a three-axis acceleration sensor is used to collect acceleration data of the human body, and a three-dimensional data change is regarded as a body movement.

[0070] An acceleration threshold is set in advance. When micro-motion acceleration is used as a parameter index for calculation, after the analysis period is intercepted, it is read whether there is acceleration data lower than the acceleration threshold in the acceleration sensor data in this period. If there is acceleration data lower than the acceleration threshold in the analysis period, the acceleration weight is calculated according to the following formula. The acceleration weight calculation formula is:

[0071] Gesensor_Weight=(S-SP02_min)×k4+b,

[0072] Wherein, Gesensor_Weight is the acceleration weight, S is the acceleration threshold corresponding to the blood oxygen saturation in the analysis period, SP02_min is the minimum blood oxygen saturation in the analysis period, k4 is the weight value corresponding to the acceleration, and b is an arbitrary correction value.

[0073] There is a certain linear relationship between acceleration and minimum blood oxygen saturation,

[0074] 4. Sleep stage:

[0075] Through a large number of researches on human pulse data, to a certain extent, pulse rate variability (PRV) and HRV (a kind of electrocardiogram result, i.e. heart rate variability) can accurately reflect the activity of the autonomic nervous system.

[0076] Apnea is a phenomenon that occurs during sleep, and the variability of pulse rate changes in different stages of sleep. The different stages of sleep can be divided into rapid eye movement, light sleep and deep sleep by the variability of pulse rate. Through experiments, the deep sleep stage time of people with sleep apnea syndrome is shorter than that of ordinary people, and the blood oxygen saturation value is higher when apnea occurs in the deep sleep stage than in the other two sleep stages. After analyzing the test data, the frequency of apnea gradually increases in the order of rapid eye movement-light sleep-deep sleep, and the decrease in blood oxygen saturation is also gradually reduced. Therefore, the weight value of the respiratory stage should gradually decrease with the increase of the number of apnea in a complete sleep sampling period, that is, the number of apnea is negatively related to the weight value of the sleep stage, and each sleep stage should have different weight values.

[0077] In another different sleep stage division method, similarly, different sleep stages should correspond to different weight values, and the weight value corresponding to the sleep stage decreases continuously with the deepening of the sleep stage, that is, the increase of sleep time.

[0078] In this embodiment, after the analysis time period is intercepted, the sleep stage in which the analysis time period is located is determined according to the sleep time, and in another embodiment, the sleep stage can be determined by other external devices for determining sleep stage to obtain more accurate physiological state. The corresponding weight value is obtained according to the determined sleep stage, and the sleep stage weight is calculated according to the following formula:

[0079] SLEEP_Weightt=-k5×X+b,

[0080] Wherein, SLEEP_Weightt is the sleep stage weight, k5 is the weight value corresponding to the sleep stage in which the analysis time period is located, X is the number of marked apnea events, and b is an arbitrary correction value set.

[0081] In this embodiment, as shown in Figure 1 the process of measuring the apnea event during sleep includes:

[0082] Step S1. Determine whether the detected object is asleep. If asleep, mark the sleep time point, start timing the sleep time, and record the human body data; if awake, mark the wake-up time point, calculate the sleep duration, enter the apnea event measurement process, go to step S2, and summarize the number of all marked apnea events to obtain the number of apnea events during sleep.

[0083] Step S2. Obtain all parameter data for measuring apnea during sleep, including micro-motion acceleration, heart rate and blood oxygen saturation.

[0084] The blood oxygen saturation and heart rate of the detected object are acquired by the PPG device. The wearable PPG device is similar to a smart watch and a smart bracelet, and can detect and acquire the heart rate and blood oxygen saturation when worn on the wrist during sleep. The micro-motion acceleration is achieved by setting a three-axis acceleration sensor, and the wearing mode is similar to that of the PPG device.

[0085] Step S3. Collecting data processing.

[0086] In this process, the quality of the data collected by the PPG device is mainly analyzed. Since the PPG device collects relevant data by detecting light, it is easily affected by environmental light and other external factors such as mechanical interference during the collection process, resulting in certain quality problems in the data collected by the PPG device. Therefore, the data collected by the PPG device needs to be processed.

[0087] Generally, the collected data is filtered to screen out interference caused by external light and the like, and only the waveform data meeting the required frequency is retained, and the high-frequency noise signal and low-frequency baseline drift signal irrelevant to the human body signal are removed.

[0088] In addition, the PPG signal is an optical signal, and there are many interference factors for such signals. Therefore, the collected signal may cause abnormal calculation of blood oxygen saturation, such as instantaneous decrease of the collected blood oxygen saturation to below 80%. In this application, such calculation abnormality is not discussed.

[0089] Step S4. Extracting the time point at which the blood oxygen saturation is lower than the low blood oxygen threshold value, recording it as a low blood oxygen event, and judging whether the blood oxygen saturation in the low blood oxygen event has recovered to the normal level after the low peak. If the blood oxygen saturation has recovered to the normal level, the time period from when the blood oxygen saturation is lower than the low blood oxygen threshold value to when the blood oxygen saturation recovers to the normal level is defined as an analysis time period, the low blood oxygen concentration range interval to which the low peak value of the blood oxygen saturation in the analysis time period belongs is judged, and it is judged whether the low blood oxygen is caused by non-pathological reasons; if the blood oxygen concentration has not recovered to the normal level after the low peak, the low blood oxygen event caused by the analysis time period is caused by non-pathological reasons.

[0090] According to the time point of the blood oxygen saturation being lower than the low blood oxygen threshold as the analysis time period judgment cut-in point, whether the blood oxygen saturation can recover to the original normal level is judged. If it can recover to the original normal level, the cause of this low blood oxygen event may be caused by apnea. Since there is a non-pathological reason for the decrease in blood oxygen concentration, the phenomenon of the decrease in blood oxygen concentration caused by the non-pathological reason is very obvious, and it is not the data to be measured by the present application, so it is excluded. The exclusion basis is that when a certain low blood oxygen peak of the blood oxygen saturation belongs to the third blood oxygen concentration interval, and the duration of this low blood oxygen event is obviously longer than the average duration of a normal apnea event, it can be judged that the decrease in blood oxygen concentration is caused by a non-pathological reason, and is not counted in the measurement times. If the blood oxygen saturation cannot recover to the original normal level, it is obvious that the cause of this low blood oxygen event is not a pathological cause, and is not counted in the measurement times.

[0091] Regarding the judgment of pathological and non-pathological reasons, the acceleration change of the blood oxygen saturation and the heart rate measured by PPG can be considered comprehensively.

[0092] Step S5. Judge whether this low blood oxygen event is an apnea event, if yes, mark this low blood oxygen event as an apnea event, if not, continue to analyze the next low blood oxygen event. After completing all the low blood oxygen event analysis, count the number of marked apnea events, and record it as the total number of apnea events in this sleep time.

[0093] From the process of the blood oxygen saturation being lower than the low blood oxygen threshold until recovering to the normal level, the analysis time period is set, and after step S4, this low blood oxygen event is caused by a pathological reason, so the heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight are obtained from the blood oxygen concentration, heart rate, micro-movement acceleration and sleep stage respectively.

[0094] The blood oxygen concentration weight is obtained. The minimum blood oxygen saturation in the analysis time period is obtained, the low blood oxygen concentration range interval where it is located is judged, the corresponding weight value is obtained, and the duration of the analysis time period is counted to obtain the set analysis time period weight value. The blood oxygen weight SPO2_Weight is obtained through the blood oxygen weight calculation formula, and the blood oxygen weight calculation formula is SPO2_Weight=(95-SP02_min)×k1+t×k2+b.

[0095] The heart rate weight is obtained. The minimum heart rate, maximum heart rate and fixed heart rate weight value in the analysis time period are obtained, and the heart rate weight HR_Weight is obtained through the heart rate weight formula, and the heart rate weight calculation formula is HR_Weight=(HR_END-HR_START)×k3+b.

[0096] The acceleration weight is obtained when the obtained acceleration is lower than the acceleration threshold value, the minimum blood oxygen saturation in the analysis time period is obtained, and the blood oxygen saturation corresponding to the change point of the acceleration drop and being lower than the acceleration threshold value is obtained, the corresponding weight value is collected, the acceleration weight Gesensor_Weight is calculated according to the acceleration weight formula, and the acceleration weight formula is Gesensor_Weight=(S-SP02_min)×k4+b.

[0097] The sleep stage weight is obtained according to the sleep state judged by the sleep time length or the external device, the sleep stage in which the analysis time period is in is judged, the time at the starting point of the analysis time period is taken as the judgment basis, the weight value corresponding to the corresponding sleep stage is obtained, the number of recorded apnea events is counted, the sleep stage weight is obtained through the sleep stage weight calculation formula, and the sleep stage weight calculation formula is SLEEP_Weightt=-k5×X+b.

[0098] After the heart rate weight, the blood oxygen weight, the acceleration weight and the sleep stage weight are obtained, the heart rate weight, the blood oxygen weight, the acceleration weight and the sleep stage weight are comprehensively calculated, and the calculation threshold value is compared, in the embodiment, the sum of the heart rate weight, the blood oxygen weight, the acceleration weight and the sleep stage weight is calculated, the calculation threshold value is 1, that is

[0099] When SPO2_Weight+HR_Weight+Gesensor_Weight+SLEEP_Weight>1, it is judged that the low blood oxygen event is an apnea event, and is marked, all the marked low blood oxygen events are counted, and the total number of apnea events in the sleep period is obtained.

[0100] The above is only the preferred embodiment of the present application, and does not limit the present application, any modification, equivalent replacement and improvement within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for measuring sleep apnea frequency based on multiple parameters, characterized in that: The time of falling asleep is recorded, and the heart rate, blood oxygen concentration, and micro-acceleration of the human body during sleep are obtained respectively. The analysis time period is divided according to the blood oxygen concentration. According to the state of the human body, the heart rate, blood oxygen concentration, and micro-acceleration in the analysis time period are given different weights to form heart rate weight, blood oxygen weight, and acceleration weight. The sleep stage weight is obtained according to the time of falling asleep. When the combined index of heart rate weight, blood oxygen weight, acceleration weight, and sleep stage weight exceeds the set calculation threshold, the analysis time period is regarded as an apnea event and marked. Finally, the total number of marks within the sleep time is counted; Set a hypoxemia threshold, collect the blood oxygen saturation of the human body, extract and mark the first instantaneous time point when the blood oxygen saturation is lower than the hypoxemia threshold and the second instantaneous time point after the first instantaneous time point when the blood oxygen saturation is higher than the hypoxemia threshold, and the time period between the first instantaneous time point and the second instantaneous time point is an analysis time period; When calculating the blood oxygen weight, several low blood oxygen concentration range intervals are set, and each low blood oxygen concentration range interval corresponds to a different weight value; Record the minimum blood oxygen concentration during the analysis period, determine the low blood oxygen range interval of the analysis period based on the minimum blood oxygen concentration, determine the weight value corresponding to the current analysis period, and combine the minimum blood oxygen concentration with the weight value of the corresponding low blood oxygen range interval to obtain the blood oxygen weight; According to the time of falling asleep, the sleep stage of the analysis period is determined and the weight value of the corresponding sleep stage is obtained. The sleep stage weight is obtained by combining the number of marked apnea events and the weight value.

2. The method for measuring sleep apnea frequency based on multiple parameters according to claim 1, characterized in that: The length of the analysis time period is recorded. When calculating the blood oxygen weight, the weight value of the analysis time period is assigned to form the hypoxic time weight. The hypoxic time weight and the weight corresponding to the minimum blood oxygen concentration are superimposed to obtain the blood oxygen weight.

3. The method for measuring sleep apnea frequency based on multiple parameters according to claim 1, characterized in that: Obtain human heart rate data and determine whether there is a process in which the heart rate decreases and then increases during the analysis period. If so, extract the maximum and minimum heart rates during the analysis period, and combine the maximum and minimum heart rates with the corresponding weight values ​​to obtain the heart rate weight.

4. The method for measuring sleep apnea frequency based on multiple parameters according to claim 1, wherein: An acceleration threshold is set to obtain human micro-motion acceleration data. When the micro-motion acceleration is less than the acceleration threshold, the acceleration weight is formed by combining the acceleration threshold, the minimum blood oxygen concentration in the analysis time period, and the corresponding weight value.

5. The method for measuring sleep apnea frequency based on multiple parameters according to claim 1, characterized in that: The process for measuring apnea events during sleep includes: Step S1. Determine whether the subject is asleep. If so, mark the time of falling asleep, start counting the time of falling asleep, start all data collection devices, and record human body data. If the subject is awake, mark the time of waking up, calculate the sleep duration, and proceed to step S2. Step S2. Acquire all parameter indicators for measuring apnea during sleep, including micro-motion acceleration, heart rate, and blood oxygen saturation; Step S3. Collect data and process it; Step S4. Extract the time point when the blood oxygen saturation falls below the hypoxemia threshold and record it as a hypoxemia event. Determine whether the hypoxemia event is caused by a non-pathological cause. If so, continue to extract the next time point when the blood oxygen saturation falls below the hypoxemia threshold for analysis. If not, extract the collected data within the analysis time period corresponding to the hypoxemia event. Step S5. Determine whether the hypoxemia event is an apnea event. If so, mark the hypoxemia event as an apnea event. If not, continue analyzing the next hypoxemia event. After completing the analysis of all hypoxemia events, count the number of marked apnea events and record it as the total number of apnea events during this sleep period.

6. The method for measuring sleep apnea frequency based on multiple parameters according to claim 5, characterized in that: Low blood oxygen concentration range division: First blood oxygen concentration interval: blood oxygen saturation ∈ [85, 95); Second blood oxygen concentration interval: blood oxygen saturation ∈ [80, 85); The third blood oxygen concentration interval: blood oxygen saturation ∈ (-∞, 80); In step S4, it is determined whether the blood oxygen saturation has returned to a normal level after the low peak in this hypoxemia event. If it has returned to a normal level, the time period from when the blood oxygen saturation is below the hypoxemia threshold to when the blood oxygen saturation returns to a normal level is defined as the analysis time period. The low blood oxygen concentration range interval to which the low peak value of the blood oxygen saturation in the analysis time period belongs is determined: If the low peak value of blood oxygen saturation falls within the third blood oxygen concentration interval and the duration of the analysis period is longer than the average duration of a normal apnea event, the cause of the hypoxemia event is determined to be non-pathological; If the low peak value of blood oxygen saturation does not belong to the third blood oxygen concentration range, the cause of the hypoxemia event is determined to be pathological; If the duration of the analysis period is less than or equal to the average duration of a normal apnea event, the cause of the hypoxemia event is considered to be pathological; If the blood oxygen concentration does not return to normal levels after the low peak, the cause of the hypoxemia event is determined to be non-pathological.

7. The method for measuring sleep apnea frequency based on multiple parameters according to claim 5, characterized in that: In step S5, the heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight are obtained, and the sum of the heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight is calculated. The calculation threshold is set to 1. When the sum of the heart rate weight, blood oxygen weight, acceleration weight and sleep stage weight is greater than 1, it is judged that the hypoxic event is a sequential apnea event.

Images