Abnormal breathing event detection method, device and equipment and storage medium

Abstract

The invention discloses an abnormal breathing event detection method and device, equipment and a storage medium. The method comprises the following steps: acquiring an original vibration signal, wherein the original vibration signal comprises a vibration signal generated by a user and collected by a vibration sensor; preprocessing the original vibration signal to obtain a respiration signal in a required bandwidth range; generating an amplitude time sequence waveform corresponding to the respiratory signal; performing amplitude analysis on the amplitude time sequence waveform, and determining a baseline value and a drop value of the amplitude time sequence waveform; and based on the baseline value and the drop value, determining whether an abnormal breathing event exists in the original vibration signal. The breathing condition of the user in the sleep period can be noninductively detected, and the accuracy of abnormal breathing event detection is improved.

Description

Technical Field

[0001] This invention relates to the field of respiratory detection technology, and in particular to a method, apparatus, device, and storage medium for detecting abnormal respiratory events. Background Technology

[0002] Sleep apnea and hypoventilation are two common sleep disorders that affect sleep patterns and respiratory function, significantly impacting sleep quality and causing certain individuals to snore loudly at night, feel tired, irritable, sleepy during the day, and experience a variety of potentially more serious complications.

[0003] Polysomnography (PSG) is the gold standard for diagnosing sleep apnea / hypopnea and plays a crucial role in guiding effective management strategies to improve sleep quality and overall health in individuals with sleep-disordered breathing. However, PSG typically requires patients to undergo a full night of sleep monitoring at a specialized facility, incurring travel and time costs. PSG testing involves attaching multiple sensors to the patient's head, chest, abdomen, and legs, which can easily create a physiological burden, causing the "first night effect"—difficulty falling asleep or poor sleep—significantly impacting sleep quality and leading to biased or inaccurate test results. PSG testing is expensive and requires specialized equipment and professional personnel. After the test, sleep stages and respiratory events need to be manually scored and interpreted, which is very time-consuming and labor-intensive. Despite the large population suffering from sleep-disordered breathing, the high time cost, complex process, and relatively high cost of PSG testing significantly reduce patients' willingness to seek professional treatment. Interpreting respiratory events through PSG mainly relies on nasal and oral airflow, chest and abdominal breathing, blood oxygenation, and electroencephalogram (EEG) information, with decreased blood oxygenation being a key indicator for identifying hypopnea events. Relying solely on vibration sensors, without airflow and blood oxygenation information, significantly increases the difficulty of identifying hypopnea. Therefore, existing technologies often incorporate blood oxygenation or snoring detection to enhance accuracy through multimodal data analysis. However, adding blood oxygenation detection requires wearing a photoelectric sensor; even with lightweight designs like rings, earrings, or other forms, it's not truly imperceptible. Adding snoring detection involves microphone sensors, and capturing snoring may involve sensitive user information, potentially leading to privacy breaches and user resistance. Therefore, developing a non-invasive, home-based method and device for detecting sleep disturbances is crucial. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and storage medium for detecting abnormal breathing events during sleep, enabling non-intrusive, home-based detection of abnormal breathing events while improving the accuracy of such detection.

[0005] According to one aspect of the present invention, a method for detecting abnormal respiratory events is provided, comprising:

[0006] Acquire raw vibration signals, which include vibration signals generated by the user collected by vibration sensors;

[0007] The original vibration signal is preprocessed to obtain the breathing signal within the required bandwidth range;

[0008] Generate the amplitude-time waveform corresponding to the respiratory signal;

[0009] Amplitude analysis is performed on the amplitude timing waveform to determine the baseline value and drop value of the amplitude timing waveform;

[0010] Based on the baseline value and the drop value, determine whether there is an abnormal breathing event in the original vibration signal.

[0011] According to another aspect of the present invention, a respiratory abnormality event detection device is provided, comprising:

[0012] The acquisition module is used to acquire the raw vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor;

[0013] The preprocessing module is used to preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range;

[0014] The generation module is used to generate the amplitude-time waveform corresponding to the respiratory signal;

[0015] The first determining module is used to perform amplitude analysis on the amplitude time-series waveform to determine the baseline value and drop value of the amplitude time-series waveform;

[0016] The second determining module is used to determine whether there is an abnormal breathing event in the original vibration signal based on the baseline value and the drop value.

[0017] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0018] At least one processor; and

[0019] A memory communicatively connected to the at least one processor; wherein,

[0020] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method described in any embodiment of the present invention.

[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the method described in any embodiment of the present invention.

[0022] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the method described in any embodiment of the present invention.

[0023] The technical solution of this invention acquires the original vibration signal and collects the vibration signal generated by the user imperceptibly through a vibration sensor; preprocesses the original vibration signal to obtain a breathing signal within the required bandwidth range, eliminating noise interference in the original vibration signal; generates an amplitude-time sequence waveform corresponding to the breathing signal to characterize the user's breathing status in time sequence, providing a basis for detecting abnormal breathing events, and performs processing only in the time domain, reducing computational resource overhead; performs amplitude analysis on the amplitude-time sequence waveform to determine the baseline value and drop value of the amplitude-time sequence waveform, and determines whether there is an abnormal breathing event in the original vibration signal based on the baseline value and the drop value. This allows for the acquisition and analysis of vibration signals during the user's sleep based on a vibration sensor, enabling imperceptible home-based detection of abnormal breathing events during sleep, while improving the accuracy of abnormal breathing event detection.

[0024] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a waveform diagram of the original vibration signal and the filtered original vibration signal during a sleep apnea event, provided in an embodiment of the present invention.

[0027] Figure 2 This is a waveform diagram of the respiratory waveform within the time window provided in the embodiments of the present invention;

[0028] Figure 3 This is a flowchart of a respiratory abnormality event detection method provided in Embodiment 1 of the present invention;

[0029] Figure 4This is a waveform diagram of the original vibration signal provided in an embodiment of the present invention;

[0030] Figure 5 This is a waveform diagram of a filtered and denoised respiratory signal provided in an embodiment of the present invention;

[0031] Figure 6 This is a schematic diagram of the respiratory signal and the generated amplitude timing waveform provided in an embodiment of the present invention;

[0032] Figure 7 This is a flowchart of a respiratory abnormality event detection method provided in Embodiment 2 of the present invention;

[0033] Figure 8 This is a schematic diagram of determining the baseline region based on amplitude-time waveform provided in an embodiment of the present invention;

[0034] Figure 9 This is a schematic diagram of determining the drop region based on amplitude-time waveform provided in an embodiment of the present invention;

[0035] Figure 10 This is a schematic diagram illustrating the determination of the baseline region and the drop region based on the amplitude-time waveform provided in an embodiment of the present invention;

[0036] Figure 11 This is a flowchart of a respiratory abnormality event detection method provided in Embodiment 3 of the present invention;

[0037] Figure 12 This is a schematic diagram of the percentage drop in low ventilation amplitude provided in an embodiment of the present invention;

[0038] Figure 13 This is a schematic diagram of obtaining the time-corrected amplitude waveform based on respiratory width according to an embodiment of the present invention;

[0039] Figure 14 This is a schematic diagram of the structure of a respiratory abnormality event detection device provided in Embodiment 4 of the present invention;

[0040] Figure 15 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0041] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0042] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0043] Abnormal breathing events during sleep include sleep apnea events and hypopnea events. When a sleep apnea event occurs, the respiratory amplitude decreases significantly, and the waveform characteristics of the respiratory signal are very consistent with the chest and abdominal band waveform characteristics of a traditional PSG. Figure 1 This is a waveform diagram of the original vibration signal and the filtered original vibration signal during a sleep apnea event, provided in an embodiment of the present invention. Figure 1 As shown, Figure 1 The image above shows the raw, unfiltered vibration signal. Figure 1 Below is the original vibration signal after filtering. Figure 1 An obstructive sleep apnea event occurred because the chest cavity still moves during the event, and the reduced amplitude respiratory cycle can be easily identified from the filtered respiratory waveform.

[0044] When a central sleep apnea event occurs, since there is no breathing action, only the heartbeat vibration amplitude is present in the original vibration signal, and the amplitude in the filtered breathing waveform will be very low. Therefore, an extremely low amplitude threshold detection is added for the breathing amplitude. When a sleep apnea event is detected, if the amplitude is lower than this threshold, it can be considered that a central sleep apnea event has occurred. Figure 2 This is a waveform diagram of the breathing waveform within a time window provided in an embodiment of the present invention. There are a total of 7 sleep apnea events within this window, but "Event 4" is a central sleep apnea event. Because there is no breathing action during this event, there is no breathing amplitude on the breathing waveform. Therefore, central sleep apnea events can be easily distinguished by a threshold detection with extremely low amplitude.

[0045] When a hypopnea event occurs, the degree of decrease in the amplitude of the respiratory waveform can affect the judgment of sleep apnea, and blood oxygen saturation analysis is often used as an auxiliary means of judgment. This invention only needs to process the respiratory waveform in the time domain to determine whether an abnormal respiratory event has occurred.

[0046] Example 1

[0047] Figure 3 This is a flowchart of a method for detecting abnormal breathing events according to Embodiment 1 of the present invention. This embodiment is applicable to detecting abnormal breathing events during a user's sleep. The method can be executed by a device for detecting abnormal breathing events, which can be implemented in hardware and / or software and can be configured in an electronic device. The electronic device can be bedding such as a smart mattress or sleep belt, or a terminal such as a computer, mobile phone, or personal digital assistant. Figure 3 As shown, the method includes:

[0048] S110. Acquire the original vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor.

[0049] A vibration sensor is a non-contact, imperceptible sensor that captures physiological vibrations of the human body. In this embodiment, the original vibration signal can be the vibration signal generated by a user lying on bedding, including but not limited to mattresses, bed frames, and sofas. Lying on the bed can be flat or at a certain angle. Generally, a vibration sensor can be placed below the upper body, without direct contact with the body, to imperceptibly collect human vibration signals. The vibration sensor can be placed inside or near the bedding, as long as it can collect vibration signals. The original vibration signal is collected by a vibration sensor mounted in an electronic device. The vibration sensor can be, for example, an acceleration sensor, a pressure sensor, or a displacement sensor, or it can be a sensor of changes in quantities such as radar waves, magnetic fields, or electric fields, converting the changes into micro-motion signals of the body. The original vibration signal can also be collected by a sensor that converts physical quantities based on acceleration, pressure, or displacement (such as a piezoelectric ceramic sensor, an inflatable micro-motion sensor, or a fiber optic sensor). The original vibration signal includes, but is not limited to, human breathing signals, heartbeat signals, body movement signals, and noise. Vibration sensors can be single or multiple; multiple sensors can improve the accuracy and reliability of identification. This invention is described with reference to a single-channel approach, but it can be easily extended to multi-channel solutions, and will not be elaborated further. The acquisition, storage, use, and processing of data in this invention comply with relevant laws and regulations.

[0050] Specifically, the following explanation uses a smart mattress as an example of bedding. When a user is on a smart mattress, the vibration sensor in the smart mattress collects the user's original vibration signal and transmits it to the processor in the electronic device for processing. Figure 4 This is a waveform diagram of the original vibration signal provided in an embodiment of the present invention. The acquired original vibration signal is as follows: Figure 4 As shown, it contains multiple signals.

[0051] In this operation, the original vibration signal can be the vibration signal generated by the user during sleep on the smart mattress. The smart mattress can be a sleep product that integrates at least one technological function, such as an integrated non-contact vibration sensor.

[0052] S120. Preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range.

[0053] In this embodiment, the bandwidth range can be the bandwidth range of the respiratory signal. Since the original vibration signal contains body motion signals, whose bandwidth is much larger than that of the respiratory signal, it is necessary to select an appropriate bandwidth range to filter out noise, such as body motion signals. The bandwidth range can be determined empirically.

[0054] Specifically, the original vibration signal can be preprocessed using filtering methods such as Infinite Impulse Response (IIR) filter, Finite Impulse Response (FIR) filter, wavelet filter, zero-phase bidirectional filter, polynomial fitting smoothing filter, median filter, or mean filter. Other transformations such as differentiation and integration can also be performed. One or more combinations of these methods can be selected to filter the original vibration signal and obtain the breathing signal.

[0055] In one example, preprocessing involves selecting an appropriate bandwidth range (e.g., 0.05Hz–0.5Hz) to filter out body motion signals from the original vibration signal. By scaling the signal appropriately according to its dynamic range, the time-domain waveform of the respiratory signal can be obtained. Figure 5 This is a waveform diagram of a filtered and denoised respiratory signal provided in an embodiment of the present invention. Figure 5 As shown, each waveform has distinct characteristics and good consistency, regular periodicity, clear outline, and stable baseline.

[0056] S130. Generate the amplitude timing waveform corresponding to the respiratory signal.

[0057] In this embodiment, the amplitude-time waveform can be a waveform diagram showing the change of respiratory signal amplitude over time. The amplitude-time waveform can characterize the frequency and depth of respiration.

[0058] Specifically, amplitude timing waveforms can be generated in a variety of ways:

[0059] 1. Set a fixed time window (the window length is set empirically), calculate the deviation between the maximum and minimum values ​​of the respiratory waveform amplitude within each time window, also known as the maximum and minimum value deviation, and perform linear interpolation on the discrete data points to obtain the amplitude time-series waveform;

[0060] 2. Set a variable time window, with the window time mapped to the average respiratory cycle width (e.g., proportional relationship), calculate the maximum and minimum deviation of the respiratory waveform amplitude within the time window, and perform linear interpolation on the discrete data points to obtain the amplitude time-series waveform;

[0061] 3. Search for the peaks and valleys of each cycle of the respiratory signal one by one, perform amplitude mapping by the amplitude of the rising edge or the amplitude of the falling edge, or take the weighted result of the two as the amplitude mapping, and obtain the amplitude time-series waveform by interpolation and smoothing of the obtained discrete data points;

[0062] 4. Search for the peaks and valleys of each respiratory cycle one by one, and calculate the average amplitude by weighting the average amplitude through a preset number of nearby respiratory cycles. Then, interpolate and smooth the obtained discrete data points to obtain the amplitude time-series waveform.

[0063] 5. Through one or more of the above weighted mappings; or rule-based mappings, such as taking the third method when the breathing pattern is stable and good, and taking the first method when it is difficult to search for effective peaks and valleys, such as body movement or no breathing action, discrete data points are obtained, and amplitude time-series waveforms are obtained through interpolation and smoothing.

[0064] For example, Figure 6 This is a schematic diagram of the respiratory signal and the generated amplitude-time waveform provided in an embodiment of the present invention. Figure 6 As shown, Figure 6 The top image shows the filtered respiratory signal, and the bottom image shows the generated amplitude-time waveform. This amplitude-time waveform is generated by interpolating the maximum and minimum deviations of the respiratory waveform amplitude values ​​based on 4-second, 5-second, and 6-second time windows.

[0065] S140. Perform amplitude analysis on the amplitude timing waveform to determine the baseline value and drop value of the amplitude timing waveform.

[0066] In this embodiment, the baseline value can be a stable reference value when breathing is smooth. The baseline value can be associated with the maximum amplitude value of the amplitude-time waveform. The baseline value can represent the user's normal respiratory amplitude benchmark. The drop value can be the value at which the amplitude-time waveform drops significantly from the baseline level. The drop value can be associated with the minimum amplitude value of the amplitude-time waveform. The drop value and the baseline value can jointly characterize the presence and severity of abnormal respiratory events. Each respiratory cycle corresponds to one baseline value and one drop value.

[0067] Specifically, the baseline value of the amplitude time series waveform can be calculated by the statistical value of the amplitude value at or within a certain range of the peak of the amplitude time series waveform, and the drop value of the amplitude time series waveform can be calculated by the statistical value of the amplitude value at or within a certain range of the trough of the amplitude time series waveform.

[0068] S150. Based on the baseline value and the drop value, determine whether there is an abnormal breathing event in the original vibration signal.

[0069] In this embodiment, an abnormal breathing event can be defined as an event in which the amplitude of the breathing signal drops by more than a certain threshold compared to the normal breathing amplitude. Common abnormal breathing events include sleep apnea events and hypopnea events. Abnormal breathing events can be determined by whether the drop in amplitude time-series waveform value exceeds a certain range compared to the baseline value.

[0070] Specifically, based on the relationship between the baseline value and the drop value, it is determined whether there are abnormal respiratory events in the original vibration signal. The relationship between the baseline value and the drop value can be the ratio or difference between the baseline value and the drop value, or other statistics that can reflect the relationship between the two.

[0071] The technical solution of this invention acquires raw vibration signals and collects user-generated vibration signals imperceptibly using a vibration sensor mounted in the mattress. The raw vibration signals are preprocessed to obtain breathing signals within the required bandwidth, eliminating noise interference. An amplitude-time-series waveform corresponding to the breathing signals is generated to represent the user's breathing status in time, providing a basis for detecting abnormal breathing events. Processing is performed only in the time domain, reducing computational resource overhead. Amplitude analysis is performed on the amplitude-time-series waveform to determine its baseline and drop values. Based on the baseline and drop values, it is determined whether abnormal breathing events exist in the raw vibration signals. This invention enables the acquisition and analysis of vibration signals during user sleep using a vibration sensor, allowing for imperceptible, home-based detection of abnormal breathing events during sleep, while improving the accuracy of abnormal breathing event detection.

[0072] Example 2

[0073] Figure 7 This is a flowchart of a respiratory abnormality event detection method provided in Embodiment 2 of the present invention. This embodiment is an optimization based on any of the above embodiments, and mainly includes: determining a baseline region and a drop region, then determining a baseline value and a drop value respectively, determining a drop percentage based on the baseline value and the drop value, and comparing it with a percentage threshold to determine whether a respiratory abnormality event exists. It should be noted that technical details not described in detail in this embodiment can be found in any of the above embodiments. Figure 7 As shown, the method includes:

[0074] S210. Acquire the original vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor.

[0075] S220. Preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range.

[0076] S230, Generate the amplitude timing waveform corresponding to the respiratory signal, and execute S240 and S260.

[0077] S240. Analyze the maximum amplitude range in the amplitude timing waveform to determine the baseline region in the amplitude timing waveform.

[0078] In this embodiment, the baseline region can refer to the area where the amplitude timing waveform is in a stable state. This region can be located within a certain range of the maximum amplitude value of the amplitude timing waveform. There is one baseline region within a respiratory cycle, and the baseline region can be characterized by the start and end positions of the baseline region (including the start position and the end position).

[0079] Specifically, the baseline region can be determined by defining the start and end positions as follows, with the area between the start and end positions being the baseline region:

[0080] 1. The start and end points of the maximum amplitude value within one respiratory cycle in the amplitude time-series waveform are used as the start and end points of the baseline region;

[0081] 2. Use the decrease of X% in the maximum amplitude value within one respiratory cycle in the amplitude time-series waveform as the start and end positions. X can be 10%, 20%, etc.

[0082] 3. Determine the maximum amplitude value within a respiratory cycle in the amplitude time-series waveform, and then use the corresponding peaks, troughs, or preset amplitude quantiles (such as 10% or 20% of the maximum amplitude difference) on the rising or falling edges as the start and end positions of the baseline region.

[0083] 4. Determine the maximum amplitude value within a respiratory cycle in the amplitude time-series waveform, and then take the position corresponding to the decrease of the maximum amplitude value by a specific percentage (X%, such as 10% or 20%), or the specific quantile (N%, such as 10% or 20%) of the peak, trough, rising edge or falling edge of the cycle as the start and end positions of the baseline region.

[0084] 5. The weighted statistics of the start and end positions of at least one of the baseline regions determined above are used to determine the final start and end positions of the baseline region.

[0085] For example, Figure 8 This is a schematic diagram illustrating the determination of the baseline region based on amplitude-time waveforms provided in an embodiment of the present invention. For example... Figure 8 As shown, the top part is a diagram of the respiratory signal waveform, and the bottom part is a schematic diagram of the baseline region determined based on the amplitude-time waveform. Figure 8 There are 6 respiratory cycles in total. The starting and ending points of the baseline region are the positions where the maximum amplitude value decreases by a certain percentage in each cycle. Each respiratory cycle corresponds to one baseline region.

[0086] S250. Based on the amplitude values ​​within the baseline region, determine the baseline value of the amplitude timing waveform, and execute S280.

[0087] Specifically, the baseline value can be determined based on the amplitude values ​​within the baseline area and their statistics. The following methods can be used: the maximum amplitude value, minimum amplitude value, average amplitude value, median amplitude value, or amplitude value with the largest proportion within the baseline area can be used as the baseline value, or the weighted statistic of at least one of the above amplitude values ​​can be used as the baseline value.

[0088] S260. Analyze the minimum amplitude range in the amplitude timing waveform to determine the drop region in the amplitude timing waveform.

[0089] In this embodiment, the drop region can be a region in the amplitude timing waveform where the amplitude drops and persists for a period of time. The drop region can be determined based on the minimum amplitude value in the amplitude timing waveform or a certain range around the minimum amplitude value. Each respiratory cycle corresponds to one drop region, which can be characterized by the start and end positions of the drop region.

[0090] Specifically, the fall zone is determined based on the range of the minimum amplitude value in each respiratory cycle. This can be achieved by determining the start and end points as follows: the area between these points is the fall zone.

[0091] 1. The start and end points of the fall zone are determined by the start and end points of the minimum amplitude value in each respiratory cycle;

[0092] 2. The starting and ending points of the drop zone are defined as the points where the minimum amplitude value increases by X% in each respiratory cycle. X can be 10%, 20%, etc.

[0093] 3. Determine the minimum amplitude value corresponding to each respiratory cycle, and use it, or a specific amplitude quantile (e.g., 10%, 20%) of the peak, trough, rising edge, or falling edge within that cycle, as the start and end points of the drop zone.

[0094] 4. Determine the minimum amplitude value in each respiratory cycle, and the position corresponding to when it rises by a specific percentage (X%, such as 10% or 20%), or the specific amplitude quantiles (such as 10% or 20%) of the peak, trough, rising edge or falling edge in that cycle, as the start and end positions of the falling area.

[0095] 5. The weighted statistics of the start and end positions of at least one of the above-determined fall areas are used to determine the final start and end positions of the fall areas.

[0096] For example, Figure 9 This is a schematic diagram illustrating the determination of the drop region based on amplitude-time waveforms provided in an embodiment of the present invention. For example... Figure 9As shown, Figure 9 The image above shows a respiratory waveform. Figure 9 The area below is the drop region determined based on the amplitude timing waveform. Figure 9 It includes 4 breathing cycles, with the position corresponding to the rise of a specific percentage in each breathing cycle serving as the start and end position of the fall zone.

[0097] S270. Based on the amplitude value within the drop region, determine the drop value of the amplitude timing waveform.

[0098] Specifically, the drop value is determined based on the amplitude values ​​within the drop area and its statistical measures. This can be done by using the maximum amplitude value, minimum amplitude value, average amplitude value, median amplitude value, or the amplitude value with the highest proportion within the drop area as the drop value, or by weighting the statistical measures of at least one of the above amplitude values ​​and mapping the weighted statistical measures to the drop value.

[0099] For example, Figure 10 This is a schematic diagram illustrating the determination of the baseline region and drop region based on amplitude-time waveforms provided in an embodiment of the present invention. For example... Figure 10 As shown, Figure 10 The top part shows the waveform of the respiratory signal, and the bottom part shows the baseline region and drop region determined based on the amplitude-time waveform. Each respiratory cycle corresponds to a baseline region and a drop region. The maximum amplitude value in the baseline region is used as the baseline value, and the minimum amplitude value in the drop region is used as the drop value.

[0100] S280. Determine the percentage drop based on the baseline value and the drop value.

[0101] In this embodiment, the drop percentage can be a numerical value used to quantify the degree of decline in respiratory amplitude. The drop percentage can be expressed as the degree of decrease in the drop value relative to the baseline value, and the drop percentage can be used to determine whether an abnormal respiratory event has occurred.

[0102] Specifically, first calculate the ratio of the drop value to the baseline value, and then subtract the percentage of this ratio from 1 to get the drop percentage.

[0103] S290. Based on the comparison results of the drop percentage and the percentage threshold, determine whether there is an abnormal breathing event in the original vibration signal.

[0104] In this embodiment, the percentage threshold can be a threshold used to determine whether an abnormal respiratory event exists. The percentage threshold can be determined by the following methods: the percentage threshold can be determined empirically as a fixed value; the percentage drop of the most recent identified abnormal respiratory event can be used as the percentage threshold, or it can be weighted with a fixed threshold to obtain the percentage threshold; the percentage threshold can be used as a statistical measure (e.g., the maximum, minimum, mean, or median of the percentage threshold of at least one most recent identified abnormal respiratory event), or it can be weighted with a fixed threshold to obtain the percentage threshold; or the weighted mapping of the above at least one percentage threshold can be used to determine the final percentage threshold.

[0105] Specifically, the percentage of the fall is compared with a percentage threshold. If the percentage of the fall is greater than the percentage threshold, then a respiratory abnormality event is determined to have occurred at this time.

[0106] For example, Table 1 shows the baseline values, drop values, and drop percentages for four respiratory abnormalities provided in this embodiment of the invention. At this point, the percentage threshold is 90%, and the drop percentages for all four respiratory cycles are greater than the percentage threshold, thus confirming that a respiratory abnormality event has occurred.

[0107] Table 1. Baseline values, drop values, and drop percentages for four respiratory abnormalities.

[0108]

[0109] The technical solution of this invention involves acquiring an original vibration signal, preprocessing the original vibration signal to obtain a respiratory signal within a required bandwidth range, generating an amplitude-time waveform corresponding to the respiratory signal, analyzing the maximum amplitude range in the amplitude-time waveform to determine a baseline region, and determining the baseline value of the amplitude-time waveform based on the amplitude values ​​within the baseline region. Determining the baseline value by identifying the baseline region avoids interference in the amplitude-time waveform and improves the accuracy of the baseline value. The invention also involves analyzing the minimum amplitude range in the amplitude-time waveform to determine a drop region, and determining the drop value of the amplitude-time waveform based on the amplitude values ​​within the drop region. Determining the drop value by identifying the drop region avoids noise interference and improves the accuracy of the drop value. Finally, the invention determines a drop percentage based on the baseline value and the drop value, and determines whether an abnormal respiratory event exists in the original vibration signal based on a comparison between the drop percentage and a percentage threshold. The presence of an abnormal respiratory event is determined by judging the degree of decrease in the drop value compared to the baseline value, thus improving the accuracy of abnormal respiratory event detection.

[0110] Example 3

[0111] Figure 11This is a flowchart of a respiratory abnormality event detection method provided in Embodiment 3 of the present invention. This embodiment is an optimization based on any of the above embodiments, and mainly includes: a detailed description of the process of generating a width time-series waveform, updating the amplitude time-series waveform to obtain an updated amplitude time-series waveform, and determining the baseline value and drop value through the updated amplitude time-series waveform. It should be noted that technical details not described in detail in this embodiment can be found in any of the above embodiments. Figure 11 As shown, the method includes:

[0112] S310. Acquire the original vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor.

[0113] S320. Preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range.

[0114] S330. Generate the amplitude timing waveform corresponding to the respiratory signal.

[0115] S340. Generate the width-time waveform corresponding to the respiratory signal.

[0116] In this embodiment, the width-time waveform can be a waveform that describes the change in the duration of each respiratory cycle. Using a width-time waveform, the phenomenon of increased respiratory rate at the end of a hypoventilation event can be reflected through changes in respiratory rhythm.

[0117] Specifically, the method for generating a wide time series waveform can be as follows: First, set a time window of fixed length, and take the maximum, minimum, average or median of the respiratory cycle within the time window. Calculate the statistics for each respiratory cycle using a sliding window, connect the obtained discrete data through linear interpolation and smooth them to obtain the wide time series waveform.

[0118] For example, Figure 12 This is a schematic diagram illustrating the percentage drop in low ventilation amplitude provided in an embodiment of the present invention. For example... Figure 12 As shown, a hypoventilation event occurred at this time. Figure 12 The topmost graph shows the change in blood oxygen saturation over time. Figure 12 The second image from the top is a thermal airflow breathing waveform diagram. Figure 12 The third image from the top is a diagram of the pressure airflow breathing waveform. Figure 12 The fourth image from the top in the middle is the waveform of the filtered respiratory signal. Figure 12The bottom of the image shows the amplitude-time waveform obtained from the respiratory signal. According to the hypoventilation criteria, a hypoventilation event occurs when the pressure-airflow amplitude decreases by more than 30%, lasts for more than 10 seconds, and the blood oxygen saturation drops from 91% to 88% (≥3%). However, when calculating the percentage drop from the baseline amplitude to the drop amplitude using the amplitude-time waveform, it is only 43.49%. For respiratory events with such a percentage drop within a certain preset range, the original amplitude-time waveform needs to be updated using a width-time waveform.

[0119] S350. Update the amplitude timing waveform based on the width timing waveform to obtain the updated amplitude timing waveform.

[0120] Specifically, the updated amplitude timing waveform is obtained by the ratio of the amplitude waveform to the width waveform.

[0121] Optionally, updating the amplitude timing waveform based on the width timing waveform to obtain the updated amplitude timing waveform includes:

[0122] A1. Convert the amplitude timing waveform into a first function.

[0123] In this embodiment, the first function can be a function obtained by converting an amplitude-time waveform. The first function can indicate the intensity of breathing.

[0124] Specifically, the method for converting the amplitude time-series waveform into the first function can be as follows: First, the acquired amplitude time-series waveform is discretely sampled to obtain a series of amplitude values ​​corresponding to each time point. Then, a suitable function model (such as a sine function) is selected, and optimization algorithms such as the least squares method are used to fit the function, determine the parameters in the function model, and obtain the first function.

[0125] A2. Convert the width timing waveform into a second function.

[0126] In this embodiment, the second function can be a function obtained by transforming a width-time waveform. The second function can indicate the rate or frequency of breathing.

[0127] Specifically, a method for converting a width-time-series waveform into a second function could be, for example, identifying the start and end positions of each respiratory cycle, approximating these discrete data points using a function fitting method, and finally obtaining a continuous second function.

[0128] A3. The ratio of the first function to the second function raised to a predetermined power is determined as the third function.

[0129] In this embodiment, the third function can be the ratio of the first function to a predetermined power of the second function. For example, the third function can be the ratio of the square of the first function to the square of the second function.

[0130] Specifically, the ratio of the first function to the second function raised to a predetermined power is used as the first function.

[0131] A4. Use the waveform corresponding to the third function as the updated amplitude timing waveform.

[0132] Specifically, the obtained third function is used as the updated amplitude timing waveform.

[0133] S360. Perform amplitude analysis on the updated amplitude timing waveform to determine the baseline value and drop value of the updated amplitude timing waveform.

[0134] Specifically, amplitude analysis is performed on the updated amplitude time series waveform. The baseline value is obtained based on the statistics of the peak or amplitude values ​​within a certain range of the peak of the updated amplitude time series waveform. The drop value is obtained based on the statistics of the trough or amplitude values ​​within a certain range of the trough of the updated amplitude time series waveform.

[0135] For example, Figure 13 This is a schematic diagram of the time-corrected amplitude waveform obtained based on respiratory width, provided in an embodiment of the present invention. Figure 13 As shown, Figure 13 The image above shows the respiratory signal waveform, with the horizontal axis representing time and the vertical axis representing respiratory amplitude. Figure 13 The middle section shows the width time series waveform calculated based on a fixed time window. The amplitude time series waveform is then calculated based on the fixed window, and this width time series waveform is used to update the amplitude time series waveform, resulting in the updated amplitude time series waveform (solid line). A comparison is shown between the updated amplitude time series waveform and the original amplitude time series waveform (dashed line). Figure 13 The boxes in the upper respiratory signal waveform diagram represent the fixed time window when calculating the amplitude time-series waveform. The percentage drop of the updated amplitude time-series waveform is 58.97%, which is significantly greater than the original amplitude time-series waveform's 43.49%. Therefore, this breath can be identified as hypoventilation. Figure 13 The following is a comparison chart of the amplitude time series waveform calculated based on a fixed window. The amplitude time series waveform is calculated based on the peak and trough amplitude values ​​of each respiratory cycle and the fixed window. The amplitude time series waveform is updated using this amplitude time series waveform to obtain the updated amplitude time series waveform (solid line). The updated amplitude time series waveform (dashed line) is compared with the original amplitude time series waveform (dashed line). The percentage drop of the updated amplitude time series waveform is 56.80%, which is significantly greater than the 39.49% of the original amplitude time series waveform.

[0136] S370. Based on the baseline value and the drop value, determine whether there is an abnormal breathing event in the original vibration signal.

[0137] The technical solution of this invention involves acquiring an original vibration signal, preprocessing the original vibration signal to obtain a respiratory signal within a required bandwidth range, generating an amplitude-time waveform corresponding to the respiratory signal, generating a width-time waveform corresponding to the respiratory signal, updating the amplitude-time waveform based on the width-time waveform to obtain an updated amplitude-time waveform, and improving the accuracy of hypoventilation event identification by correcting the amplitude-time waveform using the width-time waveform; performing amplitude analysis on the updated amplitude-time waveform to determine the baseline value and drop value of the updated amplitude-time waveform, thus avoiding interference from hypoventilation events and improving the accuracy of sleep abnormality event detection; and determining whether a respiratory abnormality event exists in the original vibration signal based on the baseline value and the drop value.

[0138] In another embodiment, the step of performing amplitude analysis on the amplitude timing waveform to determine the baseline value and drop value of the amplitude timing waveform includes:

[0139] B1. Perform amplitude analysis on the updated amplitude timing waveform to determine a first analysis result, the first analysis result including the baseline value and drop value of the updated amplitude timing waveform.

[0140] In this embodiment, the first analysis result can be an amplitude analysis result obtained based on the updated amplitude timing waveform. The first analysis result includes a baseline value and a drop value, which are obtained from the updated amplitude timing waveform.

[0141] Specifically, amplitude analysis is performed on the updated amplitude time series waveform. The baseline value is obtained based on the statistical value of the peak or a certain range of the peak amplitude of the updated amplitude time series waveform. The drop value is obtained based on the statistical value of the trough or a certain range of the trough amplitude of the updated amplitude time series waveform. The baseline value and the drop value are used as the first analysis results.

[0142] B2. Perform amplitude analysis on the amplitude timing waveform before the update to determine the second analysis result, which includes the baseline value and drop value of the amplitude timing waveform before the update.

[0143] Wherein, the first analysis result is used to determine the first drop percentage of the updated amplitude time series waveform, and the second analysis result is used to determine the second drop percentage of the unupdated amplitude time series waveform;

[0144] In this embodiment, the second analysis result can be an amplitude analysis result obtained based on the amplitude time series waveform before the update. The second analysis result includes a baseline value and a drop value, determined by the amplitude time series waveform before the update. The first drop percentage can be a drop percentage determined by the baseline value and the drop value in the first analysis result. The first drop percentage can describe the degree of drop of the drop value represented by the updated amplitude time series waveform relative to the baseline value. The second drop percentage can be a drop percentage determined based on the baseline value and the drop value in the second analysis result. The second drop percentage can describe the degree of drop of the drop value represented by the amplitude time series waveform before the update relative to the baseline value.

[0145] Specifically, amplitude analysis is performed on the amplitude time series waveform before the update. The baseline value is obtained based on the statistical value of the peak or a certain range of the peak amplitude time series waveform before the update. The drop value is obtained based on the statistical value of the trough or a certain range of the trough amplitude time series waveform before the update. The baseline value and the drop value are used as the second analysis results.

[0146] B3. When the first drop percentage and the second drop percentage are both greater than the corresponding percentage thresholds, it is determined that there is a respiratory abnormality event in the original vibration signal.

[0147] Specifically, the first drop percentage is calculated using the first analysis result, and the second drop percentage is calculated using the second analysis result. If both the first drop percentage and the second drop percentage are greater than their respective percentage thresholds, then a respiratory abnormality event is determined to have occurred.

[0148] For example, Table 2 is a comparison table of the accuracy of respiratory abnormality event judgment before and after the amplitude-time waveform update provided in the embodiments of the present invention. As shown in Table 2, for five nightly monitoring data collected in internal clinical settings where hypoventilation events were dominant, the time correction method was used to identify respiratory events, which improved the sensitivity by nearly 30% compared to the original method.

[0149] Table 2. Comparison of Accuracy of Respiratory Abnormality Event Judgment Before and After Amplitude-Time Waveform Update

[0150]

[0151] The present invention will be described below by way of example, where “filtered respiratory waveform” represents “respiratory signal”, “baseline” represents “baseline value”, “baseline amplitude region” represents “baseline region”, “amplitude” represents “amplitude value”, “drop amplitude region” represents “drop region”, “drop percentage threshold” represents “percentage threshold”, and “amplitude waveform” represents “amplitude timing waveform”:

[0152] Polysomnography (PSG) is the standard for diagnosing sleep apnea / hypopnea and plays a crucial role in guiding effective management strategies to improve sleep quality and overall health in individuals with sleep-disordered breathing. PSG testing is costly and requires specialized facilities and personnel; furthermore, manually scoring sleep apnea and hypopnea events after measurement is very time-consuming. Therefore, obtaining a non-invasive, home-based method and device for detecting sleep apnea and hypopnea is of paramount importance.

[0153] During sleep, the human body typically lies on bedding, including but not limited to beds and sofas, either lying flat or at a slight angle. Generally, a vibration sensor can be placed below the upper body, collecting vibration signals (i.e., raw vibration signals) without direct contact. Vibration sensor signals (i.e., raw vibration signals) can be collected using accelerometers, pressure sensors, displacement sensors, or sensors that convert physical quantities based on acceleration, pressure, and displacement (such as piezoelectric ceramic sensors, inflatable micro-motion sensors, fiber optic sensors, etc.). Generally, to ensure the quality of the collected signals, measurements need to be taken in a relatively quiet environment. During sleep, the human body is in a resting state, resulting in higher signal quality. Since the sensor senses vibration signals, the collected raw data includes components of human respiratory signals and heartbeat signals, as well as interference from environmental micro-vibrations, body movements, and circuit noise. This embodiment uses a piezoelectric ceramic sensor, which is sensitive to changes in acceleration caused by vibration. Figure 1 The diagram shows the waveform of the original vibration signal obtained. The general outline of the signal at this time is the signal envelope generated by human respiration, while the heartbeat and other interference noise are superimposed on the respiration envelope curve.

[0154] The raw vibration signal contains a wealth of information, so preprocessing is necessary to capture the signal within the bandwidth required for calculating relevant parameters. During filtering and denoising, it's also necessary to analyze whether the raw vibration signal carries power frequency interference. If so, a power frequency notch filter may be needed to remove the noise. Of course, the raw vibration signal itself is superimposed with body motion signals. However, since the amplitude or energy of body motion is much greater than that of heartbeats or respiration, the signal acquired by the sensor is prone to exceeding the range. Therefore, defining appropriate upper and lower thresholds (i.e., bandwidth range) is sufficient to identify the body motion state and remove the signal. Then, by reasonably scaling the signal dynamic range, the desired time-domain signal waveform (i.e., respiratory signal) can be obtained. Taking the acquisition of the time-domain respiratory waveform (i.e., respiratory signal) as an example, a frequency range of 0.05Hz–0.5Hz is taken as the required bandwidth range. By designing a reasonable filter, relevant signal data can be obtained. Individual respiratory cycles can be analyzed, i.e., peak-to-peak or trough-to-trough intervals are determined through waveform peak-to-trough search, and the respiratory rate can be calculated accordingly.

[0155] Amplitude-time waveforms can be generated based on the filtered breathing waveform, and can be generated in various ways.

[0156] 1. Set a fixed time window and calculate the maximum and minimum deviations of the respiratory waveform within the time window;

[0157] 2. Set a variable time window, with the window time mapped to the average respiratory cycle width, and calculate the maximum and minimum deviations of the respiratory waveform within the time window;

[0158] 3. Search for peaks and troughs in each respiratory cycle, and perform amplitude mapping by the amplitude of the rising edge or the amplitude of the falling edge, or take the weighted result of the two as the amplitude mapping;

[0159] 4. Search for peaks and troughs in each respiratory cycle and calculate a weighted average using multiple nearby respiratory cycles;

[0160] 5. Through one or more of the above weighted mappings; or rule-based mappings, such as taking the third method when the breathing pattern is stable and good, and taking the first method when it is difficult to search for effective peaks and valleys, such as when there is no body movement or breathing action, etc.

[0161] After generating the amplitude-time waveform, respiratory events can be identified through amplitude analysis. Baseline calculations can be performed on the amplitude region before (or after) a non-respiratory event. However, it is necessary to determine the start and end positions of the baseline region, which can be done in several ways:

[0162] 1. Use the maximum amplitude value as the start and end positions;

[0163] 2. Use the maximum decrease in value by X% as the starting and ending points. X can be 10%, 20%, etc.

[0164] 3. Use the peak or trough of the respiratory cycle corresponding to the maximum amplitude value, or the N% percentile of the rising or falling edge, as the start and end positions. N can be 10%, 20%, etc.

[0165] 4. The starting and ending positions are the peak or trough of the respiratory cycle corresponding to the maximum amplitude decrease of X%, or the N% percentile of the rising or falling edge. X can be 10%, 20%, etc., and N can be 10%, 20%, etc.

[0166] 5. Through one or more of the above weighted mappings; or rule-based mappings.

[0167] Once the start and end points of the baseline region are determined, baseline calculation can be performed, and it can be generated in several ways:

[0168] 1. Use the maximum amplitude value within the baseline amplitude range as the baseline value;

[0169] 2. Use the minimum amplitude value within the baseline amplitude range as the baseline value;

[0170] 3. Use the average amplitude value within the baseline amplitude area as the baseline value;

[0171] 4. Use the median amplitude value within the baseline amplitude range as the baseline value;

[0172] 5. Use the amplitude value that accounts for the largest proportion within the baseline amplitude range as the baseline value;

[0173] 6. Through one or more of the above weighted mappings; or rule-based mappings.

[0174] When a respiratory event occurs, the waveform amplitude (i.e., the amplitude value of the amplitude-time waveform) drops. Similarly, the baseline method can be used to determine the start and end points of the drop area:

[0175] 1. Use the minimum amplitude value as the start and end positions;

[0176] 2. Use the minimum increase of X% as the starting and ending points, where X can be 10%, 20%, etc.

[0177] 3. Use the peak or trough of the respiratory cycle corresponding to the minimum amplitude value, or the N% percentile of the rising or falling edge, as the start and end positions. N can be 10%, 20%, etc.

[0178] 4. The starting and ending positions are the respiratory cycle peak or trough, or the N% percentile of the rising or falling edge, corresponding to the minimum amplitude value increase of X%. X can be 10%, 20%, etc., and N can be 10%, 20%, etc.

[0179] 5. Through one or more of the above weighted mappings; or rule-based mappings.

[0180] Once the start and end points of the fall zone are determined, the fall value can be calculated, which can be generated in several ways:

[0181] 1. The maximum amplitude value within the drop range is taken as the drop value;

[0182] 2. The minimum drop value within the drop range is taken as the drop value;

[0183] 3. The average drop value within the drop range is used as the drop value;

[0184] 4. Use the median value within the drop range as the drop value;

[0185] 5. The drop value is determined by the percentage of the drop range that is most significant within that range.

[0186] 6. Through one or more of the above weighted mappings; or rule-based mappings.

[0187] This allows us to determine the baseline region and the drop region. In this embodiment, the maximum amplitude value in the baseline region is used as the baseline value, and the minimum amplitude value in the drop region is used as the drop value for event analysis.

[0188] Calculate the percentage drop = (1 - drop value / baseline value) By statistically analyzing the percentage of falls in a large number of sleep apnea events and examining the sensitivity and specificity curves, a suitable percentage fall threshold can be determined to identify sleep apnea events. This threshold can also be dynamically adjusted when identifying an event.

[0189] 1. A fixed percentage drop threshold is used as the judgment threshold;

[0190] 2. Use the percentage drop threshold of the most recent identified event as the judgment threshold, or use it as a weighted average of a fixed threshold;

[0191] 3. Use the maximum, minimum, average, or median of the percentage drop thresholds of the most recent identified events as the decision threshold, or use a weighted average of these thresholds with a fixed threshold as the decision threshold;

[0192] 4. Through one or more of the above weighted mappings; or rule-based mappings, such as mappings that are initially fixed and dynamically adjusted over time based on identified events.

[0193] Using, for example, a 70% threshold as a criterion, the threshold will decrease during hypoventilation. Similar calculation methods can be used, which will not be elaborated upon here. Of course, additional calculations can be made to adjust the percentage drop for the same event based on baseline values ​​before and after the event to avoid misclassification of certain non-respiratory events. For example, if there is a slight change in body position without causing body movement, the percentage drop before the event might exceed the threshold, but the percentage drop compared to the baseline value after the event might be very small. That is, the body has remained in a stable baseline breathing state after the positional adjustment. In this case, a supplementary model based on the post-event baseline can be used to eliminate the misidentification of the event.

[0194] However, when hypoventilation occurs, insufficient amplitude reduction may lead to many missed hypoventilation events and an underestimation of the number of events when using a large threshold such as 50%. Conversely, setting the threshold very low, such as 30%, may result in many false positives and an overestimation of the number of events. Relying solely on respiratory waveforms acquired by vibration sensors, without blood oxygen saturation data as an auxiliary indicator, makes it difficult to determine whether hypoventilation constitutes a respiratory event. Therefore, it is necessary to improve the accuracy of respiratory event identification in this gray area.

[0195] Generally, during a sleep apnea event, the body's normal breathing is disrupted by insufficient oxygen supply due to hypoventilation, leading to deeper and faster breathing towards the end of the event. This deeper breathing is reflected in the recovery of respiratory amplitude, as described earlier in this invention. The faster breathing is reflected in the recovery of the respiratory cycle width. For example, at the beginning of a hypoventilation event, the time width is approximately 1660 ms, which initially widens to approximately 1950 ms during the event, before gradually narrowing back to around 1700 ms. Therefore, the time-corrected amplitude waveform can be calculated by combining the amplitude and width ratio. The width waveform can be calculated based on a fixed window (the width can be the maximum, minimum, average, or median value within the window, etc.), and the time-corrected amplitude waveform (shown by the solid line) can be obtained by the ratio of the amplitude waveform to the width waveform. Note that the numerator of this ratio is a function transformation of the amplitude waveform, and the denominator is a function transformation of the width waveform. If the amplitude waveform is defined as... The width waveform is The corrected waveform is The simplest is , To increase time-width sensitivity, it can be... , The mapping function can be selected based on the signal characteristics acquired by the actual sensor.

[0196] Time correction is also applicable to the processing of sleep apnea events mentioned earlier. However, since the amplitude waveform itself has sufficient sensitivity and specificity (the drop percentage is relatively large, and the difference from the normal breathing waveform is obvious), there is no need to add additional time correction, thus reducing computational load and saving computing resources and power consumption. When the drop percentage is in an ambiguous range, the time correction method can be introduced for processing. Then, the judgment is made based on the new corrected drop percentage threshold (i.e., the percentage threshold of the updated amplitude time-series waveform), or by using the original drop percentage threshold (i.e., the percentage threshold of the amplitude time-series waveform before the update) + the corrected drop percentage threshold as a dual constraint, thereby improving the accuracy of event identification. Note that the corrected drop percentage threshold here can also be determined by statistically analyzing the corrected drop percentage of a large number of breathing events (or only for hypoventilation breathing events), analyzing the sensitivity and specificity curves, and determining a suitable corrected drop percentage threshold to judge breathing events. When judging breathing events, this corrected threshold can also be dynamically adjusted with reference to the original threshold mentioned earlier, which will not be elaborated further.

[0197] The respiratory event identification method and the time-correction-based method proposed in this invention only require processing in the time domain, are easy to understand and implement, consume little computational resources, and improve the accuracy of respiratory event identification, especially hypoventilation events.

[0198] The detection results of respiratory abnormalities in this invention can be used for reporting and statistics, real-time user intervention, or evaluation of intervention effects. Depending on the usage scenario, the data can ultimately be displayed or commands can be interacted with on terminals such as mobile phones, tablets, and computers. Alternatively, data can be displayed directly through a configured display screen or connected to other terminals via a wired connection.

[0199] Example 4

[0200] Figure 14 This is a schematic diagram of a respiratory abnormality event detection device provided in Embodiment 4 of the present invention. Figure 14 As shown, the device includes:

[0201] The acquisition module 410 is used to acquire the original vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor.

[0202] The preprocessing module 420 is used to preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range;

[0203] The generation module 430 is used to generate the amplitude-time waveform corresponding to the respiratory signal;

[0204] The first determining module 440 is used to perform amplitude analysis on the amplitude timing waveform and determine the baseline value and drop value of the amplitude timing waveform;

[0205] The second determining module 450 is used to determine whether there is an abnormal breathing event in the original vibration signal based on the baseline value and the drop value.

[0206] The technical solution of this invention involves acquiring the original vibration signal through an acquisition module, and then collecting the vibration signal generated by the user imperceptibly through a vibration sensor mounted in the mattress. A preprocessing module preprocesses the original vibration signal to obtain a breathing signal within the required bandwidth range, eliminating noise interference in the original vibration signal. A generation module generates an amplitude-time-series waveform corresponding to the breathing signal, representing the user's breathing status in time sequence and providing a basis for detecting abnormal breathing events. Processing is performed only in the time domain, reducing computational resource overhead. A first determination module performs amplitude analysis on the amplitude-time-series waveform to determine its baseline and drop values. A second determination module, based on the baseline and drop values, determines whether an abnormal breathing event exists in the original vibration signal. This allows for the acquisition and analysis of vibration signals during the user's sleep based on the vibration sensor, enabling imperceptible, home-based detection of abnormal breathing events during sleep, while improving the accuracy of abnormal breathing event detection.

[0207] In another embodiment, the first determining module 440 is specifically used for:

[0208] The maximum amplitude range in the amplitude time-series waveform is analyzed to determine the baseline region in the amplitude time-series waveform;

[0209] Based on the amplitude values ​​within the baseline region, the baseline value of the amplitude timing waveform is determined;

[0210] The minimum amplitude range in the amplitude time-series waveform is analyzed to determine the drop region in the amplitude time-series waveform;

[0211] The drop value of the amplitude timing waveform is determined based on the amplitude value within the drop region.

[0212] In another embodiment, the second determining module 450 is specifically used for:

[0213] The percentage drop is determined based on the baseline value and the drop value;

[0214] Based on the comparison between the drop percentage and the percentage threshold, it is determined whether there is an abnormal breathing event in the original vibration signal.

[0215] In another embodiment, the device further includes:

[0216] A width-time waveform generation module is used to generate the width-time waveform corresponding to the respiratory signal;

[0217] The update module is used to update the amplitude timing waveform based on the width timing waveform to obtain the updated amplitude timing waveform.

[0218] In another embodiment, the update module is specifically used for:

[0219] Convert the amplitude timing waveform into a first function;

[0220] The width timing waveform is converted into a second function;

[0221] The ratio of the first function to the second function raised to a predetermined power is used as the third function;

[0222] The waveform corresponding to the third function is used as the updated amplitude timing waveform.

[0223] In another embodiment, the second determining module 450 is specifically used for:

[0224] Amplitude analysis is performed on the updated amplitude timing waveform to determine the baseline value and drop value of the updated amplitude timing waveform.

[0225] In another embodiment, the first determining module 440 is specifically used for:

[0226] Amplitude analysis is performed on the updated amplitude timing waveform to determine a first analysis result, which includes the baseline value and drop value of the updated amplitude timing waveform.

[0227] Amplitude analysis is performed on the amplitude timing waveform before the update to determine a second analysis result, which includes the baseline value and drop value of the amplitude timing waveform before the update.

[0228] Wherein, the first analysis result is used to determine the first drop percentage of the updated amplitude time series waveform, and the second analysis result is used to determine the second drop percentage of the unupdated amplitude time series waveform;

[0229] When the first drop percentage and the second drop percentage are both greater than their respective percentage thresholds, it is determined that there is a respiratory abnormality event in the original vibration signal.

[0230] The respiratory abnormality event detection device provided in this embodiment of the invention can execute the method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0231] Example 5

[0232] Figure 15 This is a structural block diagram of an electronic device provided in an embodiment of the present invention, such as... Figure 15The diagram illustrates a schematic representation of an electronic device 10 that can be used to implement embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0233] like Figure 15 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor 11, and the computer program is executed by the at least one processor 11 to enable the at least one processor 11 to perform the method provided by the present invention.

[0234] The processor 11 can perform various appropriate actions and processes based on a computer program stored in the read-only memory (ROM) 12 or a computer program loaded from the storage unit 18 into the random access memory (RAM) 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0235] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0236] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the methods provided in this invention.

[0237] In some embodiments, the methods provided herein may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to execute the methods by any other suitable means (e.g., by means of firmware).

[0238] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard parts (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0239] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0240] In the context of this invention, a computer-readable storage medium stores computer instructions that are used to cause a processor to execute and implement the method provided by this invention.

[0241] The present invention also provides a computer program product comprising a computer program that, when executed by a processor, implements the method provided according to embodiments of the present invention. A computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0242] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube or liquid crystal display) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0243] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0244] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0245] This invention also provides a computer program product, including a computer program that, when executed by a processor, can implement the methods provided in any embodiment of this application.

[0246] In the implementation of the computer program product, computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0247] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0248] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for detecting abnormal respiratory events, characterized in that, include: Acquire raw vibration signals, which include vibration signals generated by the user collected by vibration sensors; The original vibration signal is preprocessed to obtain the breathing signal within the required bandwidth range; Generate the amplitude-time waveform corresponding to the respiratory signal; Amplitude analysis is performed on the amplitude timing waveform to determine the baseline value and drop value of the amplitude timing waveform; Based on the baseline value and the drop value, determine whether there is an abnormal breathing event in the original vibration signal.

2. The method according to claim 1, characterized in that, The step of performing amplitude analysis on the amplitude time-series waveform to determine the baseline value and drop value of the amplitude time-series waveform includes: The maximum amplitude range in the amplitude time-series waveform is analyzed to determine the baseline region in the amplitude time-series waveform; Based on the amplitude values ​​within the baseline region, the baseline value of the amplitude timing waveform is determined; The minimum amplitude range in the amplitude time-series waveform is analyzed to determine the drop region in the amplitude time-series waveform; The drop value of the amplitude timing waveform is determined based on the amplitude value within the drop region.

3. The method according to claim 1, characterized in that, The step of determining whether an abnormal respiratory event exists in the original vibration signal based on the baseline value and the drop value includes: The percentage drop is determined based on the baseline value and the drop value; Based on the comparison between the drop percentage and the percentage threshold, it is determined whether there is an abnormal breathing event in the original vibration signal.

4. The method according to any one of claims 1-3, characterized in that, After generating the amplitude-time waveform corresponding to the respiratory signal, the method further includes: Generate the width-time waveform corresponding to the respiratory signal; The amplitude timing waveform is updated based on the width timing waveform to obtain the updated amplitude timing waveform.

5. The method according to claim 4, characterized in that, The step of updating the amplitude time-series waveform based on the width time-series waveform to obtain the updated amplitude time-series waveform includes: Convert the amplitude timing waveform into a first function; The width timing waveform is converted into a second function; The ratio of the first function to the second function raised to a predetermined power is used as the third function; The waveform corresponding to the third function is used as the updated amplitude timing waveform.

6. The method according to claim 4, characterized in that, The step of performing amplitude analysis on the amplitude time-series waveform to determine the baseline value and drop value of the amplitude time-series waveform includes: Amplitude analysis is performed on the updated amplitude timing waveform to determine the baseline value and drop value of the updated amplitude timing waveform.

7. The method according to claim 4, characterized in that, The step of performing amplitude analysis on the amplitude time-series waveform to determine the baseline value and drop value of the amplitude time-series waveform includes: Amplitude analysis is performed on the updated amplitude timing waveform to determine a first analysis result, which includes the baseline value and drop value of the updated amplitude timing waveform. Amplitude analysis is performed on the amplitude timing waveform before the update to determine a second analysis result, which includes the baseline value and drop value of the amplitude timing waveform before the update. Wherein, the first analysis result is used to determine the first drop percentage of the updated amplitude time series waveform, and the second analysis result is used to determine the second drop percentage of the unupdated amplitude time series waveform; When the first drop percentage and the second drop percentage are both greater than their respective percentage thresholds, it is determined that there is a respiratory abnormality event in the original vibration signal.

8. A respiratory abnormality event detection device, characterized in that, include: The acquisition module is used to acquire the raw vibration signal, which includes the vibration signal generated by the user collected by the vibration sensor; The preprocessing module is used to preprocess the original vibration signal to obtain the breathing signal within the required bandwidth range; The generation module is used to generate the amplitude-time waveform corresponding to the respiratory signal; The first determining module is used to perform amplitude analysis on the amplitude time-series waveform to determine the baseline value and drop value of the amplitude time-series waveform; The second determining module is used to determine whether there is an abnormal breathing event in the original vibration signal based on the baseline value and the drop value.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method of any one of claims 1-7.

Images