A method and apparatus for identifying SAS patients based on PPG signals

Abstract

This application discloses a method and apparatus for determining SAS (Self-Rating Anxiety) patients based on PPG signals, relating to the field of biomedical technology. The method includes: acquiring a target continuous PPG signal and blood oxygen saturation data of the patient within a set time period; determining the trough position of the target continuous PPG signal based on a window sliding minimum detection method, and segmenting the target continuous PPG signal into at least two individual PPG signals; determining at least one risk period corresponding to the patient from the set time period based on the blood oxygen saturation data, and the blood oxygen change amplitude corresponding to each risk period; performing multi-peak fitting processing on each individual PPG signal using a Gaussian function to obtain the Gaussian mean change amplitude corresponding to each risk period; and determining the patient's SAS result based on the blood oxygen change amplitude and Gaussian mean change amplitude corresponding to each risk period, along with the judgment criteria. In this application, the Gaussian mean is used instead of heart rate, which effectively ignores the interference of high-frequency signals, making the SAS detection results more reliable.

Description

Technical Field

[0001] This application relates to the field of biomedical technology, and in particular to a method and apparatus for identifying SAS patients based on PPG signals. Background Technology

[0002] SAS usually refers to Sleep Apnea Syndrome (SAS), a common sleep disorder characterized by repeated pauses in breathing or irregular breathing during sleep, leading to insufficient oxygen at night and poor sleep quality.

[0003] In existing technologies, determining whether a patient has SAS typically involves using a photoplethysmography (PPG) sensor to collect the patient's heart rate and blood oxygen saturation. Wavelet transform is then used to process the PPG signal, extracting blood oxygen saturation (SpO2), heart rate (HR), and their amplitude changes (ΔSpO2, ΔHR). The relationship between the ratio of ΔSpO2 to ΔHR and screening criteria is then analyzed to determine if the patient has SAS. However, research has found that the accuracy of this method depends entirely on the stability of HR, which is affected by environmental noise and motion artifacts, thus impacting SAS screening results and increasing the risk of misdiagnosis. Summary of the Invention

[0004] The purpose of this application is to address at least one of the aforementioned technical deficiencies.

[0005] On one hand, embodiments of this application provide a method for identifying SAS patients based on PPG signals, the method comprising:

[0006] Acquire the patient's target continuous PPG signal and blood oxygen saturation data within a set time period. The target continuous PPG signal is the continuous PPG signal corresponding to the pulse data.

[0007] The trough position of the target continuous PPG signal is determined based on the window sliding minimum detection method, and the target continuous PPG signal is cut into at least two individual PPG signals according to the trough position.

[0008] Based on blood oxygen saturation data, determine at least one risk duration for the patient from the set duration, as well as the blood oxygen change range corresponding to each risk duration;

[0009] A Gaussian function was used to perform multi-peak fitting on each individual PPG signal to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration.

[0010] The patient's SAS result is determined based on the blood oxygen saturation variation for each risk duration, the mean Gaussian wave variation for each risk duration, and the judgment criteria.

[0011] Optionally, acquiring the patient's target continuous PPG signal over a set time period includes:

[0012] Acquire the initial continuous PPG signal corresponding to the patient's pulse data within a set time period;

[0013] Obtain the cutoff frequency and high-pass filter, and determine the filter coefficients based on the cutoff frequency and high-pass filter.

[0014] The baseline drift interference is removed from the initial continuous PPG signal based on the filter coefficients to obtain the first continuous PPG signal.

[0015] Obtain the notch filter frequency and the notch filter itself, and determine the notch filter coefficients based on the notch filter frequency and the notch filter itself;

[0016] The first continuous PPG signal is processed to remove specific frequency interference based on the notch filter coefficient, and the target continuous PPG signal is obtained.

[0017] Optionally, the location of the trough in the target continuous PPG signal is determined based on the window sliding minimum detection method, including:

[0018] Obtain the initial window. The number of sampling points contained in the initial window is determined based on the sampling rate of the target continuous PPG signal and the heart rate reference value.

[0019] Based on the initial window, a set distance is slid across the target continuous PPG signal each time to obtain at least one first window, and at least one second window is determined from the at least one first window according to the data point values ​​in each first window;

[0020] Based on the values ​​of the data points in each second window, determine the minimum data point corresponding to each second window, and determine the rising and falling trend between each minimum data point and its adjacent data points;

[0021] The position of the smallest data point whose rise and fall trend meets the preset requirements is taken as the trough position of the target continuous PPG signal.

[0022] Optionally, if the number of extremely small data points that meet the preset requirements for the rising and falling trend reaches a set number, and the sliding distance does not reach the endpoint of the target continuous PPG signal, the method further includes:

[0023] The initial window length and the set distance are adjusted to obtain the adjusted window and the adjusted distance. Based on the adjusted window and the adjusted distance, the trough position of the target continuous PPG signal is further determined.

[0024] Optionally, at least one risk duration corresponding to the patient can be determined from a set duration based on blood oxygen saturation data, including:

[0025] Determine at least one duration during which the blood oxygen saturation value is lower than a preset value in the blood oxygen saturation data corresponding to a set duration;

[0026] The duration during which the blood oxygen level is below a preset value is considered as at least one risk duration for the patient.

[0027] Optionally, a Gaussian function is used to perform multi-peak fitting on each individual PPG signal to obtain the mean Gaussian wave variation amplitude corresponding to each risk duration, including:

[0028] Based on the time information of each individual PPG signal, determine at least two target individual PPG signals corresponding to each risk duration from at least two individual PPG signals;

[0029] For each risk duration, a Gaussian function is used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration, so as to obtain the mean change amplitude of the Gaussian wave for each risk duration.

[0030] Optionally, a Gaussian function is used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration, including:

[0031] A Gaussian function was used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration to obtain the Gaussian mean of the single PPG signal of each target.

[0032] Based on the Gaussian mean of the individual PPG signal for each target, the amplitude of the change in the Gaussian mean for each risk duration is obtained.

[0033] Optionally, a Gaussian function is used to perform multi-peak fitting on the individual PPG signal of each target corresponding to the risk duration to obtain the Gaussian mean of the individual PPG signal of each target, including:

[0034] For each target single PPG signal, the Gaussian function and the initial parameter values ​​of the function are obtained, and the target single PPG signal is processed based on the Gaussian function to obtain the function value corresponding to each data point in the target single PPG signal;

[0035] Based on the values ​​of each data point in the target single PPG signal and the corresponding function values ​​of each data point in the target single PPG signal, a corresponding target function is constructed;

[0036] Substitute the initial parameter values ​​of the function into the objective function and perform iterative calculations until the convergence condition is met to obtain the Gaussian mean of the target single PPG signal.

[0037] Optionally, the judgment criteria include the product standard value and the low ventilation rate standard. Based on the blood oxygen saturation variation for each risk duration, the mean Gaussian wave variation for each risk duration, and the judgment criteria, the patient's SAS result is determined, including:

[0038] For each risk duration, the product of the blood oxygen change amplitude corresponding to the risk duration and the corresponding Gaussian mean change amplitude is processed to obtain the first value corresponding to the risk duration.

[0039] The first value corresponding to each risk duration is compared with the product standard to determine the number of target risk durations from all risk durations.

[0040] The number of target risk durations is compared with the criteria for low ventilation frequency to determine the patient's SAS outcome.

[0041] On the other hand, embodiments of this application provide a device for identifying SAS patients based on PPG signals, comprising:

[0042] The data acquisition module is used to acquire the patient's target continuous PPG signal and blood oxygen saturation data within a set time period. The target continuous PPG signal is the continuous PPG signal corresponding to the pulse data.

[0043] The signal cutting module is used to determine the trough position of the target continuous PPG signal based on the window sliding minimum detection method, and cut the target continuous PPG signal into at least two individual PPG signals according to the trough position.

[0044] The data processing module is used to determine at least one risk duration for the patient from a set duration based on blood oxygen saturation data, as well as the blood oxygen change range corresponding to each risk duration.

[0045] The data fitting module is used to perform multi-peak fitting on each individual PPG signal using a Gaussian function to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration.

[0046] The results determination module is used to determine the patient's SAS results based on the blood oxygenation variation amplitude corresponding to each risk duration, the Gaussian mean variation amplitude corresponding to each risk duration, and the judgment criteria.

[0047] In another aspect, embodiments of this application provide an electronic device, including a processor and a memory:

[0048] The memory is configured to store machine-readable instructions that, when executed by the processor, cause the processor to perform any of the methods in the PPG signal-based SAS patient determination method.

[0049] The beneficial effects of the technical solutions provided in this application include at least the following:

[0050] In this embodiment, the continuous PPG signal is segmented into PPG signals of one cardiac cycle length (i.e., a single PPG signal) based on the window minimum detection method. Then, the single PPG signal is decomposed into a Gaussian mean value through Gaussian multi-peak fitting processing, and the amplitude of the Gaussian mean value change is determined. Finally, the patient's SAS is determined based on the amplitude of the Gaussian mean value change and the amplitude of the blood oxygen change. That is, in this application, the Gaussian mean value (μ1) is used instead of heart rate (HR), which can effectively ignore the interference of high-frequency signals, making the SAS detection results more reliable. It is especially suitable for environments with large heart rate changes or signal interference, providing an efficient and accurate heart rate variability detection technology for medical health monitoring.

[0051] In this embodiment, the acquired initial continuous PPG signal is processed using a high-pass filter and a notch filter. This process removes low-frequency baseline drift and specific frequencies caused by physiological changes, motion artifacts, or other non-cardiovascular signals from the initial continuous PPG signal, effectively ignoring the influence of some high-frequency signals and making the detection results more reliable. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application, 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 A flowchart illustrating a method for identifying SAS patients based on PPG signals, provided for an embodiment of this application;

[0054] Figure 2 A flowchart illustrating the process of determining the location of a trough, provided as an embodiment of this application;

[0055] Figure 3 A waveform diagram of a single PPG signal provided in an embodiment of this application;

[0056] Figure 4 A schematic flowchart illustrating another method for identifying SAS patients based on PPG signals, provided for an embodiment of this application;

[0057] Figure 5 A schematic diagram of a device for identifying SAS patients based on PPG signals is provided in this application embodiment;

[0058] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0059] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting the invention.

[0060] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this application means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0062] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0063] Specifically, such as Figure 1 As shown, the method may include:

[0064] Step S101: Acquire the patient's target continuous PPG signal and blood oxygen saturation data within a set time period. The target continuous PPG signal is the continuous PPG signal corresponding to the pulse data.

[0065] Optionally, the duration can be set according to the patient's actual situation, and this application embodiment does not limit it. For example, the duration can be set to 1 hour, in which case the continuous PPG signal and blood oxygen saturation data corresponding to the patient's pulse data within that 1 hour can be obtained based on the PPG sensor.

[0066] In an optional embodiment of this application, acquiring the patient's target continuous PPG signal within a set time period includes:

[0067] Acquire the initial continuous PPG signal corresponding to the patient's pulse data within a set time period;

[0068] Obtain the cutoff frequency and high-pass filter, and determine the filter coefficients based on the cutoff frequency and high-pass filter.

[0069] The baseline drift interference is removed from the initial continuous PPG signal based on the filter coefficients to obtain the first continuous PPG signal.

[0070] Obtain the notch filter frequency and the notch filter itself, and determine the notch filter coefficients based on the notch filter frequency and the notch filter itself;

[0071] The first continuous PPG signal is processed to remove specific frequency interference based on the notch filter coefficient, and the target continuous PPG signal is obtained.

[0072] Optionally, the initial continuous PPG signal obtained by the PGG sensor may be subject to low-frequency baseline drift and specific frequency interference caused by physiological changes, motion artifacts or other non-cardiovascular signals. Since these drifts and interferences may affect the final judgment results, this application will perform baseline drift interference removal processing and specific frequency interference removal processing on the acquired initial continuous PPG signal.

[0073] Specifically, the cutoff frequency ω can be obtained. C While high-pass filters are commonly used for PPG signals, the baseline drift frequency is relatively low in practical applications, typically between 0.01Hz and 0.5Hz. Because of its flat response within the bandpass, the Butterworth filter is frequently employed for PPG signals. In this case, a Butterworth high-pass filter can be obtained, as shown below:

[0074]

[0075] Where s is the complex frequency, ω C Q is the cutoff frequency, and Q is the quality factor.

[0076] Furthermore, the filter coefficients can be determined based on the obtained cutoff frequency and Butterworth high-pass filter. Specifically, the filter coefficients can be calculated using the filter design function in Matlab (a programming tool). The calculated filter coefficients are then applied to the initial continuous PPG signal to obtain the first continuous PPG signal, which has had low-frequency components removed.

[0077] Correspondingly, for specific frequency interference in PPG signals, such as 50Hz power frequency interference in power systems, the notch filter frequency and a notch filter with the notch filter frequency set to 50Hz can be obtained:

[0078]

[0079] Where ω0 is the notch frequency and ε is the damping ratio.

[0080] Furthermore, the notch filter coefficients are determined based on the obtained notch filter frequency and the notch filter itself. Specifically, the notch filter coefficients can be calculated using the notch filter design function in Matlab. Then, the calculated notch filter coefficients are applied to the first continuous PPG signal to obtain the target continuous PPG signal. The obtained target continuous PPG signal is used to remove interference at a specific frequency.

[0081] It is understandable that in practical applications, for the acquired initial continuous PPG signal, there is no specific order between the high-pass filter and the notch filter used to remove low-frequency baseline drift and interference at specific frequencies. The low-frequency baseline drift can be removed first by high-pass filtering, and then the interference at specific frequencies can be removed by notch filter. Alternatively, the interference at specific frequencies can be removed first by notch filter, and then the low-frequency baseline drift can be removed by high-pass filtering. The embodiments of this application do not limit the order of the high-pass filter and notch filter used.

[0082] Step S102: Determine the trough position of the target continuous PPG signal based on the window sliding minimum detection method, and cut the target continuous PPG signal into at least two individual PPG signals according to the trough position.

[0083] Optionally, after obtaining the target continuous PPG signal from the patient within a set time period, the trough position of the target continuous PPG signal can be determined based on the window sliding minimum detection method. Then, based on the determined trough position, the target continuous PPG signal can be cut into multiple individual PPG signals. The trough position is the end point of the previous individual PPG signal and also the starting point of the next individual PPG signal. The target continuous PPG signal is like a rope; at this point, the trough position can be marked with a ruler at the whole centimeter position, and then the target continuous PPG signal can be cut with scissors at the marked point to obtain multiple individual PPG signals.

[0084] In an optional embodiment of this application, determining the trough position of the target continuous PPG signal based on the window sliding minimum detection method includes:

[0085] Obtain the initial window. The number of sampling points contained in the initial window is determined based on the sampling rate of the target continuous PPG signal and the heart rate reference value.

[0086] Based on the initial window, a set distance is slid across the target continuous PPG signal each time to obtain at least one first window, and at least one second window is determined from the at least one first window according to the data point values ​​in each first window;

[0087] Based on the values ​​of the data points in each second window, determine the minimum data point corresponding to each second window, and determine the rising and falling trend between each minimum data point and its adjacent data points;

[0088] The position of the smallest data point whose rise and fall trend meets the preset requirements is taken as the trough position of the target continuous PPG signal.

[0089] Optionally, the procedure for determining the trough position of the target continuous PPG signal based on the window sliding minimum method can be as follows: Figure 2 As shown, this can specifically include acquiring the target continuous PPG signal (i.e. Figure 2 The process begins by analyzing the preprocessed PPG signal and the initial window size. Within the initial window, local minima are searched and valleys are confirmed. Simultaneously, it is determined whether the endpoint of the target continuous PPG signal has been reached (i.e., the end of data traversal in the diagram?). If the endpoint of the target continuous PPG signal has been reached (i.e., the Y branch in the diagram), all valleys can be output. If the endpoint of the target continuous PPG signal has not been reached (i.e., the N branch in the diagram), the sliding window's sliding distance and window length are adjusted based on the distance to adjacent valleys (i.e., the sliding window and adjusted window length in the diagram). New valleys are then searched within the updated window based on the method of finding local minima, until the endpoint is reached and all valleys are output.

[0090] The choice of initial window size is crucial for the accuracy of trough detection. A window that is too small may fail to capture the complete waveform, while a window that is too large may merge multiple waveforms together. Therefore, the number of sampling points included in the initial window can be determined based on the sampling rate of the target continuous PPG signal and a heart rate reference value. For example, a normal human heart rate is between 60-120 beats per minute. When the PPG signal sampling rate is 500, a heart rate of 90 beats per minute can be initially selected, and the number of sampling points included in the initial window length can be set to 334.

[0091] Furthermore, based on the initial window sliding a set distance on the target continuous PPG signal each time, such as sliding a window for 0.67s each time, at least one first window is obtained. Then, based on the data point values ​​within each first window, at least one second window is determined from the at least one first window. The determination of the second window can be achieved by comparing the value of each data point within each first window with the values ​​of its neighboring data points. If the value of a data point in a first window is less than the values ​​of its immediate and adjacent data points, then that first window is the second window.

[0092] Accordingly, based on the value of the data point in each second window, the minimum data point corresponding to each second window (i.e., the value of the data point is less than the value of its neighboring data points) is determined, and the rising and falling trend between each minimum data point and its neighboring data points is determined. If the rising and falling trend of the minimum data point meets the preset requirements, then the position of the minimum data point is the trough position of the target continuous PPG signal.

[0093] In an optional embodiment of this application, if the number of extremely small data points that meet the preset requirements for the upward and downward trend reaches a set number, and the sliding distance does not reach the endpoint of the target continuous PPG signal, the method further includes:

[0094] The initial window length and the set distance are adjusted to obtain the adjusted window and the adjusted distance. Based on the adjusted window and the adjusted distance, the trough position of the target continuous PPG signal is further determined.

[0095] The specific value of the set quantity can be set according to actual needs, and this application does not limit it. Optionally, when the minimum data points that meet the preset requirements for the upward and downward trend reach the set quantity (i.e., the determined trough positions reach the set quantity), and the sliding distance has not reached the end point of the target continuous PPG signal, the initial window length and the size of each sliding distance (i.e., the set distance) can be adjusted. Then, based on the adjusted window and the adjusted set distance, the trough positions of the target continuous PPG signal are determined again using the method described above, until the end point of the target continuous PPG signal is reached. For example, when more than three trough positions are detected, the set distance is adjusted to the average length of the distances of the previous three troughs, and the window length is adjusted to 60% of the window movement distance. Furthermore, after the window slides to the end point of the target continuous PPG signal, the detected trough positions are recorded for subsequent analysis.

[0096] Step S103: Determine at least one risk duration corresponding to the patient from the set duration based on the blood oxygen saturation data, and the blood oxygen change range corresponding to each risk duration.

[0097] The risk duration can refer to the duration during which the patient may have SAS symptoms. Optionally, after obtaining the patient's blood oxygen saturation data within a set time period, the risk duration within the set time period can be determined based on the specific blood oxygen saturation values. Then, the blood oxygen variation range corresponding to each risk duration can be determined based on the specific blood oxygen saturation values ​​within each risk duration. This blood oxygen variation range is the difference between the highest and lowest blood oxygen saturation values ​​within each risk duration, which can be represented by ΔSpO2.

[0098] In an optional embodiment of this application, determining at least one risk duration corresponding to the patient from a set duration based on blood oxygen saturation data includes:

[0099] Determine at least one duration during which the blood oxygen saturation value is lower than a preset value in the blood oxygen saturation data corresponding to a set duration;

[0100] The duration during which the blood oxygen level is below a preset value is considered as at least one risk duration for the patient.

[0101] Specifically, the risk duration refers to the time during which the blood oxygen saturation value drops by more than 4% from a stable state. For example, if the standard blood oxygen saturation is 90% or higher, the duration during which the blood oxygen saturation data falls below 86% is considered a risk duration. Optionally, for the blood oxygen saturation data corresponding to the set duration, at least one duration during which the blood oxygen value is below a preset value can be determined based on the specific blood oxygen saturation value and its corresponding time information. This at least one duration during which the blood oxygen value is below the preset value is then considered as at least one risk duration for the patient. For example, the duration during which the blood oxygen saturation data falls below 86% can be considered a risk duration.

[0102] Step S104: Use a Gaussian function to perform multi-peak fitting on each individual PPG signal to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration.

[0103] Optionally, the waveform of each individual PPG signal can be as follows: Figure 3 As shown, each individual PPG signal is a complete pulse wave waveform of a cardiac cycle, which may include the main wave ( Figure 3 A wave in the middle), tidal wave ( Figure 3 B wave and diphthous wave (in the middle) Figure 3 (C wave in the data). Furthermore, a Gaussian function can be used to perform multi-peak fitting on each individual PPG signal to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration.

[0104] In an optional embodiment of this application, a Gaussian function is used to perform multi-peak fitting processing on each individual PPG signal to obtain the mean Gaussian wave variation amplitude corresponding to each risk duration, including:

[0105] Based on the time information of each individual PPG signal, determine at least two target individual PPG signals corresponding to each risk duration from at least two individual PPG signals;

[0106] For each risk duration, a Gaussian function is used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration, so as to obtain the mean change amplitude of the Gaussian wave for each risk duration.

[0107] Optionally, for each risk duration and each individual PPG signal, there will be corresponding time information. In this case, for a risk duration, based on the risk duration and the time information corresponding to each individual PPG signal, the individual PPG signal whose time information matches the time information of the risk duration can be determined, and the corresponding individual PPG signal can be used as the target individual PPG signal for that risk duration. Furthermore, a Gaussian function can be used to perform multi-peak fitting processing on each target individual PPG signal corresponding to the risk duration to obtain the mean change amplitude of the Gaussian wave for that risk duration.

[0108] In an optional embodiment of this application, a Gaussian function is used to perform multi-peak fitting processing on each target single PPG signal corresponding to the risk duration to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration, including:

[0109] A Gaussian function was used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration to obtain the Gaussian mean of the single PPG signal of each target.

[0110] Based on the Gaussian mean of the individual PPG signal for each target, the amplitude of the change in the Gaussian mean for each risk duration is obtained.

[0111] Optionally, for each risk duration, a multi-peak fitting process can be performed separately on each target individual PPG signal corresponding to that risk duration based on a Gaussian function. In this case, each target individual PPG signal will have its own corresponding Gaussian wave mean. Further, the maximum and minimum values ​​of the Gaussian wave mean of the target individual PPG signal within that risk duration can be determined. Then, the difference between the maximum and minimum values ​​can be used as the variation amplitude of the Gaussian wave mean corresponding to that risk duration, denoted by Δμ1.

[0112] In an optional embodiment of this application, a Gaussian function is used to perform multi-peak fitting processing on each target's individual PPG signal corresponding to the risk duration to obtain the Gaussian mean of each target's individual PPG signal, including:

[0113] For each target single PPG signal, the Gaussian function and the initial parameter values ​​of the function are obtained, and the target single PPG signal is processed based on the Gaussian function to obtain the function value corresponding to each data point in the target single PPG signal;

[0114] Based on the values ​​of each data point in the target single PPG signal and the corresponding function values ​​of each data point in the target single PPG signal, a corresponding target function is constructed;

[0115] Substitute the initial parameter values ​​of the function into the objective function and perform iterative calculations until the convergence condition is met to obtain the Gaussian mean of the target single PPG signal.

[0116] Optionally, each Gaussian function can be represented as:

[0117]

[0118] Where a is the amplitude, μ is the mean of the Gaussian wave, and σ is the standard deviation.

[0119] Furthermore, since each target's individual PPG signal typically requires three Gaussian functions for fitting, the Gaussian model of each target's individual PPG signal per cardiac cycle PPG(x) can be expressed as:

[0120]

[0121] Therefore, each Gaussian function requires three parameters: a, μ, and σ. For a single target PPG signal, since three Gaussian functions are needed to fit a cardiac cycle, nine parameters need to be estimated, namely a1, μ1, σ1, a2, μ2, σ2, a3, μ3, and σ3.

[0122] Furthermore, each data point in the target single PPG signal can be substituted into the Gaussian model of PPG(x) for each cardiac cycle to obtain the corresponding function value y(x). i Then, the least squares method is used to estimate the values ​​of these 9 parameters.

[0123] Optionally, the initial values ​​of the function parameters can be obtained, i.e., the initial values ​​of a1, μ1, σ1, a2, μ2, σ2, a3, μ3, σ3. In this case, the initial values ​​of the function parameters can be set according to prior knowledge:

[0124] [a1,μ1,σ1; a2,μ2,σ2; a3,μ3,σ3]=[1,70,0.2;0.8,100,0.5;0.5,150,1.0]

[0125] Furthermore, based on the values ​​of each data point in the target single PPG signal and the corresponding function values ​​of each data point in the target single PPG signal, a corresponding objective function can be constructed, which is the sum of squared residuals between each data point in the target single PPG signal and the Gaussian model:

[0126]

[0127] Among them, PPG(x i ) is the x in a single target PPG signal i The value of the data point at y(x) i Gaussian function in x i The value at that location.

[0128] Then, the initial parameter values ​​of the function can be substituted into the objective function for iterative calculation until the convergence condition is reached, thus obtaining the Gaussian mean of the target single PPG signal. The specific tools for iterative calculation and the convergence condition can be selected and set according to actual needs; this application does not impose any limitations on this. For example, the Gauss-Newton method in Matlab can be used to iteratively update the parameters until the set convergence condition is met or the maximum number of iterations is reached. At this point, the final values ​​of the nine parameters (i.e., [a1,μ1,σ1; a2,μ2,σ2; a3,μ3,σ3]) will be obtained. In this application, μ1 is extracted as the Gaussian mean of the target single PPG signal.

[0129] Step S105: Determine the patient's SAS result based on the blood oxygen change amplitude corresponding to each risk duration, the Gaussian mean change amplitude corresponding to each risk duration, and the judgment criteria.

[0130] Optionally, after obtaining the blood oxygen change amplitude and the Gaussian mean change amplitude corresponding to each risk duration, a pre-set judgment criterion can be obtained. Then, the blood oxygen change amplitude and the Gaussian mean change amplitude corresponding to each risk duration can be compared with the judgment criterion to obtain the patient's final SAS result.

[0131] In optional embodiments of this application, the judgment criteria include a product standard value and a low ventilation rate standard. Based on the blood oxygen saturation variation amplitude corresponding to each risk duration, the Gaussian mean variation amplitude corresponding to each risk duration, and the judgment criteria, the patient's SAS result is determined, including:

[0132] For each risk duration, the product of the blood oxygen change amplitude corresponding to the risk duration and the corresponding Gaussian mean change amplitude is processed to obtain the first value corresponding to the risk duration.

[0133] The first value corresponding to each risk duration is compared with the product standard to determine the number of target risk durations from all risk durations.

[0134] The number of target risk durations is compared with the criteria for low ventilation frequency to determine the patient's SAS outcome.

[0135] Optionally, the judgment criteria may include a product standard value (which can be expressed as ΔSpO2*Δμ1) and a low ventilation rate standard (which can be expressed as AHI) within a set time period. The specific values ​​of the product standard value and the low ventilation rate standard can be set by the patient according to their actual situation, and this application embodiment does not limit them. For example, since ΔSpO2 and Δμ1 are positively correlated, when sleep apnea syndrome (SAS) occurs, ΔSpO2 and Δμ1 will increase simultaneously. At this time, the value of ΔSpO2*Δμ1 will also increase significantly. Therefore, the judgment criteria can be set to a product standard value >200 and a low ventilation rate standard ≥5 times / hour as the SAS result.

[0136] Furthermore, the amplitude of blood oxygen saturation change (ΔSpO2) and the amplitude of the mean Gaussian wave change (Δμ1) corresponding to each risk duration can be multiplied to obtain a first value (i.e., the product value ΔSpO2*Δμ1). Then, the first value for each risk duration is compared with a product standard. Risk durations with a first value greater than the product standard are identified as target risk durations. Correspondingly, the number of target risk durations can be determined and compared with a hypoventilation rate standard. If the number of target risk durations reaches the hypoventilation rate standard, the patient is diagnosed with SAS; if the number of target risk durations does not reach the hypoventilation rate standard, the patient is diagnosed without SAS. For example, when the first value ΔSpO2*Δμ1 corresponding to a risk duration > 200, it is considered that one instance of hypoventilation has occurred, and the AHI value is incremented by 1. When the AHI ≥ 5 times / hour, SAS can be considered present.

[0137] In this embodiment, combining signal preprocessing and window minimum detection techniques, the continuous PPG signal is segmented into PPG signals of one cardiac cycle length (i.e., a single PPG signal) along the time axis. Then, Gaussian multi-peak fitting is used to decompose the single PPG signal to obtain the Gaussian mean, and the amplitude of the Gaussian mean variation is determined. Finally, based on the amplitude of the Gaussian mean variation and the amplitude of blood oxygen variation, it is determined whether the patient has SAS. That is, in this application, the Gaussian mean (μ1) is used instead of heart rate (HR), which can effectively ignore the interference of high-frequency signals, making the SAS detection results more reliable. This is especially suitable for environments with large heart rate variations or interference, providing an efficient and accurate heart rate variability detection technology for medical health monitoring.

[0138] To better understand the methods provided in the embodiments of this application, the following is combined with... Figure 4 The method and process described in this application will be explained again.

[0139] Step 401, PPG signal: Acquire the patient's initial continuous PPG signal and blood oxygen saturation data within a set time period;

[0140] Step 402, Preprocessing: The initial continuous PPG signal is processed to remove baseline drift interference based on a high-pass filter to obtain the first continuous PPG signal. Then, the first continuous PPG signal is processed to remove specific frequency interference based on a notch filter to obtain the target continuous PPG signal.

[0141] Step 403, trough detection and PPG segmentation: Determine the trough position of the target continuous PPG signal based on the window sliding minimum detection method, and segment the target continuous PPG signal into at least two individual PPG signals according to the trough position.

[0142] Step 404, Gaussian multi-peak fitting: Use a Gaussian function to perform multi-peak fitting on the single PPG signal of each target to obtain the Gaussian mean of the single PPG signal of each target;

[0143] Step 405, Feature Extraction: Calculate the mean Gaussian wave change amplitude (Δμ1) and blood oxygen change amplitude (ΔSpO2) of each target's individual PPG signal;

[0144] Step 406, Statistical analysis: Calculate ΔSpO2*Δμ1. When ΔSpO2*Δμ1>200, add 1 to the AHI value. When AHI≥5 times / hour, the patient is confirmed to have SAS.

[0145] This application provides a device for identifying SAS patients based on PPG signals, such as... Figure 5 As shown, the device may include: a data acquisition module 501, a signal cutting module 502, a data processing module 503, a data fitting module 504, and a result determination module 505, wherein,

[0146] The data acquisition module is used to acquire the patient's target continuous PPG signal and blood oxygen saturation data within a set time period. The target continuous PPG signal is the continuous PPG signal corresponding to the pulse data.

[0147] The signal cutting module is used to determine the trough position of the target continuous PPG signal based on the window sliding minimum detection method, and cut the target continuous PPG signal into at least two individual PPG signals according to the trough position.

[0148] The data processing module is used to determine at least one risk duration for the patient from a set duration based on blood oxygen saturation data, as well as the blood oxygen change range corresponding to each risk duration.

[0149] The data fitting module is used to perform multi-peak fitting on each individual PPG signal using a Gaussian function to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration.

[0150] The results determination module is used to determine the patient's SAS results based on the blood oxygenation variation amplitude corresponding to each risk duration, the Gaussian mean variation amplitude corresponding to each risk duration, and the judgment criteria.

[0151] Optionally, when acquiring the patient's target continuous PPG signal within a set time period, the data acquisition module is specifically used for:

[0152] Acquire the initial continuous PPG signal corresponding to the patient's pulse data within a set time period;

[0153] Obtain the cutoff frequency and high-pass filter, and determine the filter coefficients based on the cutoff frequency and high-pass filter.

[0154] The baseline drift interference is removed from the initial continuous PPG signal based on the filter coefficients to obtain the first continuous PPG signal.

[0155] Obtain the notch filter frequency and the notch filter itself, and determine the notch filter coefficients based on the notch filter frequency and the notch filter itself;

[0156] The first continuous PPG signal is processed to remove specific frequency interference based on the notch filter coefficient, and the target continuous PPG signal is obtained.

[0157] Optionally, when the signal cutting module determines the trough position of the target continuous PPG signal based on the window sliding minimum detection method, it is specifically used for:

[0158] Obtain the initial window. The number of sampling points contained in the initial window is determined based on the sampling rate of the target continuous PPG signal and the heart rate reference value.

[0159] Based on the initial window, a set distance is slid across the target continuous PPG signal each time to obtain at least one first window, and at least one second window is determined from the at least one first window according to the data point values ​​in each first window;

[0160] Based on the values ​​of the data points in each second window, determine the minimum data point corresponding to each second window, and determine the rising and falling trend between each minimum data point and its adjacent data points;

[0161] The position of the smallest data point whose rise and fall trend meets the preset requirements is taken as the trough position of the target continuous PPG signal.

[0162] Optionally, if the number of extremely small data points that meet the preset requirements for the rising and falling trend reaches a set number, and the sliding distance does not reach the end point of the target continuous PPG signal, the signal cutting module is also used for:

[0163] The initial window length and the set distance are adjusted to obtain the adjusted window and the adjusted distance. Based on the adjusted window and the adjusted distance, the trough position of the target continuous PPG signal is further determined.

[0164] Optionally, when the data processing module determines at least one risk duration corresponding to the patient from the set duration based on the blood oxygen saturation data, it is specifically used for:

[0165] Determine at least one duration during which the blood oxygen saturation value is lower than a preset value in the blood oxygen saturation data corresponding to a set duration;

[0166] The duration during which the blood oxygen level is below a preset value is considered as at least one risk duration for the patient.

[0167] Optionally, when the data fitting module performs multi-peak fitting on each individual PPG signal using a Gaussian function to obtain the mean Gaussian wave variation amplitude corresponding to each risk duration, it is specifically used for:

[0168] Based on the time information of each individual PPG signal, determine at least two target individual PPG signals corresponding to each risk duration from at least two individual PPG signals;

[0169] For each risk duration, a Gaussian function is used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration, so as to obtain the mean change amplitude of the Gaussian wave for each risk duration.

[0170] Optionally, when the data fitting module performs multi-peak fitting on the single PPG signal of each target corresponding to the risk duration using a Gaussian function to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration, it is specifically used for:

[0171] A Gaussian function was used to perform multi-peak fitting on the single PPG signal of each target corresponding to the risk duration to obtain the Gaussian mean of the single PPG signal of each target.

[0172] Based on the Gaussian mean of the individual PPG signal for each target, the amplitude of the change in the Gaussian mean for each risk duration is obtained.

[0173] Optionally, when the data fitting module performs multi-peak fitting processing on the individual PPG signal of each target corresponding to the risk duration using a Gaussian function to obtain the Gaussian mean of the individual PPG signal of each target, it is specifically used for:

[0174] For each target single PPG signal, the Gaussian function and the initial parameter values ​​of the function are obtained, and the target single PPG signal is processed based on the Gaussian function to obtain the function value corresponding to each data point in the target single PPG signal;

[0175] Based on the values ​​of each data point in the target single PPG signal and the corresponding function values ​​of each data point in the target single PPG signal, a corresponding target function is constructed;

[0176] Substitute the initial parameter values ​​of the function into the objective function and perform iterative calculations until the convergence condition is met to obtain the Gaussian mean of the target single PPG signal.

[0177] Optionally, the result determination module, when determining the patient's SAS result based on the judgment criteria including the product standard value and the low ventilation rate standard, and according to the blood oxygen saturation change amplitude corresponding to each risk duration, the Gaussian mean change amplitude corresponding to each risk duration, and the judgment criteria, is specifically used for:

[0178] For each risk duration, the product of the blood oxygen change amplitude corresponding to the risk duration and the corresponding Gaussian mean change amplitude is processed to obtain the first value corresponding to the risk duration.

[0179] The first value corresponding to each risk duration is compared with the product standard to determine the number of target risk durations from all risk durations.

[0180] The number of target risk durations is compared with the criteria for low ventilation frequency to determine the patient's SAS outcome.

[0181] The device for identifying SAS patients based on PPG signals in this embodiment can execute the method for identifying SAS patients based on PPG signals shown in the embodiment of this application. The implementation principle is similar and will not be described again here.

[0182] This application provides an electronic device, which includes: a processor; and a memory configured to store machine-readable instructions that, when executed by the processor, cause the processor to perform a method for identifying SAS patients based on PPG signals.

[0183] This application provides an electronic device, such as... Figure 6 As shown, Figure 6 The illustrated electronic device includes a processor 2001 and a memory 2003. The processor 2001 and the memory 2003 are connected, for example, via a bus 2002. Optionally, the electronic device 2000 may further include a transceiver 2004. It should be noted that in practical applications, the transceiver 2004 is not limited to one type, and the structure of this electronic device 2000 does not constitute a limitation on the embodiments of this application.

[0184] Processor 2001 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 2001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0185] Bus 2002 may include a pathway for transmitting information between the aforementioned components. Bus 2002 may be a PCI bus or an EISA bus, etc. Bus 2002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0186] The memory 2003 may be ROM or other type of static storage device capable of storing static information and instructions, RAM or other type of dynamic storage device capable of storing information and instructions, or EEPROM, CD-ROM or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0187] The memory 2003 stores the application code that executes the scheme of this application, and its execution is controlled by the processor 2001. The processor 2001 executes the application code stored in the memory 2003 to implement... Figure 5 The embodiment shown provides a device for determining the operation of a SAS patient based on PPG signals.

[0188] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0189] The above description is only a partial embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A device for identifying SAS patients based on PPG signals, characterized in that, include: The data acquisition module is used to acquire the patient's target continuous PPG signal and blood oxygen saturation data within a set time period, wherein the target continuous PPG signal is the continuous PPG signal corresponding to the pulse data; The signal cutting module is used to determine the trough position of the target continuous PPG signal based on the window sliding minimum detection method, and cut the target continuous PPG signal into at least two individual PPG signals according to the trough position. The data processing module is used to determine at least one risk duration corresponding to the patient from the set duration based on the blood oxygen saturation data, and the blood oxygen change range corresponding to each risk duration; The data fitting module is used to perform multi-peak fitting processing on each of the individual PPG signals using a Gaussian function to obtain the mean change amplitude of the Gaussian wave corresponding to each of the risk durations. The result determination module is used to determine the patient's SAS result based on the blood oxygen change amplitude corresponding to each risk duration, the Gaussian mean change amplitude corresponding to each risk duration, and the judgment criteria. The judgment criteria include a product standard value and a low ventilation rate standard. The patient's SAS result is determined based on the blood oxygen saturation variation amplitude corresponding to each risk duration, the Gaussian mean variation amplitude corresponding to each risk duration, and the judgment criteria, including: For each of the aforementioned risk durations, the product of the blood oxygen change amplitude corresponding to the risk duration and the corresponding Gaussian mean change amplitude is processed to obtain the first value corresponding to the risk duration. The first value corresponding to each of the risk durations is compared with the product standard to determine the number of target risk durations from all the risk durations. The patient's SAS outcome is determined by comparing the number of target risk durations with the low ventilation rate criteria.

2. The device for identifying SAS patients based on PPG signals according to claim 1, characterized in that, The acquisition of the patient's target continuous PPG signal within a set time period includes: Acquire the initial continuous PPG signal corresponding to the patient's pulse data within a set time period; Obtain the cutoff frequency and the high-pass filter, and determine the filter coefficients based on the cutoff frequency and the high-pass filter; Based on the filter coefficients, the initial continuous PPG signal is processed to remove baseline drift interference, resulting in a first continuous PPG signal. Obtain the notch frequency and the notch filter, and determine the notch filter coefficient based on the notch frequency and the notch filter; The first continuous PPG signal is processed to remove specific frequency interference based on the notch filter coefficients to obtain the target continuous PPG signal.

3. The device for identifying SAS patients based on PPG signals according to claim 1, characterized in that, The method for determining the trough position of the target continuous PPG signal based on the window sliding minimum detection method includes: An initial window is obtained, the number of sampling points contained in the initial window is determined based on the sampling rate of the target continuous PPG signal and the heart rate reference value; Based on the initial window sliding a set distance on the target continuous PPG signal each time, at least one first window is obtained, and at least one second window is determined from the at least one first window according to the data point value in each first window; Based on the value of each data point in the second window, determine the minimum data point corresponding to each second window, and determine the rising and falling trend between each minimum data point and its adjacent data points; The position of the smallest data point whose rise and fall trend meets the preset requirements is taken as the trough position of the target continuous PPG signal.

4. The device for identifying SAS patients based on PPG signals according to claim 3, characterized in that, If the number of extremely small data points that meet the preset requirements for the upward and downward trend reaches a set number, and the sliding distance does not reach the endpoint of the target continuous PPG signal, the method further includes: The window length of the initial window and the size of the set distance are adjusted respectively to obtain the adjusted window and the adjusted distance, and the trough position of the target continuous PPG signal is further determined based on the adjusted window and the adjusted distance.

5. The device for identifying SAS patients based on PPG signals according to claim 1, characterized in that, The step of determining at least one risk duration corresponding to the patient from the set duration based on the blood oxygen saturation data includes: Determine at least one duration in the blood oxygen saturation data corresponding to the set duration where the blood oxygen value is lower than a preset value; The duration during which the blood oxygen value is lower than a preset value is taken as at least one risk duration for the patient.

6. The device for identifying SAS patients based on PPG signals according to claim 1, characterized in that, The step of performing multi-peak fitting processing on each individual PPG signal using a Gaussian function to obtain the mean Gaussian wave variation amplitude corresponding to each risk duration includes: Based on the time information of each individual PPG signal, at least two target individual PPG signals corresponding to each risk duration are determined from the at least two individual PPG signals; For each of the aforementioned risk durations, a Gaussian function is used to perform multi-peak fitting on each of the target single PPG signals corresponding to the risk duration, thereby obtaining the mean change amplitude of the Gaussian wave for each of the aforementioned risk durations.

7. The device for identifying SAS patients based on PPG signals according to claim 6, characterized in that, The step of performing multi-peak fitting processing on each target single PPG signal corresponding to the risk duration using a Gaussian function to obtain the mean change amplitude of the Gaussian wave corresponding to each risk duration includes: A Gaussian function is used to perform multi-peak fitting on each target single PPG signal corresponding to the risk duration to obtain the Gaussian mean of each target single PPG signal. The mean Gaussian wave value of each risk duration is obtained based on the mean Gaussian wave value of each individual PPG signal of the target.

8. The device for identifying SAS patients based on PPG signals according to claim 7, characterized in that, The step of performing multi-peak fitting processing on each target individual PPG signal corresponding to the risk duration using a Gaussian function to obtain the Gaussian mean of each target individual PPG signal includes: For each target single PPG signal, a Gaussian function and initial parameter values ​​are obtained, and the target single PPG signal is processed based on the Gaussian function to obtain the function value corresponding to each data point in the target single PPG signal; Based on the values ​​of each data point in the target single PPG signal and the function values ​​corresponding to each data point in the target single PPG signal, a corresponding target function is constructed; The initial parameter values ​​of the function are substituted into the objective function for iterative calculation until the convergence condition is reached, thereby obtaining the Gaussian mean of the target single PPG signal.

Images