A portable blood oxygen signal OSA intelligent detection method

Abstract

The present application relates to the technical field of medical monitoring, and discloses a portable blood oxygen signal OSA intelligent detection system, which comprises the following steps: collecting a pulse wave signal through a blood oxygen sensor, generating an amplitude difference sequence and a period difference sequence in parallel, analyzing the rhythm synchronism of the two, and determining a respiratory effort event when a real-time relationship index deviates from a reference value. The present application captures the pulse wave timing-amplitude decoupling phenomenon, realizes direct sensing of the specific mode process of the invalid respiratory effort signal, avoids the traditional lag monitoring mode which relies on blood oxygen decline, has the inherent ability to distinguish physiological fluctuations from the specific mode struggle of the signal, and provides a new solution for early screening of mild and moderate OSA.

Description

TECHNICAL FIELD

[0001] The present application relates to a portable blood oxygen signal OSA intelligent detection method, belonging to the technical field of medical monitoring. BACKGROUND

[0002] In the field of portable sleep respiratory disorder remote monitoring, the existing technology mainly relies on physiological result signals such as blood oxygen saturation drop and heart rate variation as the basis for determining obstructive sleep apnea (OSA). Whether through polysomnography or simplified portable devices, the common limitation is focusing on the lagging physiological compensation phenomenon after airway obstruction, such as blood oxygen reduction or pulse rate fluctuation, but it cannot capture the ineffective respiratory effort process that occurs in the early stage of obstruction. When facing mild to moderate OSA patients, such schemes fail in the night struggle stage when the patient has not yet appeared significant indicator abnormalities, leading to a large number of missed diagnoses of hidden cases.

[0003] Further analysis shows that the existing technology has a fundamental constraint: its signal processing logic passively relies on physiological results rather than the specific pattern of the signal itself. For example, conventional schemes eliminate respiratory pulse wave modulation through filtering, but ignore the real-time airway dynamics information contained in the modulation signal. Even if a multi-sensor fusion strategy is introduced to improve accuracy, it is still difficult to deploy on edge devices due to the reliance on complex algorithms and high computing power, and it cannot distinguish between physiological fluctuations such as body position changes and the essence of the specific pattern of the signal respiratory struggle.

[0004] Specifically, the existing technology faces three core bottlenecks: 1. It can only identify the obstruction results that have occurred, and cannot perceive the ongoing respiratory effort; 2. It is difficult to distinguish the essential difference between physiological interference and the specific pattern of the signal struggle; 3. Complex models are difficult to adapt to the hardware constraints of large-scale home screening scenarios. Therefore, how to provide a method that can sensitively reflect the rhythm changes in the pulse wave signal related to respiration and the specific pattern of the signal has become a technical problem to be solved by the present application. SUMMARY

[0005] The present application provides a portable blood oxygen signal OSA intelligent detection method, which mainly aims to solve the problem of how to directly capture the specific pattern process of the respiratory effort signal through a single blood oxygen signal and achieve high robustness classification and recognition of edge devices.

[0006] To achieve the above purpose, the present application provides a portable blood oxygen signal OSA intelligent detection method, which comprises the following steps:

[0007] Step a, continuously collecting photoplethysmogram signals and detecting the pulse wave peak of the photoplethysmogram signals in real time to obtain a pulse wave peak sequence;

[0008] Step b, generating, in parallel, an amplitude difference sequence reflecting the amplitude variation of the pulse wave and a period difference sequence reflecting the period interval variation of the pulse wave based on the sequence of pulse wave peaks;

[0009] Step c, analyzing the rhythm synchronism of the amplitude difference sequence and the period difference sequence within a preset time window, the rhythm synchronism being obtained by calculating the sign consistency ratio of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence, the sign consistency ratio representing the sign consistency degree of the two sequences;

[0010] Step d, comparing the real-time obtained sign consistency ratio with a reference sign consistency ratio, the reference sign consistency ratio being determined under the stable breathing state at the start of the user monitoring;

[0011] Step e, when the real-time sign consistency ratio continuously falls below a first preset threshold of the reference sign consistency ratio, determining that the current signal segment is a rhythm desynchronism state signal segment.

[0012] Preferably, in step c, the sign consistency ratio is calculated by counting the number of times that the product of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence is positive within a preset time window, and taking the ratio of the number of times to the total number of heartbeats within the time window as the sign consistency ratio.

[0013] Preferably, the reference sign consistency ratio is adaptively established by calculating the moving average of the sign consistency ratio under the stable breathing state at the start of the user monitoring.

[0014] Preferably, the method further comprises the following step: when the cumulative duration or the occurrence frequency of the determined rhythm desynchronism state signal segment reaches or exceeds a preset reference threshold, outputting a high-risk level warning.

[0015] Preferably, after determining the occurrence of the rhythm desynchronism state signal segment in step e, the method further comprises the following steps: calculating the second-order difference value of the amplitude difference sequence and the second-order difference value of the period difference sequence at the time point of the occurrence of the rhythm desynchronism state signal segment; if either of the absolute value of the second-order difference value of the amplitude difference sequence or the absolute value of the second-order difference value of the period difference sequence is greater than a preset morphological discrimination threshold, marking the event as a high-frequency interference segment and reducing or excluding its weight in the calculation of the risk index.

[0016] Preferably, the preset morphological discrimination threshold is set according to historical physiological data, and is used to distinguish the relatively gentle variation caused by respiratory effort from the instantaneous sharp variation caused by high-frequency signal disturbance of non-respiratory origin.

[0017] Preferably, the method further comprises the steps of: continuously monitoring the body movement of the user through the acceleration sensor; and suspending the determination of step e when the amplitude of the body movement is detected to be greater than a preset body movement threshold, so as to avoid the interference of the body movement artifact on the determination of the rhythm desynchronization state signal segment.

[0018] Preferably, the method further comprises the steps of: continuously calculating the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence during the time period in which step e does not determine the occurrence of the rhythm desynchronization state signal segment, the modulation energy being obtained by calculating the variance of the difference sequence in the time period; and identifying a specific signal segment representing signal modulation energy silence when the modulation energy E A of the amplitude difference sequence and the modulation energy E T of the period difference sequence are both continuously lower than the preset lower limit threshold of the respective energy baseline within a preset duration.

[0019] Preferably, the lower limit threshold of the energy baseline is set to a lower limit proportion value of the corresponding modulation energy in the normal and stable sleep state of the user, and the lower limit proportion value ranges from 0.1 to 0.3.

[0020] Preferably, the modulation energy baseline in the normal and stable sleep state is adaptively established by moving average of the variances of the amplitude difference sequence and the period difference sequence in the stable breathing state of the user in the initial monitoring stage.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] 1. The present application analyzes the rhythm coupling relationship between the pulse wave amplitude and the period, and proposes a new signal feature (such as the sign consistency ratio), which can reflect the change of the breathing rhythm earlier and more sensitively than the traditional blood oxygen saturation index. This method provides a new technical means for dynamic monitoring of the breathing state, and can effectively distinguish the signal interference caused by body movement, etc. At the same time, based on the sign consistency analysis of the pulse wave difference sequence, the system naturally distinguishes the physiological fluctuations (such as turning over and sleep stage conversion) from the specific mode of the signal. When the body movement sensor suspension logic is introduced synchronously, the determination process is further controlled dynamically by the action amplitude threshold, so that the body movement noise no longer triggers false judgment. The cooperation of this timing rhythm analysis and physical movement monitoring establishes an anti-interference double barrier from the signal essence level, and still maintains the recognition reliability in the home multi-disturbance environment.

[0023] 2. By reusing the same set of pulse wave characteristic data streams, the system can identify obstructive events while simultaneously activating the central apnea judgment logic by monitoring the abnormal attenuation of differential sequence modulation energy. This specific patterned interpretation of the signal silence enables a single pulse oximeter to distinguish between obstructive and central apnea for the first time, improving the differentiation between different signal patterns and providing richer technical information for subsequent analysis by professionals.

[0024] 3. The rhythm synchronization baseline established during the user's resting phase is adaptively updated through moving average during continuous monitoring. When combined with the second-order differential morphological analysis of the pulse wave differential sequence, the system can autonomously identify transient artifacts caused by high-frequency signal disturbances from non-respiratory sources, thus avoiding ECG interference from contaminating the respiratory event database. This time-series management mechanism, which integrates calibration, monitoring, and verification, significantly reduces the equipment's dependence on idealized operating conditions. Attached Figure Description

[0025] Figure 1 This is a timing diagram for the body movement amplitude interference identification and respiratory analysis control of the present invention;

[0026] Figure 2 This is a time-series analysis diagram of the rhythm and modulation activity of the present invention, wherein... Figure 2 (a) is a schematic diagram showing the change of rhythm synchronicity ratio over time. Figure 2 (b) is a schematic diagram of modulatory activity at different sleep stages;

[0027] Figure 3 This is a flowchart of the respiratory event determination process of the present invention;

[0028] Figure 4 This is a 24-hour monitoring curve of the symbol consistency ratio of this invention;

[0029] Figure 5 This is a comparative analysis diagram of the differential sequence modulation energy of the present invention;

[0030] Figure 6 This is a timing diagram of the body motion interference and system state of the present invention.

[0031] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0032] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0033] This application provides a portable OSA (Oxygen Spectrum Absorption) intelligent detection method, which includes the following steps:

[0034] Step a, continuously collecting the photoplethysmogram signal and detecting the pulse wave peak of the photoplethysmogram signal in real time to obtain a pulse wave peak sequence;

[0035] Step b, generating an amplitude difference sequence reflecting the amplitude change of the pulse wave and a period difference sequence reflecting the period interval change of the pulse wave based on the pulse wave peak sequence in parallel;

[0036] Step c, analyzing the rhythm synchronism of the amplitude difference sequence and the period difference sequence within a preset time window, the rhythm synchronism being obtained by calculating the sign consistency ratio of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence, the sign consistency ratio representing the sign consistency degree of the two sequences;

[0037] Step d, comparing the real-time obtained sign consistency ratio with a reference sign consistency ratio, the reference sign consistency ratio being determined under the stable breathing state of the user in the initial stage of the user monitoring;

[0038] Step e, when the real-time sign consistency ratio continuously falls below a first preset threshold of the reference sign consistency ratio, determining that the current signal segment is a rhythm desynchronization state signal segment.

[0039] Preferably, in step c, the sign consistency ratio is calculated in the following manner: within the preset time window, the number of times that the product sign of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence is positive is counted, and the ratio of the number of times to the total number of heartbeats in the time window is taken as the sign consistency ratio.

[0040] Preferably, the reference sign consistency ratio is adaptively established by calculating the moving average of the sign consistency ratio under the stable breathing state of the user in the initial stage of the user monitoring.

[0041] Preferably, the method further comprises the following step: when the cumulative duration or the occurrence frequency of the determined rhythm desynchronization state signal segment reaches or exceeds a preset reference threshold, outputting a high-risk level warning.

[0042] Preferably, after the rhythm desynchronization state signal segment is determined in step e, the method further comprises the following steps: calculating the second-order difference value of the amplitude difference sequence and the second-order difference value of the period difference sequence at the time point of the occurrence of the rhythm desynchronization state signal segment; if either of the absolute value of the second-order difference value of the amplitude difference sequence or the absolute value of the second-order difference value of the period difference sequence is greater than a preset morphological discrimination threshold, marking the event as a high-frequency interference segment and reducing or excluding the weight of the high-frequency interference segment in the calculation of the risk index.

[0043] Preferably, the preset morphological discrimination threshold is set according to historical physiological data and is used to distinguish the relatively gentle change caused by respiratory effort from the instantaneous sharp change caused by high-frequency signal disturbance of non-respiratory origin.

[0044] Preferably, the method further comprises the steps of continuously monitoring the user's body movement by the acceleration sensor; and suspending the determination of step e when detecting a body movement amplitude greater than a preset body movement threshold, to avoid the interference of body movement artifact on the determination of the rhythm desynchronization state signal segment.

[0045] Preferably, the method further comprises the steps of continuously calculating the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence during the time period when step e does not determine the occurrence of the rhythm desynchronization state signal segment, the modulation energy being obtained by calculating the variance of the difference sequence within the time period; and identifying a specific signal segment representing signal modulation energy silence when the modulation energy E A of the amplitude difference sequence and the modulation energy E T of the period difference sequence are both continuously below the respective energy baseline lower threshold within a preset duration.

[0046] Preferably, the energy baseline lower threshold is set to a lower limit proportion value of the corresponding modulation energy in the normal and stable sleep state of the user, and the lower limit proportion value ranges from 0.1 to 0.3.

[0047] Preferably, the modulation energy baseline in normal stable sleep state is adaptively established by moving average of the variance of the amplitude difference sequence and the period difference sequence in the stable breathing state at the beginning of user monitoring; meanwhile, in the present application, the setting of the reference symbol consistency ratio is based on the stable breathing state at the initial monitoring stage of the user, and the moving average value of the symbol consistency ratio at this time is calculated to adaptively establish the reference value, which can dynamically adjust the reference value to adapt to individual physiological differences. In specific application, the preliminary reference value is calculated by stable breathing state data in the resting state of the user at the initial monitoring stage, and then the system will adaptively update the reference value according to the long-term monitoring data of the user, and the update period of the reference value is closely related to the stability of the monitoring data. When the system detects that the breathing state changes significantly, the reference value will be corrected according to the new data to ensure that the breathing characteristics of each user are fully considered; as for the preset threshold, the first consideration is the adaptability to individual differences of different users, and the setting of these thresholds is not only based on the percentage change of the reference symbol consistency ratio, but also needs to consider the individual physiological characteristics of the user. The selection range of the preset threshold is usually between 20% and 30% of the reference value, and the specific range is optimized according to the actual test and physiological response of the user. The threshold range is based on the engineering trade-off between the normal fluctuation of respiratory rhythm change and the specific mode change of possible signals to ensure that the system can effectively identify the respiratory abnormalities of mild to moderate OSA patients while avoiding misjudgment of normal physiological fluctuations. The reasonable setting of the threshold is to achieve the best balance between sensitivity and specificity to ensure the accuracy of identification; in event identification, the present application analyzes the rhythm synchrony of the amplitude difference sequence and the period difference sequence to judge the ineffective respiratory effort event and the rhythm desynchronization state signal segment. The ineffective respiratory effort event usually refers to the situation that when the airway is blocked, the patient's respiratory effort fails to effectively improve the airflow, and the feature of this event is the occurrence of rhythm decoupling phenomenon, which is reflected by the significant decrease of the symbol consistency ratio. This phenomenon can be detected at the initial stage before the decrease of blood oxygen saturation, which is an important basis for early screening of mild to moderate OSA; while the rhythm desynchronization state signal segment refers to the respiratory effort event caused by complete or partial obstruction of the airway, which is characterized by continuous decrease of the symbol consistency ratio and increase of physiological interference when monitored by blood oxygen signal. Therefore, by comparing the rhythm synchrony ratio with the reference value, the system can effectively distinguish between the two events. In order to ensure the accuracy of this process, the system will also combine the second-order difference calculation and morphological discrimination mechanism to identify different types of artifacts, for example, when the second-order difference value of the signal exceeds the preset morphological discrimination threshold, the event will be marked as a cardiogenic artifact and excluded from the final determination.The setting basis of the morphological discrimination threshold is the fluctuation rule of historical physiological data, aiming to distinguish the gentle fluctuation caused by respiratory effort from the sharp fluctuation caused by high-frequency signal disturbance of non-respiratory source; and in the present application, in order to prevent the interference of body motion artifact on the judgment of rhythm desynchronization state signal segment, the system monitors the user's body motion in real time through the acceleration sensor, and decides whether to suspend event judgment according to the preset body motion amplitude threshold. When the body motion amplitude exceeds the preset body motion threshold, the system suspends the respiratory event judgment of the current data segment. The selection standard of the body motion threshold is to set it according to the typical body motion amplitude variation range of the device wearing position. Specifically, under normal circumstances, the body motion amplitude threshold is set at about 0.5G to ensure that within the normal body motion range, it will not be misjudged as an artifact, and the setting of body motion monitoring takes into account the adaptability in different use environments, for example, in different sleep stages, sleep postures and living environments, the body motion amplitude will vary, therefore, the system will adjust the body motion threshold according to the long-term monitoring data of the user, to ensure that in the changing environment, the influence of body motion artifact is minimized, all of which belong to the extended embodiments known to those skilled in the art.

[0048] Embodiment 1: The embodiment proposes a portable blood oxygen signal OSA intelligent detection method, based on the photoplethysmogram signal collected by a single blood oxygen sensor, by extracting the pulse wave peak value sequence, and constructing the corresponding amplitude difference sequence and period difference sequence, further calculating the synchronization degree of the two types of difference sequences in the rhythm level, and identifying the rhythm desynchronization state signal segment according to this, on this basis, combining the morphological features to exclude artifacts, and determining the specific signal segment representing signal modulation energy silence when the specific cumulative condition is met; Specifically, in the signal acquisition stage, a conventional photoelectric volume sensor is used to realize continuous sampling of the user's pulse wave signal, usually through photoelectric detection at the fingertip or earlobe position, to obtain the pulse waveform reflecting the heart beat activity, the system extracts the peak value of the pulse waveform in real time to construct the pulse wave peak value sequence, which provides basic data for subsequent difference processing, and its stability directly affects the accuracy of the entire determination process, in actual implementation, multi-scale smoothing and false peak correction strategy can be used to improve the peak extraction robustness under low signal-to-noise ratio conditions; In addition, in the construction process of the pulse wave peak value sequence, in order to ensure the stability and accuracy of the peak extraction, the amplitude threshold is set based on the average pulse wave peak-valley difference of the input signal in the static segment, which is typically set to fifty percent of the average value, to eliminate the false peak response caused by instantaneous disturbance or noise, with this threshold strategy, the system introduces a false peak correction mechanism, specifically including: first, after preliminary detection of the peak candidate point, set a time window according to the typical duration of adjacent heart cycles, and judge whether the candidate point constitutes a local maximum center; If there are multiple adjacent candidate points, the one with the maximum amplitude in the local time window is retained, and the others are removed; Secondly, for the case where the distance between adjacent peaks is significantly lower than the lower limit of the average heart cycle, the system will trigger the adjacent peak re-alignment and replacement process to correct the false dense peak structure induced by high-frequency interference, and ensure that the finally formed pulse wave peak value sequence has good time alignment characteristics and amplitude representativeness.Based on the above peak sequence, the system respectively constructs two difference sequences: one is the amplitude difference sequence used to represent the amplitude change between consecutive pulse wave peaks, and the elements thereof are composed of the difference between adjacent peaks; the other is the period difference sequence used to reflect the change of the interval between peaks, and the elements thereof are composed of the change amount of consecutive heart periods; the two difference sequences respectively reveal the local fluctuation of the signal from the amplitude domain and the time domain, and are the key feature dimensions for identifying the specific mode of the signal respiratory effort state; the system sets a sliding time window for calculating the rhythm synchronization index between the two difference sequences in real-time monitoring, i.e. the sign consistency ratio, which is defined as: in the time window, the number of times that the product sign of the corresponding elements of the amplitude difference sequence and the period difference sequence is positive is counted, and the ratio of the number of times to the total number of heartbeats in the time window is obtained, which is the sign consistency ratio. The ratio can be used to quantify the consistency of the two difference sequences in the sign fluctuation direction, thereby reflecting the rhythm coupling relationship of the respiratory modulation signal in the pulse wave.

[0049] In the normal breathing state, there is a relatively stable rhythm synchronization between the amplitude difference and the period difference, and the sign consistency ratio usually remains at a high level. When the rhythm desynchronization state signal segment occurs, due to the abnormal fluctuation of intrapulmonary negative pressure caused by airway obstruction, the rhythm decoupling phenomenon occurs in the circulation dynamics and neural regulation mechanism, thereby causing a significant decrease in the above ratio. Therefore, the ratio is an important dynamic indicator for identifying the specific mode of such signals. To achieve individualized judgment, the system establishes a baseline value of the sign consistency ratio in the stable breathing state of the user in the initial monitoring stage. The baseline value is generated by moving average of the ratio values calculated in the stable period, and has certain adaptive ability to adapt to individual physiological differences and state changes. The establishment method of the baseline value avoids the adaptability problem caused by static setting, and can be dynamically updated in the long-term monitoring process. In real-time detection, if the system finds that the sign consistency ratio of the current period is continuously lower than the baseline value minus a first preset threshold, and the decline state lasts for more than a set minimum time length, it is considered that the rhythm desynchronization state signal segment may occur. The preset threshold is preferably set as a proportional decrease of the baseline value, rather than a fixed value, to enhance the adaptability to different individuals. For example, the proportion can be selected between twenty percent and thirty percent, and the setting logic is to sensitively identify the specific mode of the signal abnormality while reducing the misjudgment probability of normal physiological fluctuations.

[0050] To further improve the accuracy of the determination, after the occurrence of the above preliminary determination event, the system will also calculate the second-order difference value of the amplitude difference sequence and the period difference sequence respectively, and judge whether the absolute value exceeds the morphological identification threshold value, if the absolute value of the second-order difference value exceeds the threshold value, the event can be determined as a false event caused by cardiogenic factors, so as to be excluded, the morphological identification threshold value is set according to historical physiological data, which is usually used to distinguish the relatively slow change phenomenon caused by respiratory effort from the sudden event caused by high-frequency signal disturbance of non-respiratory source, so as to enhance the identification ability of the system to non-respiratory factor interference; and the application also introduces body motion detection logic, which continuously monitors the body motion state of the user through the acceleration sensor, when the detection amplitude exceeds the preset body motion threshold value, the system will suspend the determination process of the above rhythm desynchronization state signal segment, so as to prevent the body motion artifact from interfering with the detection result, the mechanism builds a cooperative barrier between the rhythm analysis at the signal level and the interference monitoring at the physical behavior level, and enhances the practical adaptability of the method in daily life environment.Further, during the time period when no rhythm desynchronization state signal segment is detected, the system will continue to calculate the modulation energy of the amplitude difference sequence and the period difference sequence, which is obtained by calculating the variance of the difference sequence in this time period, to quantify the dynamic activity level of the pulse wave signal. When the two modulation energy values simultaneously and continuously remain below the lower threshold of the respective energy baseline for a set time length, a specific signal segment representing signal modulation energy silence is identified. During the time period when no rhythm desynchronization state signal segment is determined to occur, to ensure that the identification of the specific signal segment representing signal modulation energy silence has clear engineering boundaries, the duration of the continuous calculation of the modulation energy of the amplitude difference sequence and the period difference sequence can be set to be no less than the time length of one complete respiratory rhythm cycle, which usually corresponds to a sliding analysis window of thirty to sixty seconds. The selection of this time length is based on the mechanism consideration that the specific signal segment representing signal modulation energy silence has obvious persistent silence characteristics in physiological performance. During this period, the loss of respiratory center drive leads to the decline of airflow, respiratory movement and blood flow modulation, thereby causing the double significant attenuation of modulation energy in the pulse waveform. If the modulation energy values of the amplitude difference sequence and the period difference sequence continuously remain below the lower threshold of the energy baseline within the above-mentioned time window, and this state continues uninterrupted until the end point of the window, the system determines that this signal segment is the occurrence interval of the specific signal segment representing signal modulation energy silence. This determination logic is based on the characteristics of central apnea leading to complete interruption of respiratory drive, which in turn manifests as the silent decline of the signal. Therefore, low modulation energy becomes an important indicator of this type of event. The lower threshold of the modulation energy is set according to the corresponding modulation energy level of the user in the normal and stable sleep state, which is usually ten to thirty percent of this level as the proportional lower limit. At the same time, this energy baseline value can also be adaptively generated by moving average at the initial stage of monitoring, to ensure that it has the determination ability matched with the individual characteristics of the user.

[0051] Meanwhile, in the rhythm analysis process of the difference sequence, the calculation of the sign consistency ratio is based on the corresponding relationship between the amplitude difference sequence and the period difference sequence in the time window, the sign product of the elements at the same time is judged respectively, and the ratio index is constructed based on the statistical result of the positive sign, and then the coordination of the two in the rhythm fluctuation direction is quantified. In this process, the elements of each sequence are marked with time sequence number to ensure the uniqueness of the one-to-one correspondence, thereby avoiding the risk of misjudgment caused by sequence misplacement. The setting logic of the ratio is not only based on mathematical operation, but also derived from the physiological coupling characteristics of the pulse wave signal under the respiratory driving state: in the normal state, the pulse wave amplitude and period modulation have the trend of synchronous rise or fall; while in the specific mode state of the signal or the respiratory rhythm disorder, the decoupling behavior with inconsistent direction often occurs, and the sign consistency ratio decreases significantly; at the same time, the setting of the first preset threshold introduces a dynamic trade-off mechanism. The fundamental technical consideration of the threshold is to achieve the optimal engineering balance between the sensitive recognition ability of the rhythm decoupling state and the overall anti-interference performance of the system. If the threshold is set too high, the system may miss the invalid respiratory effort signal in the mild signal specific mode stage; on the contrary, if the threshold is too low, the system may be too sensitive to short-time noise or body motion disturbance, resulting in an increase in false positive rate. Therefore, in actual deployment, the threshold is selected based on the reference sign consistency ratio established in the stable breathing state at the beginning of monitoring, and the relative judgment threshold is set by the proportion of the decrease, typically in the range of twenty to thirty percent of the reference value, to ensure that the algorithm has sufficient recognition sensitivity and judgment stability. To further improve the stability of the system under low signal-to-noise ratio conditions, the signal acquisition module introduces a multi-scale smoothing mechanism and a pseudo-peak correction strategy. In the pulse wave peak extraction process, local extreme value judgment combined with amplitude threshold control is used, and the threshold value is taken from fifty percent of the average peak-to-valley difference in the resting section to filter out high-frequency pseudo-peak. At the same time, by jointly judging the time interval and amplitude gradient of adjacent peaks, a peak validity screening mechanism is established to ensure that the pulse wave peak sequence has good time alignment and representativeness at the input end. The introduction of the above strategy establishes a robust acquisition link based on signal stability and constrained by feature continuity, providing reliable support for subsequent difference calculation and rhythm analysis. For key parameters such as the sign consistency ratio, the second difference value of the amplitude difference sequence and the period difference sequence, and the lower threshold of the energy baseline, the system does not use absolute constant setting, but generates adaptively through sliding average and proportional setting mechanism based on the stable state data obtained in the initial monitoring stage of the individual. This method of establishing parameter benchmarks based on individual dynamic characteristics avoids the adaptability defects caused by static setting, and allows the system to iteratively optimize according to the actual monitoring data in long-term operation, thereby establishing a high-robustness judgment framework with endogenous feedback regulation ability.

[0052] Embodiment 2: The embodiment discloses a specific implementation process of a portable blood oxygen signal OSA intelligent detection method, the process is based on a photoplethysmogram signal collected by a single blood oxygen sensor, and aims to realize reliable detection of obstructive respiratory effort and a specific signal segment representing signal modulation energy silence without relying on additional sensors. The core of the method is that: by analyzing the rhythm synchronism between the pulse wave amplitude and the cycle difference signal, the invalid respiratory effort process is identified in time; and combining morphological feature analysis and body motion monitoring information, artifact recognition and noise suppression are realized, so that accurate determination is realized on an edge computing platform with low operation resources; the whole method includes the following processing stages: first, the system adopts a photoplethysmography sensor to collect a pulse wave signal, the typical wearing position of the sensor is a fingertip or an earlobe, and the collected original signal is first subjected to multi-scale smoothing processing to improve the signal stability; then the pulse wave peak value is extracted through a strategy based on neighborhood extreme value comparison and amplitude threshold constraint, and a peak value sequence is formed, the sequence is used as the basis for subsequent difference analysis, and the stability of the sequence directly affects the accuracy of the overall determination; on the basis of the peak value sequence, the system generates two types of difference sequences in parallel: one is an amplitude difference sequence, which is used to reflect the amplitude change between adjacent pulse wave peaks; the other is a cycle difference sequence, which is used to represent the time interval fluctuation between consecutive heartbeats. The two types of difference sequences respectively reveal the local dynamic characteristics of the pulse wave from the amplitude domain and the time domain. In a set sliding time window, the system calculates the number of times that the product of corresponding elements of the two types of difference sequences is positive, and compares it with the total number of heartbeats in the time period, to obtain a sign consistency ratio as a quantitative indicator of rhythm synchronism.

[0053] To establish individualized decision criteria, the system automatically selects a stable breathing period of data in the initial resting state of the user wearing the device, obtains the sign consistency ratio in this period through moving average, and thus constructs the baseline value of the user. In the subsequent real-time monitoring process, if the current sign consistency ratio continuously falls below the relative threshold value of the baseline value minus a preset proportion (such as 20%), and lasts for a set minimum time length, the system determines that the rhythm desynchronization state signal segment may occur. The threshold value is set in the form of relative ratio, which is beneficial to adapt to physiological differences between individuals, and improves the universality and practicality of the system. To further exclude waveform abnormal interference caused by non-breathing factors, the system performs second-order difference calculation on the amplitude difference sequence and the period difference sequence respectively after identifying the suspected event. If the absolute value of any second-order difference sequence exceeds the morphological identification threshold value, it is considered that the event may be caused by cardiac factors, and is determined as an artifact and is excluded. The morphological identification threshold value is set based on the physiological fluctuation rules disclosed in the art, and is commonly used to distinguish slowly changing signals caused by respiratory effort from sudden signals caused by abnormal heart rhythm. Considering the signal disturbance that may be caused by the change of user's body position during night sleep, the system further integrates an acceleration sensor to monitor the body movement state. If the body movement amplitude exceeds the preset threshold value, which is set according to the body movement amplitude range of the typical wearing position, the system will temporarily suspend the rhythm desynchronization state signal segment determination process, and then resume the determination operation after the body movement tends to be stable. This processing mechanism realizes the cooperative control between signal layer rhythm analysis and physical layer interference suppression, and improves the anti-interference ability of the overall detection system in the daily environment.

[0054] In the time period in which the rhythm desynchronization state signal segment is not determined to occur, the system also continuously evaluates the modulation activity of the difference signal. Specifically, the system calculates the variance of the amplitude difference sequence and the period difference sequence in a set time window to measure the degree of dynamic change of the pulse wave. When the modulation energy of both is simultaneously below the preset lower limit of the baseline value of each (such as 10% to 30%) in a continuous period of time, the system determines that the state is a specific signal segment representing signal modulation energy silence. The above baseline value can be adaptively determined by the user's fluctuation level in the initial resting period to adapt to individual differences and improve the judgment accuracy. Through the orderly cooperation of the above stages, this method can realize high robust identification of obstructive and central respiratory events based on a single sensor, avoiding the dependence on traditional blood oxygen saturation lag indicators, and has strong real-time and identification sensitivity.

[0055] Embodiment 3: This embodiment constructs a platform for portable blood oxygen signal acquisition and analysis, which contains the following main components: a PPG signal acquisition module, which adopts a photoelectric volume sensor, is worn on the subject's fingertip through a finger clip design, continuously acquires the original photoelectric volume pulse wave signal, and the sensor working frequency can be adjusted according to actual needs, aiming to ensure the stability of signal acquisition; a data preprocessing and feature extraction module: this module processes the original photoelectric volume pulse wave signal to detect the pulse wave peak value of the photoelectric volume pulse wave signal in real time, so as to obtain the pulse wave peak value sequence, the processing process can contain multi-scale smoothing and pseudo-peak correction strategy, aiming to improve the robustness of pulse wave peak value extraction under low signal-to-noise ratio conditions; a difference sequence generation and rhythm synchrony analysis module, which generates amplitude difference sequence reflecting the change of pulse wave amplitude and period difference sequence reflecting the change of pulse period interval in parallel based on the pulse wave peak value sequence, within a preset time window, this module analyzes the rhythm synchrony of the amplitude difference sequence and the period difference sequence, the rhythm synchrony is obtained by calculating the sign consistency ratio of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence, the sign consistency ratio represents the degree of sign consistency of the two sequences, in this experiment, the preset time window can be set to the duration containing the typical respiratory modulation period, so as to effectively capture the rhythm coupling relationship in the pulse wave, the calculation method of the sign consistency ratio can be to count the number of times that the product sign of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence is positive within the preset time window, and take the ratio of the number of times and the total number of heartbeats in the time window as the sign consistency ratio; a reference value adaptive establishment and event determination module, which compares the real-time obtained sign consistency ratio with the reference sign consistency ratio, the reference sign consistency ratio can be determined under the stable breathing state at the beginning of user monitoring, specifically, the reference sign consistency ratio can be adaptively established by calculating the moving average of the sign consistency ratio under the stable breathing state at the beginning of user monitoring, when the real-time sign consistency ratio continuously falls below the first preset threshold of the reference sign consistency ratio, it can be determined that the current signal segment is a rhythm desynchronization state signal segment, the first preset threshold can be set as the proportion of the reference value, in order to enhance the adaptability to different individuals, for example, the proportion can be selected between twenty percent and thirty percent, the setting logic is to reduce the false judgment probability of normal physiological fluctuations while sensitively identifying specific mode abnormalities of the signal;The artifact identification and body movement monitoring module continuously monitors the user's body movement through the acceleration sensor, and when the detected body movement amplitude is greater than the preset body movement threshold, the rhythm desynchronization state signal segment determination can be suspended to avoid the interference of body movement artifacts on the rhythm desynchronization state signal segment determination. In addition, at the rhythm desynchronization state signal segment occurrence time point, the second-order difference value of the amplitude difference sequence and the second-order difference value of the period difference sequence can be calculated. If either the absolute value of the second-order difference value of the amplitude difference sequence or the absolute value of the second-order difference value of the period difference sequence is greater than the preset morphological identification threshold, the event can be marked as a high-frequency interference segment, and its weight is reduced or excluded in the risk index calculation. The preset morphological identification threshold is set according to historical physiological data, and is used to distinguish the relatively flat changes caused by respiratory effort from the instantaneous sharp changes caused by non-respiratory source high-frequency signal disturbance. The central apnea determination module can continuously calculate the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence in the time period in which the rhythm desynchronization state signal segment is not determined to occur. The modulation energy is obtained by calculating the variance of the difference sequence in the time period. When the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence are both continuously lower than the respective energy baseline lower threshold within the preset continuous time length, a specific signal segment representing signal modulation energy silence can be identified. The energy baseline lower threshold can be set to the lower limit proportion value of the corresponding modulation energy in the user's normal stable sleep state, and the lower limit proportion value ranges from 0.1 to 0.3. The modulation energy baseline in the normal stable sleep state can be adaptively established by moving average of the variances of the amplitude difference sequence and the period difference sequence in the stable respiration state of the user in the initial monitoring stage.

[0056] The present experiment aims to verify the identification ability of the portable blood oxygen signal OSA intelligent detection method for invalid respiratory effort events, the artifact discrimination ability and the discrimination ability for different apnea types. The process is as follows: first, the photoplethysmogram signal is continuously collected by the optical plethysmograph sensor. The collected raw signal is processed by multi-scale smoothing to improve the signal stability. Then, the pulse wave peak of the photoplethysmogram signal is detected in real time by the strategy based on the comparison of neighborhood extreme value and amplitude threshold constraint, so as to obtain the pulse wave peak sequence. The stability of the sequence directly affects the accuracy of subsequent judgment. Based on the pulse wave peak sequence, the amplitude difference sequence and the period difference sequence are generated in parallel. The amplitude difference sequence is used to reflect the amplitude change between adjacent pulse wave peaks, and the period difference sequence is used to represent the fluctuation of time interval between consecutive heartbeats. In the preset time window, the number of positive values of the product of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence is calculated, and the ratio of the number to the total number of heartbeats in the time window is taken as the sign consistency ratio to obtain the rhythm synchronization. The ratio quantifies the consistency degree of the two difference sequences in the sign fluctuation direction. In the stable breathing state of the user monitoring starting stage, the baseline value is generated by moving average of the sign consistency ratio calculated in the stable period. In real-time detection, if the sign consistency ratio of the current period is continuously lower than the first preset threshold of the baseline sign consistency ratio, which can be set as a percentage of 20 to 30 percent of the baseline value, and the falling state lasts for more than a set minimum time length, then the current signal segment is determined as a rhythm desynchronization state signal segment. The preset threshold aims to enhance the adaptability to different individuals, and at the same time, to reduce the misjudgment probability of normal physiological fluctuations while sensitively identifying specific mode abnormalities of the signal. After preliminary determination of the rhythm desynchronization state signal segment, the second-order difference values of the amplitude difference sequence and the period difference sequence are calculated respectively. If the absolute value of any second-order difference value is greater than the preset morphological discrimination threshold, the event is marked as a high-frequency interference segment and excluded. The morphological discrimination threshold can be set according to historical physiological data to distinguish the relatively slow change phenomenon caused by respiratory effort from the sudden event of high-frequency signal disturbance caused by non-respiratory source. At the same time, the acceleration sensor continuously monitors the body movement of the user. When the body movement amplitude is greater than the preset body movement threshold, the determination of the rhythm desynchronization state signal segment is suspended to avoid the interference of body movement artifact on the detection result.In the period of time when the rhythm desynchronization state signal segment is not determined to occur, the modulation energy of the amplitude difference sequence and the period difference sequence is continuously calculated, and the modulation energy is obtained by calculating the variance of the difference sequence in the time period. When the two modulation energy values simultaneously and continuously fall below the lower threshold of the respective energy baseline within a set time length, it is determined that a specific signal segment representing signal modulation energy silence occurs. The lower threshold of the modulation energy can be set according to the corresponding modulation energy level of the user in the normal and stable sleep state, and usually ten percent to thirty percent of the level is taken as the proportional lower limit. At the same time, the energy baseline value can also be adaptively generated by moving average method in the initial monitoring period.

[0057] The experiment focuses on the performance of the method in identifying ineffective respiratory effort, identifying artifacts, and distinguishing different types of apnea. The symbol consistency ratio plays an important role in identifying ineffective respiratory effort. The experiment observed that when the user experienced an ineffective respiratory effort event, the symbol consistency ratio showed a significant and sustained downward trend. When the user was in a normal breathing state, there was a relatively stable rhythm synchronization between the amplitude difference and the period difference, and the symbol consistency ratio usually remained at a high level. When the rhythm desynchronization state signal segment occurred, the abnormal fluctuation of intrapulmonary negative pressure caused by airway obstruction led to rhythm decoupling in the circulatory dynamics and neural regulation mechanism, resulting in a significant decrease in the above ratio. The downward trend of this ratio, when an ineffective respiratory effort event occurs, its trough is usually earlier than the significant decrease in blood oxygen saturation, indicating that by capturing the pulse wave timing-amplitude decoupling phenomenon, the method can shift the monitoring focus from the traditional airway obstruction results (such as blood oxygen decrease) to the specific mode process of the signal itself, thus identifying early respiratory abnormalities when the patient has not yet shown significant physiological abnormalities; effective differentiation of cardiogenic artifacts and body motion artifacts, during the experiment, for scenarios that may introduce artifacts, the method shows its ability to identify, if the absolute value of the second-order difference exceeds the morphological identification threshold, the system can identify it as an artifact event caused by cardiogenic factors. This morphological identification threshold aims to distinguish the relatively slow-changing phenomenon caused by respiratory effort from the sudden and dramatic fluctuations caused by non-respiratory high-frequency signal disturbances. In addition, when the body motion amplitude exceeds the preset body motion threshold, the system can suspend the rhythm desynchronization state signal segment determination process, which effectively avoids the interference of body motion artifacts on the detection results, builds a collaborative barrier between signal-level rhythm analysis and physical behavior-level interference monitoring, and significantly enhances the practical adaptability of the method in daily life; differentiation between obstructive and central apnea, the method can distinguish specific signal segments representing signal modulation energy silence by monitoring the abnormal decay of difference sequence modulation energy. Within the time period without rhythm desynchronization state signal segment determination, when the amplitude difference sequence and the period difference sequence modulation energy are simultaneously and continuously below their respective energy baseline lower threshold for a preset duration, a specific signal segment representing signal modulation energy silence can be identified. This determination logic is based on the characteristics of central apnea, which leads to complete interruption of respiratory drive, resulting in a silent decline in signal. Therefore, low modulation energy is an important indicator of this type of event. This ability enables a single blood oxygen sensor to distinguish between obstructive and central apnea for the first time, providing a key classification basis for accurate treatment decisions.

[0058] Embodiment 4: The present embodiment combines Figures 1 to 3 A portable blood oxygen signal OSA intelligent detection method is implemented. As shown in Figure 1As shown, the three-axis acceleration data is output in real time by the accelerometer to the main processor for extraction and analysis of body motion amplitude value. When the body motion amplitude meets the condition of amplitude <0.5G, the main processor enters the alt state, and triggers the start of the breathing analysis. During this period, the system marks the current data segment as a valid signal mark, which is further processed by the event analysis module. If the amplitude is detected to be ≥0.5G, the system enters the high body motion state, and the system will perform a pause event judgment operation and synchronously cache the original data for subsequent recovery and use. After the body motion event ends, the system enters the recovery monitoring stage. If the user is detected to be stationary for 2 seconds and returns to the stable signal state, the main processor will perform a restart analysis and recovery mark, and resume the analysis and judgment of the breathing event. The body motion threshold of 0.5G and the pause delay of 200ms are explicitly set in the figure, which are used to trigger and control the boundary conditions of each processing logic. This flow intelligently pauses and resumes the breathing analysis process under body motion interference conditions, effectively avoiding the influence of false difference interference on rhythm desynchronization state signal segment judgment, and ensuring the accuracy and robustness of the detection system.

[0059] As shown in Figure 2 , Figure 2 (a) shows the change of rhythm synchronization rate with monitoring time (hh:mm) in the form of a line graph, where different event types are marked with state A, state B, and state C, reflecting that during the 00:00 to 03:00 period, the rhythm synchronization rate gradually decreased from the initial value close to 1.0, with a significant decrease between 01:30 and 02:30, and a rebound to about 0.9 at 03:00, reflecting the fluctuation characteristics of the synchronization between the amplitude difference sequence and the period difference sequence in the pulse wave, highlighting the changing trend of rhythm decoupling in the pre-event period of state B and state C. The baseline shown by the dashed line in the figure is used to determine the trigger threshold for synchronization reduction, and the gray background emphasizes the change range of the index, Figure 2 (b) is a column chart showing the distribution of amplitude modulation activity values in each sleep stage state under the same monitoring time. The annotations include wake period, N1 period, N2 period, N3 period, and REM period. From 00:00 to 03:00, the activity overall shows a downward trend, especially at the 02:30 (REM period) stage, which dropped to the lowest, supporting the correlation between respiratory-driven silence and state C pause.

[0060] As shown in Figure 3As shown, first, by continuously collecting and extracting the peak of the photoplethysmogram signal, the amplitude difference sequence and the period difference sequence are generated in parallel to construct the basic feature data for subsequent analysis. The system establishes a baseline consistency ratio under stable breathing conditions, forms an individualized baseline by moving average baseline, then calculates the symbol consistency ratio of the difference sequence (amplitude and period difference symbol matching degree) and compares it with the baseline value. If the real-time symbol consistency ratio is lower than the baseline threshold, it is determined that the rhythm is out of sync, and the system enters the abnormal analysis stage. To improve the accuracy of the determination and reduce the misjudgment, the system introduces the cardiogenic artifact exclusion, which includes two-order difference morphology analysis and body motion monitoring pause logic, to effectively identify non-breathing source high-frequency signal disturbance and body motion artifact interference. Under the premise of not triggering the obstructive event, the system also monitors the modulation activity. If the modulation energy is continuously lower than the baseline, a specific signal segment representing signal modulation energy silence is identified, thereby constructing a high-robustness respiratory event recognition process based on the rhythm synchronization, morphology change, and energy feature triple judgment mechanism, ensuring the accuracy and adaptability of the technical solution in the edge device environment. As shown in Figure 4 As shown, the trend of symbol consistency ratio during 24-hour monitoring is shown. The solid line represents the real-time symbol consistency ratio, the dotted line represents the baseline value, and the dashed line represents the determination threshold (70% of the baseline value). When the real-time ratio is continuously lower than the threshold, the system determines that the rhythm is out of sync; as shown in Figure 5 As shown, the modulation energy changes of the amplitude difference sequence and the period difference sequence under different breathing states are shown. The solid line represents the amplitude difference sequence modulation energy, the long dashed line represents the period difference sequence modulation energy, and the dotted line represents the energy baseline lower threshold; as shown in Figure 6 As shown, the relationship between body motion amplitude and breathing analysis state is shown. The upper part shows the body motion amplitude change (solid line), and the lower part shows the system analysis state (block marks represent normal analysis, and triangular marks represent pause analysis).

[0061] In the construction of the difference sequence, the system first extracts the pulse peak sequence based on the continuously collected photoplethysmogram signal using a pulse peak extraction method combining neighborhood local extremum judgment and amplitude threshold control, and generates the pulse wave peak sequence with a pseudo-peak suppression strategy. To enhance the stability of heart beat alignment, the system sets up a sampling frequency standardization module during peak extraction and introduces a numbering labeling mechanism to ensure sequence consistency in different heart rate intervals. On this basis, the system calculates the amplitude difference and period difference between adjacent peaks to form amplitude difference and period difference sequences reflecting the trend of blood oxygen signal changes, and ensures a one-to-one element relationship between the two difference sequences at each time node through numbering association. In terms of rhythm synchrony index generation, the system introduces a fixed-length sliding analysis window to calculate the rhythm coupling degree between the difference sequences, with a window length usually set to thirty to forty seconds, covering several complete respiratory cycles. This setting takes into account the timing characteristics of respiratory rhythm and the stability of algorithm response, effectively covering representative signal change patterns while effectively suppressing short-term abnormal interference. Within this time window, the system compares the signs of the products of corresponding elements in the amplitude difference and period difference sequences one by one, counts the number of positive products, and forms a ratio with the total number of heartbeats in the window. This ratio is the sign consistency ratio, which quantifies the consistency level of respiratory regulation rhythm in hemodynamic response.

[0062] To adapt to individual differences, the system selects the signal segment of the user in the resting state in the initial stage of wearing the device to construct the baseline value. In the specific process, the system adopts multi-window smoothing calculation to obtain the initial stage of the sign consistency ratio, and establishes the statistical expectation by eliminating abnormal values deviating from the average level, which is used as the individual baseline sign consistency ratio for subsequent event judgment. This method can realize baseline self-learning without relying on additional physiological parameters, and is suitable for the diversity of the basic state of users. In actual operation, the system continuously monitors the change of the current sign consistency ratio. If it is found that the ratio decreases by more than the set proportion of the baseline value in a continuous time period, and the duration reaches the system response threshold, the system determines that there is a respiratory rhythm decoupling phenomenon in this period, and further judges whether it constitutes a rhythm desynchronization state signal segment combined with other signal characteristics. The proportion change threshold is usually set in the range of twenty-five percent to thirty percent, which takes into account the rhythm fluctuation amplitude in normal sleep process and the need to distinguish abnormal states. To improve the accuracy of event recognition, the system also introduces a second-order difference calculation mechanism for the difference sequence. The amplitude difference sequence and the period difference sequence are respectively processed by the second-order difference, and the absolute value of the obtained value is calculated to evaluate the fluctuation intensity. When a suspected respiratory event is detected, the system compares the second-order difference intensity of the current time period with the morphological threshold value constructed in the resting state or normal rhythm state previously. If the difference exceeds the threshold range, it indicates that there is an abnormal sharp fluctuation in the signal, which may be caused by non-respiratory related reasons such as high-frequency signal disturbance of non-respiratory sources. The system determines that the event is a non-respiratory artifact event, and excludes it from the final event record. The setting principle of the morphological threshold value is based on the change characteristics of the typical heart beat waveform mutation, considers the influence law of common noise types on the blood oxygen signal, and maintains its dynamic adaptability through a smoothing update mechanism.

[0063] In the detection process of central apnea, the system evaluates the fluctuation intensity of the amplitude difference sequence and the period difference sequence, and calculates the signal variance of each sliding window. The calculated variance value is used to reflect the modulation activity of the blood oxygen signal in this time period. If the variance value is lower than the preset lower limit of the energy baseline in continuous time windows, it is considered that the modulation energy is in a silent state, and the system identifies a specific signal segment that represents the signal modulation energy silence. The lower limit of the energy baseline is usually based on the resting state samples obtained during the initial operation of the device. The threshold is set by calculating the average variance level and selecting a range of 10% to 30%. The selection of the range combines the signal noise level and the system false alarm tolerance, making it adaptable between different users. In terms of body motion interference judgment, the system collects the three-axis acceleration signal of the wearing part in real time, calculates the current body motion amplitude through vector synthesis, and constructs a dynamic interference recognition model combined with the judgment time length. When the body motion amplitude value continuously exceeds 0.5 gravity acceleration units and the duration exceeds 200 milliseconds, the system enters a high interference state and temporarily suspends the respiratory event judgment process. All signal data will enter the cache area. After that, if the system detects that the acceleration signal continuously returns to a stationary state and remains for more than two seconds, the event analysis process will be restarted, and the cache area data will be included in the subsequent processing. The above judgment condition uses a dual judgment mechanism of amplitude threshold and time threshold to ensure good anti-interference ability for night body position adjustment, slight shaking, and other non-breathing related factors, further ensuring the continuity and accuracy of the analysis. Through the timing coordination, data intercommunication, and logic linkage between the above modules, the system realizes the cooperative detection of obstructive and central sleep respiratory events under the condition of single blood oxygen signal input, ensuring its robustness and implementability in different use environments and user states.

[0064] Embodiment 6: This embodiment aims to elaborate the specific calculation process of the core index of the present application, the symbol consistency ratio. The whole calculation process can be divided into the following steps:

[0065] Step 1: Obtain the pulse wave peak value sequence. The pulse wave signal is continuously collected by the photoelectric volume (PPG) sensor, and its peak value is detected in real time, thereby obtaining a pulse wave peak value sequence, which can be formally represented as S peak : S peak = {P1, P2, P3, …, P k}, where P k = (A k , T k ) represents the information of the kth pulse wave peak point. A k : the peak amplitude of the kth pulse wave. T k : the time of the occurrence of the peak of the kth pulse wave.

[0066] Step 2: Generate amplitude difference sequence and period difference sequence, based on pulse wave peak sequence S peak The system generates two core difference sequences in parallel. The amplitude difference sequence (S...) dA This sequence reflects the variation in peak amplitude between consecutive pulse waves. Its i-th element dA i Defined as: dA i =A i+1 -A i Periodic difference sequence (S dP This sequence reflects the changes in the intervals of continuous heartbeat cycles. First, the pulse cycle sequence S needs to be calculated from the peak time series. P′ Its i-th element P′ i Defined as the time interval between two adjacent peaks: P′ i =T i+1 -T i Subsequently, the periodic difference sequence S dP The i-th element dP i Defined as the difference between two adjacent pulse cycles: dP i =P′ i+1 -P′ i =(T i+2 -T i+1 )-(T i+1 -T i By the above definition, the element dA of the amplitude difference sequence i The element dP of the periodic difference sequence i A one-to-one correspondence was established through the shared heartbeat index number i, laying the foundation for subsequent synchronization analysis.

[0067] Step 3: Calculate the symbol consistency ratio (R0). sync Within a preset time window (e.g., 30-40 seconds), the symbol consistency ratio R... sync The calculation and statistics of the elements dA of the amplitude difference sequence within this window are performed. i The corresponding element dP of the periodic difference sequence i The product sign is positive the number of times, and the ratio of this number to the total number of heartbeats within the window is taken as the final result. The calculation formula can be expressed as: Where N is the total number of heartbeats within the preset time window; i: is the index of the difference sequence, from 1 to N-2; CountIf(·): is a counting function, which counts the heartbeats when the condition within the parentheses (dA) is met. i ×dP i When the condition is true (>0), its value is 1; when the condition is false, its value is 0. The symbol consistency ratio R proposed in this invention... syncThe concept is derived from the deep insight into the specific mode physiological process of obstructive sleep apnea (OSA) signal, aiming to solve the fundamental limitation of the prior art which can only passively rely on the lagging obstruction results such as blood oxygen saturation reduction, and the core principle is that when the normal smooth breathing, the respiratory movement of the human body through the intrathoracic pressure change and autonomic nervous regulation, the amplitude and period of the photoplethysmogram (PPG) signal are synchronously modulated, which is manifested as the physiological coupling of the change trend of the two; and when the airway obstruction occurs in the OSA patient and the ineffective respiratory effort is carried out, the severe and irregular intrathoracic negative pressure fluctuation will destroy this regulation channel, resulting in a significant decoupling phenomenon of the amplitude and period change of the pulse wave; the formula is to quantitatively capture this transition from coupling to decoupling: by calculating the product sign of the amplitude difference sequence (dA i ) and the period difference sequence (dP i ), it is judged whether the change directions of the two are consistent, so as to identify the coupled heartbeats; the final R sync ratio, i.e. the proportion of coupled heartbeats to total heartbeats, constructs a new OSA detection dimension which is independent of blood oxygen saturation, and the formula can directly perceive the specific mode process of the core signal in the early stage of obstruction, thereby realizing effective screening of mild and moderate patients.

[0068] The physical meaning of the formula is to quantify the degree of synchronization of the amplitude change trend and the period change trend, and the higher the value of R sync , the better the synchronization, and the smoother the breathing; on the contrary, it indicates that the rhythm decoupling phenomenon occurs, which may indicate the occurrence of ineffective respiratory effort.

[0069] Step 4: In order to make the calculation process of the application more clear, a simplified numerical example is provided. It is assumed that in a time window, the system collects 6 consecutive pulse wave peaks, and the total number of heartbeats N = 6. Referring to Table 1, the original data (corresponding to step 1).

[0070] Table 1: Pulse wave peak value sequence original data example table.

[0071] Heartbeat index (k) Peak time T k (seconds) Peak amplitude A k (arbitrary units) 1 10.00 95 2 10.80 98 3 11.70 96 4 12.55 92 5 13.55 90 6 14.45 93

[0072] The intermediate sequence calculation (corresponding to step 2) calculates the pulse period P' i :

[0073] P'1 = T2-T1 = 0.80s;

[0074] P'2 = T3-T2 = 0.90s;

[0075] P'3 = T4-T3 = 0.85s;

[0076] P'4 = T5-T4 = 1.00s;

[0077] P'5 = T6 - T5 = 0.90 s;

[0078] The difference sequence dA is calculated i and dP i Table 2 of the following table shows the results of the calculation from index i = 1 to i = 4 (since dP4 requires P'5).

[0079] Table 2: Example table of the results of the difference sequence calculation.

[0080] Index (i) Amplitude difference dA i = A i+1 - A i ]]> Periodic difference dP i = P' i+1 - P' i ]]> 1 98-95＝+3 0.90-0.80＝+0.10 2 96-98＝-2 0.85-0.90＝-0.05 3 92-96＝-4 1.00-0.85＝+0.15 4 90-92＝-2 0.90-1.00＝-0.10

[0081] Table 3: Example table of the symbol consistency determination process.

[0082] Index (i) dA i ×dP i of the symbol CountIf(dA i ×dP i >0)]]> 1 (+3) x (+0.10) -> positive 1 2 (-2) x (-0.05) -> positive 1 3 (-4) x (+0.15) -> negative 0 4 (-2) x (-0.10) -> positive 1 Total 3

[0083] Final ratio calculation: According to the formula, the numerator is the sum of the number of times the product is positive (3 times), and the denominator is the total number of heartbeats N (6 times) in the window, The real-time symbol consistency ratio calculated in this way (0.5 in this example) will be sent to the event determination module and compared with the reference symbol consistency ratio established by the moving average. If this value continues to be lower than the reference threshold, the system determines that a rhythm desynchronization status signal segment has occurred.

[0084] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A platform for portable blood oxygen signal acquisition and analysis, characterized in that, The platform comprises: a PPG signal acquisition module, continuously acquiring a photoplethysmogram signal; a data preprocessing and feature extraction module, real-time detecting a pulse wave peak value of the photoplethysmogram signal to obtain a pulse wave peak value sequence; a difference sequence generation and rhythm synchronism analysis module, based on the pulse wave peak value sequence, generating an amplitude difference sequence reflecting the amplitude change of the pulse wave and a period difference sequence reflecting the period interval change of the pulse wave in parallel; and the difference sequence generation and rhythm synchronism analysis module further analyzes the rhythm synchronism of the amplitude difference sequence and the period difference sequence within a preset time window, the rhythm synchronism is obtained by calculating the sign consistency ratio of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence, and the sign consistency ratio represents the sign consistency degree of the two sequences; wherein, the calculation method of the sign consistency ratio is: within the preset time window, the number of times that the product sign of the elements of the amplitude difference sequence and the corresponding elements of the period difference sequence is positive is counted, and the ratio of the number of times to the total number of heartbeats in the time window is taken as the sign consistency ratio; a reference value adaptive establishment and event determination module compares the real-time sign consistency ratio with a reference sign consistency ratio, which is determined under a stable breathing state in the initial stage of user monitoring; and the reference value adaptive establishment and event determination module further includes that when the real-time sign consistency ratio continuously falls below a first preset threshold of the reference sign consistency ratio, the current signal segment is determined as a rhythm desynchronization state signal segment.

2. The platform for portable blood oxygen signal acquisition and analysis according to claim 1, wherein, In the reference value adaptive establishment and event determination module, the reference sign consistency ratio is adaptively established by calculating the moving average of the sign consistency ratio under a stable breathing state in the initial stage of user monitoring.

3. The platform for portable blood oxygen signal acquisition and analysis of claim 1, wherein, It further includes an artifact discrimination and body movement monitoring module, which at the rhythm desynchronization state signal segment occurrence time point, can calculate the second-order difference value of the amplitude difference sequence and the second-order difference value of the period difference sequence after the rhythm desynchronization state signal segment, and if either of the absolute value of the second-order difference value of the amplitude difference sequence or the absolute value of the second-order difference value of the period difference sequence is greater than a preset morphological discrimination threshold, the event is marked as a high-frequency interference segment, and its weight is reduced or excluded in the risk index calculation.

4. The platform for portable blood oxygen signal acquisition and analysis of claim 3, wherein, The preset morphological discrimination threshold is set according to historical physiological data, and is used to distinguish the relatively gentle change caused by respiratory effort from the instantaneous sharp change caused by high-frequency signal disturbance of non-respiratory source.

5. The platform for portable blood oxygen signal acquisition and analysis of claim 1, wherein, Also included is a central apnea determination module, which, in a time period in which the rhythm desynchronization state signal segment does not determine occurrence, can continuously calculate the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence, and the modulation energy is obtained by calculating the variance of the difference sequence in the time period; when the modulation energy E A of the amplitude difference sequence is greater than the upper threshold of the energy baseline of the amplitude difference sequence, and the modulation energy E T of the period difference sequence is greater than the upper threshold of the energy baseline of the period difference sequence, and at the same time, the modulation energy of the amplitude difference sequence and the modulation energy of the period difference sequence are continuously lower than the lower threshold of the respective energy baseline within the preset duration, a specific signal segment representing signal modulation energy silence is identified.

6. The platform for portable blood oxygen signal acquisition and analysis according to claim 5, wherein, The lower limit threshold of the energy baseline is set as a lower limit proportion value of the corresponding modulation energy in the normal and stable sleep state of the user, and the range of the lower limit proportion value is 0.1 to 0.

3.

7. The platform for portable blood oxygen signal acquisition and analysis of claim 5, wherein, The modulation energy baseline in the normal and stable sleep state is adaptively established by moving average of the variances of the amplitude difference sequence and the period difference sequence under a stable breathing state in the initial stage of user monitoring.

Images