Blood oxygen monitoring system fused with sleep rhythm modeling

Abstract

The invention relates to the technical field of medical diagnosis, and discloses a blood oxygen monitoring system fused with sleep rhythm modeling, which comprises the following steps: identifying a sleep stage conversion moment through a photoelectric volume pulse wave signal, detecting a blood oxygen saturation valley value in a window after conversion, and calculating response lag duration; according to the method, the physiological rhythm is converted into a natural probe for the compensatory ability of the respiratory system, dual diagnosis of sleep staging and respiratory function is realized through a single-path signal, and meanwhile, body movement interference is converted into an observation window for the respiratory stress recovery ability. The diagnosis reliability in a home monitoring scene is remarkably improved, the system completes core analysis at an edge node, and a respiratory function evaluation report which can be directly used for clinical decision making is output.

Description

Technical Field

[0001] The present invention relates to a blood oxygen monitoring system integrated with sleep rhythm modeling, belonging to the technical field of medical diagnosis. Background Art

[0002] In the field of home sleep breathing disorder diagnosis, the current mainstream technology relies on multi-sensor fusion strategies to collect blood oxygen saturation and sleep stage data, and uses cloud-based algorithms to perform static statistics on the separated parameters to calculate indicators such as the average blood oxygen value throughout the night and the proportion of REM periods.

[0003] However, this model has three systemic limitations: 1. The multi-sensor architecture leads to high equipment costs, and signal cross-interference forces the system to add complex filtering modules, which is fundamentally in conflict with the simple deployment requirements pursued in home scenarios; 2. When the user turns over at night and causes body movement interference, the existing solution directly discards the data during this period, which not only causes the loss of stress response information of respiratory compensation ability, but also reduces the effective data coverage due to frequent interruptions; 3. The analysis paradigm of independently processing sleep stage events and blood oxygen fluctuations completely ignores the dynamic coupling relationship between the two at physiological transition nodes, such as the sudden drop in throat muscle tension during the transition from deep sleep to REM stage, making mild to moderate respiratory compensatory function degradation difficult to detect.

[0004] Despite the recent emergence of edge computing optimization solutions that attempt to complete some signal processing locally, they are still unable to break through the mindset of single-dimensional data statistics. For example, although some improved devices have reduced cloud dependence, they can only output discrete event alarms and cannot capture the trend of progressive functional decline across sleep cycles. This collective neglect of the value of physiological timing associations makes it difficult for existing technologies to meet the core diagnostic needs of low-cost, full-time, and predictive home scenarios. Therefore, how to synchronously analyze sleep rhythm conversion events and blood oxygen response mechanisms through single-channel physiological signals, and convert body motion interference into compensation capacity assessment parameters in a high-interference environment, while achieving early warning of respiratory system functional decline trends, has become the technical problem to be solved by this invention. Summary of the Invention

[0005] The present invention provides a blood oxygen monitoring system integrated with sleep rhythm modeling, the main purpose of which is to solve the problem of how to synchronously realize sleep stage identification, dynamic evaluation of respiratory compensation ability and functional decline trend warning through a single-channel PPG signal.

[0006] To achieve the above objectives, the present invention provides a blood oxygen monitoring system integrated with sleep rhythm modeling, the system comprising: a signal sensing module configured to obtain a photoplethysmographic signal and a body motion signal of a monitored person; a rhythm transition recognition module configured to: identify the start time of the transition from non-rapid eye movement sleep to rapid eye movement sleep based on the amplitude envelope morphology of the photoplethysmography signal; wherein the rhythm transition recognition module identifies the start time by calculating the first-order difference of the amplitude envelope of the photoplethysmography signal or the sudden change of short-term energy; A response lag analysis module is configured to: determine the time when the valley value of blood oxygen saturation occurs in combination with the photoplethysmography signal within a predetermined time window after the start time, and calculate the response lag duration; The risk assessment decision module is configured as follows: assigning a credibility weight to each calculated response lag duration based on the body motion signal; accumulating the weighted response lag duration sequence for the entire night, and outputting the respiratory system function risk assessment result based on the aggregation pattern of response lag duration events that are higher than the risk threshold in the weighted response lag duration sequence; the risk assessment decision module evaluates the dynamic response of a single photoplethysmogram signal at a specific physiological rhythm transition point to reveal the respiratory system's instantaneous compensation ability to physiological disturbances.

[0007] Preferably, the risk assessment decision module is further configured to: when the body motion signal detects that the acceleration variance exceeds an acceleration variance threshold within a predetermined duration, mark the corresponding response lag time measurement value as unreliable and set the credibility weight to zero to discard the measurement value.

[0008] Preferably, the risk assessment decision module is further configured to: after identifying a period of body motion interference indicated by a body motion signal, determine a physiological disturbance index based on the acceleration variance of the body motion signal and the signal-to-noise ratio of the photoplethysmography signal during the period; generate a compensation coefficient based on the physiological disturbance index by referring to a preset mapping relationship table; and use the compensation coefficient to calibrate the first response lag duration measured immediately after the body motion interference period.

[0009] Preferably, the physiological disturbance index ( ) is calculated as: ,in, represents the acceleration variance of the body motion signal during the body motion interference period, It represents the signal-to-noise ratio of the photoplethysmography signal during the period of body motion interference.

[0010] Preferably, the risk assessment decision module is also configured to: divide the response lag duration sequence for the entire night into sleep cycle boundaries determined by the sleep cycle division algorithm; calculate at least one statistical characteristic value for each sleep cycle that characterizes the respiratory stability of the cycle; and determine the evolution trend of respiratory system function risk by comparing the statistical characteristic values ​​of adjacent sleep cycles to provide an early warning.

[0011] Preferably, the statistical characteristic value is the average or maximum value of the response lag time in each sleep cycle; the generation of trend warning is based on the cross-cycle deterioration index The judgment of whether the trend value is sustained or higher than a predetermined trend threshold is made. The inter-cycle deterioration index is calculated as the ratio of the statistical characteristic value of the current cycle to the statistical characteristic value of the previous cycle. The trend warning indicates that there is a risk of progressive decline in respiratory stability.

[0012] Preferably, the change in the amplitude envelope morphology of the photoplethysmography signal specifically refers to the amplitude of the photoplethysmography signal showing a form with higher irregularity and a decreased average amplitude during the rapid eye movement period relative to the non-rapid eye movement period.

[0013] Preferably, the decision logic adopted by the risk assessment decision module includes: comparing the response lag time with at least one critical threshold to generate an event risk level; and within a single sleep cycle, if the frequency of occurrence of the event risk level higher than the predetermined risk level exceeds a predetermined frequency threshold, then increasing the respiratory system function risk assessment level.

[0014] Preferably, the rhythm transition recognition module identifies the starting moment of the transition from non-rapid eye movement (NREM) to rapid eye movement (REM) sleep state by analyzing the morphological characteristics of the single-channel photoplethysmography signal at the wrist.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By capturing the critical physiological transition from non-rapid eye movement (NREM) to rapid eye movement (REM) sleep, the system transforms traditional absolute blood oxygen monitoring into a dynamic assessment of the respiratory system's compensatory capacity. At the moment of a sudden drop in throat muscle tension, the blood oxygen response in healthy individuals is rapid and stable, while that in individuals with potential airway collapse exhibits a discernible delay. This mechanism, which converts physiological rhythm phase changes into endogenous probes, enables a single PPG signal to carry both sleep staging and respiratory function diagnostic information. This avoids the complex deployment of EEG sensors while revealing the transient compensation characteristics overlooked by traditional polysomnography.

[0016] 2. The system couples acceleration signal characteristics with the PPG signal disorder to generate a physiological disturbance index. When the user rolls over and triggers the measurement blind spot of traditional solutions, the index dynamically calibrates the first subsequent valid measurement value through a preset mapping relationship. This design transforms the body movement period from a data waste zone into an observation window for the respiratory system's stress recovery ability. In high-interference scenarios in the home environment, it maintains the reliability of core indicators and significantly improves the utilization rate of effective data. By dividing the compensatory capacity indicator sequence according to sleep cycle, it establishes cross-temporal correlations between the characteristic values ​​of adjacent cycles, such as the ratio of the lag duration between deep sleep and REM sleep. The system extracts macro-trend information from micro-events. When a sustained increase in the cross-cycle deterioration index is detected, even if a single measurement does not reach the risk threshold, it can still trigger an early warning of progressive functional decline. This longitudinal comparison mechanism based on the physiological rhythm framework provides a time window for intervention in chronic respiratory diseases that is not possible with traditional single-point warnings.

[0017] 3. The system completes core operations such as rhythm conversion recognition, response lag analysis, and body movement correlation calibration at the signal perception node, and only outputs weighted time series event reports to the terminal. This marginal transformation of raw signals to diagnostic conclusions not only avoids the impact of cloud transmission delays on physiological event capture, but also enables the device to maintain full functionality in a low-bandwidth environment. The weighted lag duration series and cross-cycle trend indicators generated by all-night monitoring directly correspond to the core parameters of compensatory capacity and functional evolution that are of clinical concern to respiratory medicine. The output report does not require doctors to parse raw waveforms or complex data, and can support the initial screening and triage of obstructive sleep apnea, greatly reducing the diagnostic threshold of primary medical institutions and alleviating the resource burden of sleep centers in tertiary hospitals. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a timing diagram of the blood oxygen monitoring response hysteresis evaluation triggered by sleep rhythm conversion of the present invention; Figure 2 This is a comparison diagram of the dynamic difference of blood oxygen response hysteresis in the present invention; Figure 3 Schematic diagram of the sleep rhythm conversion identification method based on amplitude envelope difference of the present invention; Figure 4 This is an architecture diagram of the respiratory function risk analysis system that integrates rhythm recognition and body movement assessment in the present invention.

[0019] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below. Obviously, the described embodiments are only part of the present invention and not all of it. The present invention provides a blood oxygen monitoring system that integrates sleep rhythm modeling. Its overall operation process starts with sensor data acquisition, runs through real-time analysis and dynamic evaluation of signals, and finally outputs a structured risk report. The system is logically composed of a signal perception module, a rhythm conversion recognition module, a response lag analysis module, and a risk assessment decision module. In a typical application scenario, the user wears a wrist device that integrates a single photoelectric volumetric pulse wave sensor and a single three-axis acceleration sensor when sleeping at night. The signal perception module is responsible for obtaining the raw data stream from these two sensors. Specifically, the photoelectric volumetric pulse wave signal is collected at a frequency of about 50Hz. This sampling rate is sufficient to capture the fine morphology of the pulse wave, while the body motion signal is collected at a frequency of about 25Hz to accurately record the user's posture changes and limb movements during sleep; the obtained raw signal The signal is first pre-processed by a digital bandpass filter to effectively filter out baseline drift caused by activities such as breathing and high-frequency noise introduced by the circuit itself, thereby providing a high-quality data basis for subsequent feature extraction and analysis. The processed photoplethysmography signal is sent to the rhythm conversion recognition module. The core task of this module is to act as a sentinel of physiological state and accurately identify the starting moment of the transition from non-rapid eye movement to rapid eye movement sleep. In view of the fact that the regulation of the autonomic nervous system will change significantly during the transition from non-rapid eye movement to rapid eye movement, resulting in changes in peripheral vascular tension, the amplitude envelope of the photoplethysmography signal will show a unique morphological feature, that is, compared with the non-rapid eye movement period, the signal amplitude of the rapid eye movement period will show higher irregularity and the average amplitude will show a downward trend, in order to accurately capture the starting moment of this conversion. The rhythm transition recognition module first extracts the amplitude envelope of the photoplethysmography signal in real time through Hilbert transform or peak detection and envelope fitting technology. Then, the module calculates the first-order difference of the amplitude envelope sequence. When the differential value has a sudden spike exceeding the preset negative threshold, or by calculating the short-term energy of the signal, when the energy value has a significant step-like drop, the system marks the moment as the starting moment of the transition. ,This recognition process is entirely based on the morphological features of the single-channel ,photoplethysmography signal at the wrist, without the need of introducing ,additional EEG or eye movement sensors;

[0021] Once the rhythm transition recognition module identifies the start time , the response lag analysis module is activated, and its function is to quantify the dynamic response ability of the respiratory system to this endogenous physiological challenge. Physiologically, when switching from non-rapid eye movement to rapid eye movement, the tension of the throat muscles will naturally relax. For individuals with healthy respiratory function, their respiratory system can quickly compensate, and the blood oxygen saturation remains stable or has only small, short-term fluctuations. However, for individuals at risk of airway collapse, this decrease in muscle tension will induce or aggravate respiratory resistance, resulting in a significant and delayed decrease in blood oxygen saturation. To quantify this process, the response lag analysis module is activated at the starting moment After that, an observation window of 90 to 150 seconds is immediately opened. During this window, the system uses the red light and infrared light absorption ratio in the photoplethysmography signal to continuously calculate the instantaneous blood oxygen saturation value and find the valley value in this blood oxygen saturation sequence in real time. When the valley value is determined, the time point of its occurrence is recorded as Finally, the system calculates the response lag time This duration directly reflects the time delay between the transition from physiological state to the occurrence of significant stress response in the respiratory system, and becomes the core indicator for evaluating the compensatory capacity of the respiratory system.

[0022] At the same time, the risk assessment decision module processes the body motion signals from the acceleration sensor in parallel, and integrates the processing results with the output of the response lag analysis module to ensure the reliability of the assessment results in a high-interference environment at home. The module implements a set of sophisticated, situation-based decision logic: First, in order to deal with the serious pollution of the photoelectric volume pulse wave signal caused by violent body movements such as users turning over, the module sets an acceleration variance threshold. If the acceleration variance of the body motion signal exceeds the threshold within a predetermined duration within the corresponding time window for calculating a certain response lag time, it is considered that the measurement has been subjected to irreversible interference, and the measured response lag time value is then It is marked as untrustworthy, and its corresponding credibility weight is directly set to zero, so that it is discarded in the subsequent cumulative analysis. Secondly, the system can not only effectively eliminate the contaminated data, but also transform the period of body motion interference into an observation opportunity for the stress recovery ability of the respiratory system. Specifically, when a period of body motion interference indicated by the body motion signal is identified, the module will calculate the acceleration variance of the body motion signal within the period based on the acceleration variance of the body motion signal. and the signal-to-noise ratio of the photoplethysmography signal , calculate a physiological perturbation index This index comprehensively quantifies the intensity of body motion interference and its actual impact on physiological signals. Then, the system calculates The device generates a compensation coefficient by referring to a mapping table preset in the device memory, and uses the compensation coefficient to calibrate the first response lag duration measured immediately after the body motion interference period. This transforms the body motion interference from a purely negative factor into an effective information window for evaluating the subsequent recovery speed of respiratory stability.

[0023] After obtaining the whole night's response lag duration sequence that has been credibility-weighted and calibrated with body motion, the risk assessment decision module will enter a higher-dimensional analysis stage to reveal the clustering pattern and evolution trend of risks. The module first divides the whole night's response lag duration sequence into several subsequences corresponding to the sleep cycles based on the sleep cycle boundaries determined by the sleep cycle division algorithm. For each sleep cycle, the module will calculate at least one statistical characteristic value that can characterize the respiratory stability within the cycle, such as the average or maximum value of all valid response lag durations within the cycle. Furthermore, in order to provide early warning, the module will determine the evolution trend of respiratory system function risk by comparing the statistical characteristic values ​​of adjacent sleep cycles. This is achieved by calculating a cross-cycle deterioration index. This is achieved by defining the index as the ratio of the current period’s statistical characteristic value to the previous period’s statistical characteristic value. When the trend value is continuously or significantly higher than a predetermined trend threshold, the system will generate a trend warning, indicating that the user's respiratory stability may be at risk of progressive decline. In addition, within a single sleep cycle, the module will also compare the response lag time with at least one critical threshold to generate event risk levels such as low risk, medium risk, and high risk. If the frequency of event risk levels exceeding the predetermined risk level in a single sleep cycle exceeds a predetermined frequency threshold, the system will automatically increase the overall assessment level of the user's current respiratory function risk. Finally, the system integrates all analysis results and directly generates a respiratory function assessment report on the edge computing node, that is, the wrist device worn by the user, which does not require secondary interpretation by professionals. The report includes the final risk assessment level, potential trend warning information, and key quantitative indicators such as the weighted average response lag time throughout the night, thereby providing users or primary medical institutions with a reference basis that can be directly used for clinical decision-making, significantly reducing the threshold and complexity of professional-level sleep respiratory health monitoring in home scenarios.

[0024] And, the preset mapping relationship table is solidified in the device as a set of segmented functions, which converts the calculated physiological disturbance index The value is directly mapped to the compensation coefficient of the response lag time. In a specific embodiment, the functional relationship is as follows: When the value is lower than the preset resting baseline value of 1.5, the compensation coefficient is 1.0, and when When the value is between 1.5 and 7.0, the compensation coefficient The value increases along the formula Coefficient=1.0+0.09*( -1.5) increases linearly, and when When the value exceeds 7.0, the compensation coefficient is set to an upper limit of 1.495. This functional relationship is obtained by performing segmented regression analysis modeling on more than 1,000 valid body movement events and subsequent blood oxygen response data collected from at least 50 subjects with different physical signs under synchronous monitoring of a polysomnography system. A one-time individualized benchmark calibration procedure is performed when it is first activated. The procedure guides the user to continuously collect physiological signals for three minutes while lying still without significant body movement. The system uses the second-by-second variance data of the three-axis acceleration signal within these three minutes, takes the 95th percentile value of its probability distribution, and sets it as the acceleration variance threshold for the user to subsequently judge severe body movement interference. At the same time, the system calculates the mean μ and standard deviation σ of the first-order difference sequence of the amplitude envelope of the photoelectric volume pulse wave signal during this period, and sets the individualized rhythm conversion recognition threshold to μ minus 3.5 times σ. The inter-cycle deterioration index The trend threshold is uniformly set to 1.05, which represents that the rate of change of the core statistical characteristic values ​​of adjacent sleep cycles exceeds the upper limit of the normal physiological fluctuation range observed in a large physiological database. This is an extended implementation method known to ordinary technicians in this field.

[0025] Example 1: The technical solution of the present invention operates in a specific application scenario as follows: A user has no clear signs of respiratory disease in a routine physical examination, but has been troubled by daytime sleepiness and occasional waking up at night for a long time. The average blood oxygen saturation reading of a traditional home oximeter is always within the normal range and fails to provide effective diagnostic clues. The user wears a wrist device integrated with the system of the present invention to conduct home sleep monitoring for seven consecutive days, in order to reveal the true dynamic response characteristics of his respiratory function under physiological rhythm conversion. On the first night of monitoring, after the system is started, the signal perception module begins to synchronously collect the user's photoelectric volume pulse wave signal and body movement signal, and sends the data stream to the subsequent processing unit. After entering the first sleep cycle, when the user's sleep state converts from non-rapid eye movement to rapid eye movement, the rhythm conversion recognition module analyzes the negative mutation of the first-order difference of the amplitude envelope of the photoelectric volume pulse wave signal to identify the conversion starting time. Mark it down, and the response lag analysis module will be Then, a 120-second observation window was opened. During this window, the instantaneous blood oxygen saturation sequence was continuously calculated by analyzing the photoplethysmography signal. At the 58th second, a blood oxygen trough value was captured, and its occurrence time was recorded as , the system calculates the first response delay time based on this , which is 58 seconds. Here, the precise time mark of the rhythm transition recognition module provides an unbiased timing starting point for the response lag analysis module, making The measurement can truly reflect the compensatory delay of the respiratory system triggered by the change of endogenous throat muscle tension, rather than random blood oxygen fluctuations, thereby transforming an isolated physiological signal into a diagnostic carrier that carries dual information of sleep staging and respiratory function. In the middle of the night, the user has a more violent turning movement. At this time, the system faces the technical dilemma of maintaining data integrity and reliability in a high-interference environment. The traditional solution will directly discard this segment of data because the acceleration variance of the body motion signal exceeds the threshold, resulting in an information blind spot. The system of the present invention regards this body motion interference period as an observation window for the stress recovery ability of the respiratory system. The module first calculates the acceleration variance of the body motion signal within this period based on the acceleration variance of the body motion signal. and the signal-to-noise ratio of the photoplethysmography signal , calculate a physiological perturbation index The value of this indicator is generated by referring to a mapping relationship table preset in the device memory to generate a compensation coefficient. This mapping relationship table is established during the product development stage by performing regression analysis on reference polysomnographic data under different intensities of body motion interference, ensuring the physiological significance and accuracy of the compensation coefficient. For the first response lag time measured immediately after the end of the body motion interference period, the system automatically applies the compensation coefficient for calibration. In this way, the system not only does not lose data, but instead converts an interference event into a quantitative assessment of the time required for the user's respiratory control system to return to stability after being subjected to external physical disturbances, resolving the inherent contradiction between data discarding and measurement inaccuracy. At the end of the seven-day monitoring period, the system does not calibrate all responses. Instead of static statistics of values, the analysis framework is reshaped, shifting from focusing on isolated respiratory event risks to evaluating the long-term evolution trend of respiratory function stability. The risk assessment decision module first divides the weighted response lag duration sequence of each night into independent sleep cycle subsequences based on the sleep cycle division algorithm, and calculates the maximum value of the response lag duration in each cycle as the statistical characteristic value of the cycle. Subsequently, the module calculates the cross-cycle deterioration index between adjacent sleep cycles. , that is, the ratio of the current period statistical characteristic value to the previous period statistical characteristic value. The monitoring data shows that the user The value in the sleep cycle in the second half of the night showed a pattern of being continuously higher than the preset trend threshold for three consecutive nights. Although the single response lag time never reached the critical threshold of severe risk, this longitudinal comparison based on the physiological rhythm framework revealed the hidden risk of progressive decline in its respiratory compensation ability during the deepening of sleep. Finally, in addition to the conventional quantitative indicators, the system output a special warning in the respiratory function assessment report generated by the edge node, indicating that its respiratory stability has a potential progressive attenuation trend, providing a decisive time window for the user to intervene early before obvious clinical signs appear.

[0026] Example 2: In order to objectively verify the effectiveness and engineering feasibility of the technical solution of the present invention in a simulated real home high-interference environment, especially to verify the core function of the system in converting body motion interference into a respiratory stress recovery ability observation window, and the accuracy of early warning of functional decline based on the cross-cycle deterioration index, this embodiment is specially designed and implemented. In a typical application scenario, that is, for a test subject who frequently turns over during sleep at night, traditional monitoring equipment will discard a large amount of data caused by body motion and cannot form a continuous and complete respiratory stability assessment, thereby possibly missing key Signs of worsening respiratory function. This embodiment aims to verify how the system of the present invention can not only maintain monitoring continuity in this challenging scenario, but also utilize the challenge itself to extract deeper physiological information. The experimental platform was built in a standard sleep laboratory. A volunteer who was initially screened to meet the characteristics of mild obstructive sleep apnea syndrome and reported frequent body movement events in his sleep was used as the test subject. The polysomnography system (PSG) was used as the reference gold standard. However, the system of the present invention operated independently, using only its wrist device, which integrated a single photoplethysmography sensor and a single three-axis accelerometer.

[0027] First, the acceleration variance threshold used to identify unreliable measurements is set to accurately distinguish between benign minor posture adjustments and violent body movements that are sufficient to contaminate the quality of the photoplethysmography signal. The core technical trade-off of this setting is to avoid misjudging valid data as interference and discarding it, while preventing the accuracy of the overall assessment from being reduced due to the inclusion of severely interfered data. To this end, a set of deterministic calibration procedures were implemented before the formal test: the test subjects were required to wear the device and complete three standard actions in sequence according to the instructions: lying still, micro-movement of the wrist, and turning over in bed, while simultaneously recording the acceleration signal. By calculating the variance of the acceleration signal for each action and using the decile of the variance distribution for turning over in bed as the initial threshold, an objective and individualized benchmark was anchored for the selection of this parameter. Secondly, for the core component in the risk assessment decision module, namely the physiological disturbance index The preset mapping relationship table between the compensation coefficient and the acceleration is not based on empirical inference, but is derived from a pilot study. The essence of this study is to establish a quantitative model between the intensity of physiological disturbances and the calibration requirements of respiratory response delay. In this study, multiple test subjects were recruited and, under PSG monitoring, their body movements were actively induced by external physical stimulation, and the acceleration variance measured by the system of the present invention was simultaneously recorded. , photoplethysmography signal-to-noise ratio , and the blood oxygen recovery time after the apnea or hypopnea event caused by the disturbance, which was accurately captured by PSG, were established through regression analysis. The mathematical relationship between the delay in blood oxygen recovery and the actual blood oxygen recovery delay was established, thereby generating the mapping relationship table, which ensured the physiological basis for subsequent compensation calibration. During the experiment, when the system was running stably in the first half of the night, it accurately identified the starting moment of the transition from non-rapid eye movement to rapid eye movement several times and calculated the response lag time. , its baseline value is stable within a certain range, indicating the basic respiratory compensation ability of the test subject in the absence of significant disturbance. At 2:16 am, the acceleration sensor detected a violent body movement that lasted longer than the preset value, and its acceleration variance significantly exceeded the preset acceleration variance threshold. At this time, the conventional algorithm deployed on traditional equipment will directly mark the data in this period and the subsequent time window as invalid and discarded, resulting in a breakpoint in the evaluation. However, the risk assessment decision module of the present invention is triggered by this body movement event and enters its unique processing flow. The module does not discard this event information, but immediately calculates the acceleration variance based on the acceleration variance during the body movement interference period. and the signal-to-noise ratio of the photoplethysmography signal , calculate a physiological disturbance index value, according to the The value is consulted in the preset mapping relationship table to obtain a specific compensation coefficient, and this coefficient is applied to calibrate the first response lag time measured immediately after the body motion interference period. The following embedded table presents representative data points before and after this key node and during another sleep cycle, see Table 1: Table 1, Experimental data monitoring and calibration table.

[0028]

[0029] From the above data, it can be seen that at point B after body motion interference, the uncalibrated Compared with the baseline value at point A, there is a false impression of improvement, which is illogical from a physiological point of view, because physical stress usually increases the burden on the respiratory system. However, the present invention introduces a compensatory mechanism to The calibration is 21.0 seconds, which is not only higher than the baseline, but also more accurately reflects the instantaneous negative impact of the physiological disturbance on the respiratory compensation ability. In other words, the system will The sequence is divided into sleep cycles and the statistical characteristics of each cycle are calculated, in this case the average , by calculating the inter-cycle deterioration index of adjacent cycles , that is, the average of the current period Average of the previous period The ratio of the sleep cycle from the second to the fourth is found. The values ​​were 1.04, 1.08, and 1.11, respectively, showing a unidirectional increasing trend that was continuously higher than the preset trend threshold of 1.05, which caused the system to generate a trend warning in the final report.

[0030] Example 3: This example combines Figures 1 to 4 , this paper describes the implementation of a blood oxygen monitoring system that integrates sleep rhythm modeling. Figure 1 As shown in the figure, the process starts with the rhythm conversion trigger module detecting the conversion starting point t0 as the starting event, and then the observation window control module sets the observation window [t0, t0+150s], and determines the window length to be 90~150 seconds through statistical optimization based on physiological modeling. The blood oxygen calculation module starts the blood oxygen monitoring process during this window period. In the iterative calculation cycle within the window time, the system obtains the red light / infrared light ratio and calculates the instantaneous value, and The sequence is transmitted to the valley detection module, which updates the lowest value record in real time. After the valley value is detected, the valley time tv is recorded and the physiological process is annotated: a sudden drop in pharyngeal muscle tension → a decrease in ventilation → a decrease in blood oxygen, which clearly explains the physiological mechanism of the decrease in blood oxygen. When the observation window ends, the system sends a window end signal to the valley detection module, which transmits the valley time tv to the lag calculation module. The module calculates the response lag time from the conversion starting point to the occurrence of the blood oxygen valley value according to the formula of lag time = tv-t0, and then calculates the output value of the lag time. Finally, the output result module performs functional interpretation based on the lag time value, which is 4-6 seconds for healthy people and 15-20 seconds for risk people. This value can reflect the individual's compensatory ability.

[0031] like Figure 2 As shown in the figure, the horizontal axis is time (seconds) and the vertical axis is arterial oxygen saturation (%), which depicts the time evolution of blood oxygen response in the window period after conversion, in which healthy individuals are represented by solid lines and risk individuals are represented by dotted lines. After the transition moment, the healthy individual The curve remained basically stable, with only a slight decline and then a rapid rebound, while the risk of individuals The curve shows a significant delay drop and gradually recovers after reaching the lowest point at tv (valley moment), which is clearly marked in the figure. = 58 seconds, indicating that the risk individual The response lag time between tv (transition time) and tv (valley time) is used to quantify the time difference of the respiratory system's compensatory response to the physiological stress caused by sleep rhythm transition.

[0032] like Figure 3 As shown, the original PPG signal in the upper part shows high amplitude regularity in the NREM period, and turns into a low amplitude irregular state in the REM period. The marked area transition represents the transition process of the sleep state from non-rapid eye movement period (NREM period) to rapid eye movement period (REM period). In the amplitude envelope diagram in the middle, There is a significant drop at , which indicates a sudden change in the morphology of the amplitude envelope. The first-order difference curve of the amplitude envelope at the bottom further reveals the numerical change of the sudden change. A mutation point appears at the corresponding position, which is lower than the threshold, thus meeting the discrimination conditions of the rhythm transition recognition module and completing the accurate recognition of the transition start moment.

[0033] like Figure 4 As shown in the figure, the top layer is the sensor input, including the photoelectric volume pulse wave (PPG) sensor and the three-axis acceleration sensor, which are used to obtain the photoelectric signal and body motion signal of the user during sleep. Below it is the signal perception module, which is responsible for collecting PPG signals (≈50Hz) and body motion signals (≈25Hz) and transmitting the signals to the subsequent processing module. The rhythm conversion recognition module performs the recognition of the NREM→REM conversion moment based on the collected signals, and accurately determines the starting point of the rhythm change by analyzing the PPG amplitude envelope shape and calculating the first-order difference / short-time energy mutation. After identifying the conversion point, the system starts the response lag analysis module and the body motion intervention module in parallel. The system then enters the risk assessment and decision-making module, where the system assigns a credibility weight to the response lag duration, divides the data according to the sleep cycle, further calculates the cross-cycle deterioration index (CDI), identifies risk events and evaluates patterns, and finally enters the system output stage to output a respiratory system function risk assessment report, which covers risk levels, trend warnings, and key indicator summaries.

[0034] Example 4: In a specific application scenario, a new user activates a wrist device integrated with the system of the present invention for the first time. After the user wears the device, the system does not immediately enter a passive background monitoring mode. Instead, the system actively guides the user to perform a one-time, approximately five-minute system initialization and personalized calibration procedure. The underlying technical logic of this procedure is that any universal empirical threshold may introduce judgment bias when facing individual physiological differences. Only a benchmark established based on the user's own physiological data under controllable conditions can ensure the accuracy of subsequent assessments. The procedure first instructs the user to keep their wrist still for one minute while sitting still. During this period, the signal sensing module continuously collects photoplethysmography signals and triaxial acceleration signals at its standard sampling rate. The system algorithm calculates the second-by-second variance of the acceleration signal for this minute and determines the 98th percentile value of the variance sequence as the user's personalized acceleration variance threshold. This ensures that the threshold used to subsequently identify severe body motion interference is based on the user's own resting noise baseline, rather than a universal, fixed value that may be overly sensitive or insensitive. Then, the procedure enters the second stage, in which the user is guided by the device prompts to perform a two-minute rhythmic deep breathing training, that is, deep inhalation and exhalation at a frequency of six times per minute. During this controlled breathing regulation process without external physiological load, the system can still identify the slight changes in vascular tension caused by the fine-tuning of the autonomic nervous system, which is similar to the transition from non-rapid eye movement to rapid eye movement, and calculate a series of baseline values ​​of response lag time based on this. The system forms a distribution sequence of all the response lag time values ​​measured within the 120 seconds and calculates The system uses its mean and standard deviation to establish the mean as the user's individualized health response benchmark, and the sum of the benchmark value plus two times the standard deviation is set by the system as the critical threshold for the user's individual low-risk events and medium-risk events, thereby firmly anchoring the risk grading scale on the user's own physiological characteristics. During the continuous monitoring process throughout the night, the sleep cycle division algorithm within the system operates according to a deterministic temporal clustering logic, which converts all the non-rapid eye movement periods marked by the rhythm conversion recognition module into the starting time of the rapid eye movement period. As a time point sequence, the algorithm traverses the sequence and calculates any two adjacent starting moments and The time interval between ,like If the sleep cycle is between 80 and 130 minutes, which is a typical physiologically recognized sleep cycle, the system will arrive The entire time period is confirmed as a complete sleep cycle. This method of dividing the cycle based on the physiological rhythm anchor point can more accurately correspond the respiratory stability analysis to the actual sleep structure compared to the traditional segmentation method that relies on an external clock or a simple body movement signal. Furthermore, for the preset mapping relationship table embedded in the risk assessment decision module and used to calibrate the response lag time after body movement interference, its construction process follows a rigorous offline modeling procedure. During the product development stage, a group of test subjects with different physical signs are recruited. Under the synchronous monitoring of a professional polysomnography monitoring system, they are induced to produce body movements of different intensities through controllable external physical stimulation. For each body movement event, the system synchronously records the physiological disturbance index calculated by the device of the present invention. , and the respiratory disorder event induced by the body movement and the actual blood oxygen recovery time thereafter, which was accurately measured by the polysomnography system. The value and blood oxygen recovery time are subjected to piecewise linear regression analysis, and finally a curve is fitted that can accurately reflect the quantitative relationship between the intensity of physiological disturbance and the required compensatory recovery time. After the curve is discretized, it is solidified into the mapping relationship table in the device memory.

[0035] After completing the full-night monitoring, the system executes a hierarchical risk aggregation logic to generate the final assessment report. First, for each divided sleep cycle, the system calculates two core indicators based on all weighted and calibrated response lag time data within the cycle: the first is the risk event frequency, that is, the number of events in which the response lag time value exceeds the individual critical threshold; the second is the cross-cycle deterioration index Then, the system calculates a comprehensive risk score for each sleep cycle. This score is calculated by the weighted sum of the frequency of risk events and the cross-cycle deterioration index. The weight coefficient used is also the optimal value determined by regression analysis in the aforementioned offline modeling procedure. Finally, the system calculates the total risk index for the whole night. This index is the weighted average of the risk scores of all sleep cycles. Among them, the weight of the sleep cycle in the second half of the night is moderately increased to reflect the physiological characteristics of the rapid eye movement period dominating in the second half of the night. The total risk index is finally mapped to a three-level risk level system consisting of low risk, trend warning, and high risk, and combined with The continuous changing trend of the respiratory system jointly determines the output content of the final report. In this way, through the above series of interlocking deterministic procedures, the system of the present invention transforms a nighttime monitoring into an individualized, calibrated, full-time, multi-dimensional in-depth analysis of the user's respiratory system compensatory capacity. The final output assessment report not only reveals the current risk status, but also indicates its evolution direction within the framework of physiological rhythms, achieving the established technical goal of transforming body motion interference into an effective observation window and providing an early warning for chronic respiratory function decline.

[0036] Example 5: In the internal parameter configuration of the system, a preset mapping relationship table is statically stored in the non-volatile storage unit of the wrist device. The establishment of the mapping relationship table is based on a set of offline modeling procedures in the research and development stage of the present invention. The process is centered on the logical association between the physiological disturbance index PDI and the dynamic delay characteristics of the respiratory system's compensatory ability. Specifically, in the procedure, the system first selects a plurality of test subjects with significant differences in physical signs, and applies controllable body motion stimulation to induce real physiological disturbance events under the synchronous recording of the polysomnography system. At the same time, the acceleration variance of the acceleration signal and the signal-to-noise ratio of the photoelectric volume pulse wave signal are obtained in real time by the system of the present invention. On this basis, the corresponding physiological disturbance index PDI is calculated. Subsequently, the system performs multiple rounds of matching with the blood oxygen recovery time revealed in the polysomnography data, with PDI as the independent variable and the blood oxygen lag calibration requirement as the dependent variable. Segmented regression modeling is performed, and the regression results are smoothed and discretized, ultimately forming the mapping structure for lookup table compensation. The compensation coefficient is called on demand during system operation, does not rely on external computing resources, and its value is matched and found based on the real-time PDI calculation results, which can achieve rapid calibration of the first response lag time under different disturbance intensities. It should also be noted that although the mapping relationship table is built-in in a general form at the factory, the system also supports local reconstruction and dynamic optimization through an individualized calibration process when a specific user is initialized. The specific method is: during the first use stage, the user completes the set action sequence to induce representative body movement events. The system uses this as an anchor point to resample the PDI value of the corresponding time period, and automatically adjusts the calibration coefficient of the corresponding section in the mapping table based on the lag performance, so that the overall compensation curve covers the common interference range while being more in line with the individual's actual physiological response characteristics.

[0037] Furthermore, the trend threshold used in the inter-cycle deterioration index (CDI) is not determined by empirical judgment, but rather by a trade-off mechanism guided by engineering logic. The core of this threshold setting lies in the technical balance between the system's early warning capability and false alarm tolerance. If the threshold is set too low, the system will tend to be overly sensitive, easily misjudging natural physiological fluctuations as abnormal trends, reducing the specificity of the assessment. Conversely, if the threshold is set too high, it will suppress the response to early deterioration signals and may miss the intervention window. Therefore, the threshold is ultimately set within a slightly increasing range that covers the upper limit of the normal compensatory fluctuation amplitude within a typical sleep cycle. The specific value is combined with the dynamic stability boundary of the photoplethysmography signal used in this system in a high-interference environment and the upper limit of the confidence band calculated from the mean rate of change of the statistical characteristic values ​​within consecutive sleep cycles. It is calibrated through a harmonic regression model, giving it both judgment acuity and effective avoidance of misjudgment caused by random statistical fluctuations.

[0038] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A blood oxygen monitoring system integrating sleep rhythm modeling, characterized in that: The system comprises: a signal sensing module configured to obtain a photoplethysmographic signal and a body motion signal of a monitored person; a rhythm transition recognition module configured to: identify the start time of the transition from non-rapid eye movement sleep to rapid eye movement sleep based on the amplitude envelope morphology of the photoplethysmography signal; wherein the rhythm transition recognition module identifies the start time by calculating the first-order difference of the amplitude envelope of the photoplethysmography signal or the sudden change of short-term energy; A response lag analysis module is configured to: determine the time when the valley value of blood oxygen saturation occurs in combination with the photoplethysmography signal within a predetermined time window after the start time, and calculate the response lag duration; The risk assessment decision module is configured as follows: assigning a credibility weight to each calculated response lag duration based on the body motion signal; accumulating the weighted response lag duration sequence for the entire night, and outputting the respiratory system function risk assessment result based on the aggregation pattern of response lag duration events that are higher than the risk threshold in the weighted response lag duration sequence; the risk assessment decision module evaluates the dynamic response of a single photoplethysmogram signal at a specific physiological rhythm transition point to reveal the respiratory system's instantaneous compensation ability to physiological disturbances.

2. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The risk assessment decision module is also configured to: when the acceleration variance of the body motion signal detected within a predetermined duration exceeds an acceleration variance threshold, mark the corresponding response lag duration measurement value as unreliable and set the credibility weight to zero to discard the measurement value.

3. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The risk assessment decision module is further configured to: after identifying a period of body motion interference indicated by the body motion signal, determine a physiological disturbance index based on the acceleration variance of the body motion signal and the signal-to-noise ratio of the photoplethysmography signal during the period; According to the physiological disturbance index, a compensation coefficient is generated by consulting a preset mapping relationship table; and the compensation coefficient is used to calibrate the first response lag time measured immediately after the body motion disturbance period.

4. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 3, characterized in that: Physiological disturbance index ( ) is calculated as: ,in, represents the acceleration variance of the body motion signal during the body motion interference period, It represents the signal-to-noise ratio of the photoplethysmography signal during the period of body motion interference.

5. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The risk assessment decision module is further configured to: divide the entire night's response lag duration sequence into sleep cycle boundaries determined by a sleep cycle division algorithm; calculate at least one statistical characteristic value for each sleep cycle that characterizes the respiratory stability of the cycle; And by comparing the statistical characteristic values ​​of adjacent sleep cycles, the evolution trend of respiratory system function risk is determined to provide early warning.

6. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 5, characterized in that: The statistical characteristic value is the average or maximum value of the response lag time in each sleep cycle; the generation of trend warning is based on the cross-cycle deterioration index The judgment of whether the trend value is sustained or higher than a predetermined trend threshold is made. The inter-cycle deterioration index is calculated as the ratio of the statistical characteristic value of the current cycle to the statistical characteristic value of the previous cycle. The trend warning indicates that there is a risk of progressive decline in respiratory stability.

7. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The change in the amplitude envelope morphology of the photoplethysmography signal specifically refers to the fact that during the rapid eye movement period, the amplitude of the photoplethysmography signal exhibits a morphology with higher irregularity and a decreased average amplitude compared to the non-rapid eye movement period.

8. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The decision logic adopted by the risk assessment decision module includes: comparing the response lag time with at least one critical threshold to generate an event risk level; and if the frequency of occurrence of the event risk level exceeding the predetermined risk level exceeds a predetermined frequency threshold within a single sleep cycle, then the respiratory system function risk assessment level is increased.

9. The blood oxygen monitoring system integrating sleep rhythm modeling according to claim 1, characterized in that: The rhythm transition recognition module identifies the starting moment of the transition from non-rapid eye movement (NREM) to rapid eye movement (REM) sleep by analyzing the morphological characteristics of the single-channel photoplethysmogram (PPE) signal at the wrist.

Images