Bioelectric signal-based intelligent oxygen sleep engine method and device

Abstract

The application provides an intelligent oxygen sleep engine method and device based on bioelectric signals, and relates to the technical field of data processing.The method comprises the following steps: collecting carotid artery blood oxygen signals and jugular vein blood oxygen signals corresponding to a user's carotid artery surface projection point, a midpoint of the anterior edge of sternocleidomastoid muscle and a supracleavicular fossa center, respectively; based on the carotid artery blood oxygen signals and the jugular vein blood oxygen signals, taking the carotid artery surface projection point as the center of an ellipse, the midpoint of the anterior edge of sternocleidomastoid muscle and the supracleavicular fossa center as the endpoints of the long axis of the ellipse, adjusting the length of the short axis of the ellipse according to the actual size of the neck region, obtaining an elliptical region of the contour of the neck region, dividing the elliptical region by orthogonal lines inscribed in the ellipse to obtain a plurality of fan-shaped sub-regions. The application realizes targeted perception and fine diffusion oxygen supply of oxygen supply demand.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and apparatus for a smart oxygenation sleep engine based on bioelectrical signals. Background Technology

[0002] In scenarios such as sleep-disordered breathing and high-altitude hypoxia, nocturnal hypoxia has become a key issue affecting sleep quality and health. Intelligent oxygen supply technology based on physiological signals has gradually become a research hotspot. Most existing home oxygen generators monitor peripheral blood oxygen saturation (SpO2) through finger clip pulse oximeters and trigger diffusion oxygen supply based on fixed thresholds. Although some devices have attempted to combine respiratory signals to optimize timing, there are still problems with insufficient accuracy.

[0003] For example, when using a smart diffusion oxygen concentrator, the device monitors SpO2 via a finger pulse oximeter and presets a fixed oxygen supply of 3L per minute when SpO2 falls below 90%. It also relies on a pressure sensor to detect respiratory status for auxiliary regulation. However, in actual use, the finger-clip pulse oximeter often shifts due to turning over, causing signal interruption or delay. Furthermore, peripheral SpO2 only reflects the average blood oxygen level throughout the body and cannot accurately capture the oxygen supply gap to the brain through the arteriovenous blood oxygen difference in the neck region. It also fails to distinguish the differences in oxygen supply needs between light and deep sleep stages. Fixed-intensity oxygen supply can easily lead to over-intervention during deep sleep and insufficient oxygen supply during light sleep. In addition, the lack of regional calibration of signals collected from the neck region makes the signals highly susceptible to interference from physiological structural differences, ultimately resulting in delayed oxygen supply triggering and intensity mismatch, which affects sleep continuity and reduces oxygen utilization efficiency. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and device for intelligent oxygenation sleep engine based on bioelectric signals, so as to realize targeted perception of oxygen supply demand and refined diffusion oxygen supply.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a smart oxygenation sleep engine method based on bioelectrical signals, the method comprising:

[0007] The carotid artery blood oxygenation signal and jugular vein blood oxygenation signal were collected from the carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa, respectively.

[0008] Based on the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal, the projection point of the carotid artery on the body surface is taken as the center of the ellipse, and the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa are taken as the endpoints of the major axis of the ellipse. The length of the minor axis of the ellipse is adjusted according to the actual size of the neck region to obtain the elliptical region of the neck region. The elliptical region is divided by the inscribed orthogonal lines of the ellipse to obtain multiple fan-shaped sub-regions.

[0009] Calibration parameters are generated based on the signal characteristics of each sub-region; the collected carotid artery blood oxygenation signal and jugular vein blood oxygenation signal are calibrated using the calibration parameters to obtain the calibrated carotid artery blood oxygenation signal and jugular vein blood oxygenation signal.

[0010] Based on the calibrated carotid artery and jugular vein blood oxygenation signals, the arteriovenous blood oxygen difference is calculated in real time.

[0011] Based on the arteriovenous blood oxygen difference and combined with the user's current sleep stage, determine whether the oxygen supply triggering condition is met, and determine the oxygen supply intensity when the condition is met to obtain the judgment result;

[0012] Based on the judgment results, when the oxygen supply triggering conditions are met, diffuse oxygen supply is executed according to the determined oxygen supply intensity, releasing oxygen into the indoor environment where the user is located, continuously monitoring changes in arterial and venous blood oxygen difference, and adjusting the oxygen supply intensity or shutting off the oxygen supply according to the changes.

[0013] Furthermore, based on the carotid artery and jugular vein oxygenation signals, with the carotid artery surface projection point as the center of an ellipse, and the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis of the ellipse, the length of the minor axis of the ellipse is adjusted according to the actual size of the neck region to obtain an elliptical region of the neck contour. This elliptical region is then divided by inscribed orthogonal lines to obtain multiple fan-shaped sub-regions, including:

[0014] Based on the established two-dimensional coordinate system, the projection point of the carotid artery on the body surface is determined as the center of the ellipse. Using the center of the ellipse as a reference, the distance between the center and the center of the ellipse is calculated according to the coordinate position of the midpoint of the anterior border of the sternocleidomastoid muscle. The center distance is set as the length of the major semi-axis of the ellipse.

[0015] Based on the center of the ellipse and the length of its major semi-axis, the projection distance to the center of the ellipse is calculated according to the projection position of the center of the supraclavicular fossa in the direction of the minor axis of the two-dimensional coordinate system. The projection distance is then weighted and corrected in combination with the physiological parameters of the user's anteroposterior diameter of the neck, and the length of the minor semi-axis of the ellipse is dynamically determined, thereby constructing an elliptical detection area that fully covers the carotid artery and jugular vein regions.

[0016] Based on the constructed elliptical detection region, with the center of the ellipse as the origin of the coordinate system, the elliptical region is divided into multiple rectangular sub-regions by using equally spaced orthogonal grid lines parallel to the major and minor axes of the ellipse.

[0017] Furthermore, calibration parameters are generated based on the signal characteristics of each sub-region; the acquired carotid artery and jugular vein oxygenation signals are then calibrated using these parameters to obtain calibrated carotid artery and jugular vein oxygenation signals, including:

[0018] Based on the division of multiple sub-regions and region identifiers, the time-domain and frequency-domain features of the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal collected in each sub-region are extracted respectively. By comparing the signal-to-noise ratio and stability of the signals in each sub-region, the dominant sub-region used for benchmark reference is determined.

[0019] Based on the signal characteristics of the dominant sub-region, the characteristic deviation of each sub-region signal relative to the dominant sub-region signal is calculated. In a two-dimensional coordinate system, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated. Using the length of the semi-major axis of the elliptical region as a reference, the attenuation coefficient is used to calculate the geometric weight of each sub-region through a predefined mathematical attenuation function. The geometric weight is combined with the characteristic deviation to obtain the signal calibration parameters corresponding to each sub-region.

[0020] The signal calibration parameters are applied to the corresponding sub-regions, and the original carotid artery blood oxygenation signal and jugular vein blood oxygenation signal collected in each sub-region are subjected to amplitude correction and time alignment processing to obtain the calibrated blood oxygenation signal of each sub-region.

[0021] Based on the calibrated blood oxygenation signals of each sub-region, spatial fusion processing is performed according to the region identifier to obtain the calibrated integrated carotid artery blood oxygenation signal and integrated jugular vein blood oxygenation signal, respectively.

[0022] Furthermore, based on the signal characteristics of the dominant sub-region, the characteristic deviation of each sub-region signal relative to the dominant sub-region signal is calculated. In a two-dimensional coordinate system, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated. Using the attenuation coefficient with the length of the semi-major axis of the elliptical region as a reference, the geometric weight of each sub-region is calculated through a predefined mathematical attenuation function. The geometric weight is combined with the characteristic deviation to obtain the signal calibration parameters corresponding to each sub-region, including:

[0023] The feature vector of the dominant sub-region signal is extracted as a reference. Based on the feature vector reference, the difference between the feature vector of each sub-region signal and the reference vector is calculated in turn to obtain the initial feature deviation of each sub-region.

[0024] Based on the identifiers of each sub-region, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated sequentially in a two-dimensional coordinate system. The calculated Euclidean distance is then used in conjunction with an attenuation coefficient determined based on the length of the semi-major axis of the elliptical region. This attenuation is then processed by a predefined mathematical attenuation function to obtain the spatial geometric weight corresponding to each sub-region.

[0025] The initial feature deviation of each sub-region and the spatial geometric weight of the corresponding sub-region are received and weighted and fused to obtain the signal calibration parameters of the original signal of each sub-region.

[0026] Furthermore, based on the calibrated carotid artery and jugular venous oxygenation signals, the arteriovenous oxygen difference is calculated in real time, including:

[0027] Based on the calibrated integrated carotid artery blood oxygenation signal and integrated jugular vein blood oxygenation signal, the real-time time series of carotid artery blood oxygen saturation value and jugular vein blood oxygen saturation value were extracted respectively.

[0028] Time alignment processing was performed on the time series of carotid artery blood oxygen saturation values ​​and jugular vein blood oxygen saturation values ​​to eliminate the timing deviation caused by different signal acquisition paths, so as to obtain the result that eliminates the timing deviation that may be caused by different signal acquisition paths.

[0029] The instantaneous arterial and venous oxygen difference at each moment is calculated by subtracting the time-aligned carotid artery oxygen saturation value sequence from the jugular vein oxygen saturation value sequence point by point.

[0030] Furthermore, based on the arteriovenous oxygen difference and the user's current sleep stage, it is determined whether the oxygen supply triggering condition is met, and if the condition is met, the oxygen supply intensity is determined to obtain the judgment result, including:

[0031] It receives arterial and venous blood oxygen difference values ​​and simultaneously receives information on the user's current sleep stage determined by analyzing heart rate variability and body movement signals;

[0032] Based on the user's current sleep stage information, the reference range of arterial and venous blood oxygen difference and the oxygen supply trigger threshold corresponding to the sleep stage are dynamically selected; the reference range and the oxygen supply trigger threshold are obtained by personalized modeling and updating based on the user's historical sleep data through machine learning model;

[0033] The real-time acquired arteriovenous blood oxygen difference data stream is compared with the oxygen supply trigger threshold under the current sleep stage, and a joint analysis is performed based on the statistical trend characteristics of the arteriovenous blood oxygen difference data stream within a preset time window to obtain the oxygen supply trigger conditions.

[0034] After determining that the oxygen supply triggering conditions are met, the required oxygen supply intensity level is determined based on the deviation of the real-time arterial and venous blood oxygen difference from the benchmark reference range and the specific sleep stage, through a predefined oxygen supply intensity mapping rule, in order to obtain the final judgment result.

[0035] Furthermore, based on the judgment results, when the oxygen supply triggering conditions are met, diffused oxygen supply is executed according to the determined oxygen supply intensity, releasing oxygen into the user's indoor environment, continuously monitoring changes in arterial and venous blood oxygen difference, and adjusting the oxygen supply intensity or shutting off the oxygen supply according to the changes, including:

[0036] Receive the final judgment result and parse the oxygen supply instructions and oxygen supply intensity level contained in the judgment result;

[0037] Based on the oxygen supply intensity level obtained from the analysis, the corresponding equipment control signal is obtained, and oxygen is released into the indoor environment according to the set initial oxygen supply intensity.

[0038] After oxygen supply is activated, it continuously receives updated arterial and venous blood oxygen difference data streams and simultaneously monitors the user's real-time sleep stage changes.

[0039] Based on the continuous monitoring of arteriovenous blood oxygen difference data stream and real-time sleep stages, the actual effect of oxygen supply intervention is dynamically evaluated, and it is determined whether the conditions for adjusting oxygen supply intensity or stopping oxygen supply are met, so as to obtain the judgment result.

[0040] Based on the judgment result, the corresponding operation is executed. If the conditions for adjusting the oxygen supply intensity are met, the control command for upgrading or downgrading the oxygen supply intensity is generated and executed according to the current value and trend of the arteriovenous blood oxygen difference and the current sleep stage, in accordance with the predefined adjustment strategy. If the conditions for stopping the oxygen supply are met, the control command for gradually reducing the oxygen supply intensity until it is completely shut down is generated and executed.

[0041] Secondly, the intelligent oxygenation sleep engine device based on bioelectrical signals includes:

[0042] The acquisition module is used to collect the carotid artery blood oxygenation signal and jugular vein blood oxygenation signal corresponding to the user's carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa, respectively.

[0043] The segmentation module is used to obtain an elliptical region of the neck region based on the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal, with the carotid artery surface projection point as the center of the ellipse, the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis of the ellipse, and the length of the minor axis of the ellipse adjusted according to the actual size of the neck region to obtain the elliptical region of the neck region. The elliptical region is then divided by the inscribed orthogonal lines of the ellipse to obtain multiple fan-shaped sub-regions.

[0044] The calibration module is used to generate calibration parameters based on the signal characteristics of each sub-region; the collected carotid artery blood oxygenation signal and jugular vein blood oxygenation signal are calibrated using the calibration parameters to obtain the calibrated carotid artery blood oxygenation signal and jugular vein blood oxygenation signal.

[0045] The calculation module is used to calculate the arteriovenous oxygen difference in real time based on the calibrated carotid artery blood oxygen signal and jugular vein blood oxygen signal;

[0046] The judgment module is used to determine whether the oxygen supply triggering conditions are met based on the arteriovenous blood oxygen difference and the user's current sleep stage, and to determine the oxygen supply intensity when the conditions are met, so as to obtain the judgment result;

[0047] The processing module is used to perform diffused oxygen supply according to the determined oxygen supply intensity when the oxygen supply triggering conditions are met, release oxygen into the indoor environment where the user is located, continuously monitor the changes in arterial and venous blood oxygen difference, and adjust the oxygen supply intensity or shut down the oxygen supply according to the changes.

[0048] Thirdly, a computing device includes:

[0049] One or more processors;

[0050] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0051] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0052] The above-described solution of the present invention has at least the following beneficial effects:

[0053] A precise acquisition scheme is developed that uses the carotid artery surface projection point as the center of an ellipse and dynamically constructs an elliptical detection area covering the arteries and veins, divided into rectangular sub-regions, based on the actual physiological dimensions of the neck region. This is combined with a multi-dimensional signal calibration technique that integrates dominant sub-region selection, feature deviation calculation, and geometric weighting using Euclidean distance and the attenuation coefficient of the ellipse's semi-major axis. Simultaneously, stable calculation of the arteriovenous oxygen difference is achieved through time-series alignment of blood oxygen saturation and sliding window smoothing filtering. Furthermore, an oxygen supply decision-making logic based on personalized threshold modeling using machine learning based on user historical data and trend analysis of the arteriovenous oxygen difference is constructed, resulting in a closed-loop diffusion oxygen supply control system encompassing acquisition, calibration, calculation, decision-making, execution, and real-time feedback adjustment. This system overcomes the limitations of existing technologies. The study addresses several technical challenges, including incomplete coverage of neck arterial and venous signal acquisition, significant deviations due to physiological structural differences, instability in arteriovenous oxygen difference calculation caused by temporal deviations and noise, fixed oxygen supply trigger thresholds that fail to meet sleep stage requirements, and a mismatch between oxygen supply intensity and actual brain oxygen demand due to a lack of dynamic feedback. The goal is to significantly improve the accuracy and stability of neck blood oxygen signal acquisition, ensuring that arteriovenous oxygen difference data accurately and in real-time reflects the brain's oxygen metabolism status. Simultaneously, it allows for dynamic adjustment of oxygen supply intervention intensity based on the user's real-time sleep stage and changes in arteriovenous oxygen difference, avoiding both over-oxygenation and under-intervention while ensuring a seamless sleep experience, ultimately improving the accuracy and efficiency of brain oxygen supply regulation. Attached Figure Description

[0054] Figure 1 This is a schematic flowchart of the intelligent oxygenation sleep engine method based on bioelectric signals provided in an embodiment of the present invention.

[0055] Figure 2This is a schematic diagram of a smart oxygenation sleep engine device based on bioelectric signals provided in an embodiment of the present invention. Detailed Implementation

[0056] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0057] like Figure 1 As shown, embodiments of the present invention propose a smart oxygenation sleep engine method based on bioelectrical signals, the method comprising the following steps:

[0058] Step 1: Collect the carotid artery blood oxygenation signal and jugular vein blood oxygenation signal corresponding to the carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa, respectively.

[0059] Step 2: Based on the carotid artery blood oxygen signal and the jugular vein blood oxygen signal, take the carotid artery surface projection point as the center of the ellipse, and the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis of the ellipse. Adjust the length of the minor axis of the ellipse according to the actual size of the neck region to obtain the elliptical region of the neck region outline. Divide the elliptical region into multiple fan-shaped sub-regions by inscribed orthogonal lines in the ellipse.

[0060] Step 3: Generate calibration parameters based on the signal characteristics of each sub-region; calibrate the collected carotid artery blood oxygenation signal and jugular vein blood oxygenation signal using the calibration parameters to obtain the calibrated carotid artery blood oxygenation signal and jugular vein blood oxygenation signal.

[0061] Step 4: Calculate the arteriovenous oxygen difference in real time based on the calibrated carotid artery and jugular vein oxygen signals.

[0062] Step 5: Based on the arteriovenous blood oxygen difference and the user's current sleep stage, determine whether the oxygen supply triggering condition is met, and determine the oxygen supply intensity when the condition is met to obtain the judgment result;

[0063] Step 6: Based on the judgment result, when the oxygen supply triggering condition is met, perform diffused oxygen supply according to the determined oxygen supply intensity, release oxygen into the indoor environment where the user is located, continuously monitor the changes in arterial and venous blood oxygen difference, and adjust the oxygen supply intensity or shut down the oxygen supply according to the changes.

[0064] In this embodiment of the invention, the invention overcomes the limitations of existing technologies by employing precise acquisition of carotid and jugular vein blood oxygenation signals from three key points: the carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa. It constructs an elliptical detection region with the carotid artery surface projection point as the center and the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis, and divides this region into sector-shaped sub-regions. Calibration parameters are then generated based on the signal characteristics of each sub-region to calibrate the original signal. Subsequently, the arteriovenous oxygen difference is calculated based on the calibrated signal. The oxygen supply triggering conditions and intensity are determined by considering the user's current sleep stage. Finally, diffuse oxygen supply is implemented, and the changes in the arteriovenous oxygen difference are continuously monitored to dynamically adjust the oxygen supply intensity or shut down the oxygen supply. The lack of targeted blood oxygenation signal acquisition points and regional coverage that do not conform to the physiological structure of the neck result in insufficient signal representativeness. The failure to perform sub-regional differential calibration of the signals leads to signal deviation. The failure to link oxygen supply decisions to sleep stages causes the triggering timing and intensity to deviate from the actual oxygen demand of the brain. The lack of real-time feedback and adjustment results in over- or under-oxygen supply. In order to improve the accuracy and signal quality stability of arterial and venous blood oxygenation signal acquisition in the neck region, the arterial and venous blood oxygen difference can truly reflect the brain's oxygen metabolism status. Oxygen supply intervention can accurately match the brain's oxygen demand under different sleep stages, which not only avoids interference with the sleep process, but also improves the adaptability and efficiency of brain oxygen supply. Ultimately, it helps to improve the user's sleep quality and prolong the duration of deep sleep.

[0065] In a preferred embodiment of the present invention, step 1 above may include:

[0066] Step 1.1: Based on the anatomical features of the neck, the three key acquisition locations are determined. The optical sensing probe and bioelectrical impedance electrode are precisely aligned with the corresponding areas using a multi-sensor array on the flexible neck strap. Multi-channel signal acquisition is initiated simultaneously to acquire arterial and venous signal sources. The raw signals are then subjected to real-time quality assessment and parameter adaptive adjustment. Finally, the quality-verified and accurately positioned multi-channel carotid artery and jugular vein blood oxygenation signals are output. Specifically, the physiological coordinates of the three key acquisition locations—the carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa—are determined by combining the anatomical features of the human neck. At the same time, adjustable positioning markers are set on the surface of the flexible neck strap according to the individual physiological differences of different users, such as neck circumference and neck thickness, to ensure that each key location can be accurately matched with the actual anatomical structure of the user's neck.

[0067] A multi-sensor array integrating optical sensing probes and bioelectrical impedance electrodes is installed on a flexible neck strap. Each key acquisition position is equipped with a set of optical sensing probes and bioelectrical impedance electrodes. When the user wears the flexible neck strap, the position of the neck strap is adjusted by the positioning marks on the neck strap, so that each set of sensors is accurately attached to the corresponding key acquisition area. The elasticity of the flexible neck strap is used to fit the curve of the neck, avoiding sensor displacement due to movements such as turning over.

[0068] Subsequently, multi-channel signal acquisition was initiated. Optical sensing probes captured real-time changes in blood oxygen concentration at key locations. Simultaneously, bioelectrical impedance analysis (BIA) was used to detect the impedance characteristics of blood flow within the blood vessels of the detection area. Based on the differences in the frequency and amplitude of impedance changes between arterial and venous blood flow, the acquired signals were initially separated, extracting carotid artery and jugular venous blood oxygen signals separately to avoid signal aliasing. During signal acquisition, the system continuously assessed the quality of the raw signals, monitoring whether the signal-to-noise ratio met preset standards and whether there were signal interruptions or fluctuations due to poor skin contact. When substandard signal quality was detected, the system automatically adjusted parameters such as the sensor's sampling frequency, signal gain, and the contact pressure between the probe and the skin to ensure continuous signal stability. The adjusted signals were then re-verified to select those with acceptable signal-to-noise ratios and no significant interference, resulting in accurate and reliable multi-channel carotid artery and jugular venous blood oxygen signals.

[0069] In this embodiment of the invention, by employing a technique that determines three key acquisition locations based on neck anatomical features, precisely aligns the optical sensing probe and bioelectrical impedance electrode in the corresponding areas using a multi-sensor array on a flexible neckband, simultaneously initiates multi-channel signal acquisition to distinguish between arterial and venous signal sources, and performs real-time quality assessment and parameter adaptive adjustment of the raw signals, ultimately outputting a quality-verified and accurately positioned multi-channel blood oxygenation signal, this invention overcomes the technical problems in existing technologies, such as positioning deviations due to lack of anatomical basis for blood oxygenation signal acquisition locations, signal instability caused by poor fit between the sensor and neck skin, signal aliasing caused by the inability of single-channel acquisition to effectively separate arterial and venous signals, and invalid data interference analysis due to the lack of real-time quality screening of raw signals. This results in improved accuracy of blood oxygenation signal acquisition locations and enhanced sensor wearing comfort, clear distinction between arterial and venous signal sources, real-time controllable raw signal quality, and high data reliability.

[0070] In a preferred embodiment of the present invention, step 2 above may include:

[0071] Step 2.1: Based on the established two-dimensional coordinate system, the projection point of the carotid artery on the body surface is determined as the center of the ellipse. Using this ellipse center as a reference, the distance between the center and the midpoint of the anterior border of the sternocleidomastoid muscle is calculated according to the coordinate position of the midpoint. This center distance is then set as the length of the semi-major axis of the ellipse. Specifically, this includes: establishing a two-dimensional coordinate system with the physiological contour of the user's neck side as a reference surface; setting the direction parallel to the vertical direction of the neck as the major axis (x-axis) and the direction perpendicular to the major axis and along the anterior-posterior direction of the neck as the minor axis (y-axis); then, accurately locating the projection point of the carotid artery on the body surface according to the anatomical features of the neck; using the preset anatomical position markers on the flexible neck strap, determining the position of this projection point in the two-dimensional coordinate system as the center of the ellipse and recording its coordinates; then, finding the actual position on the neck strap corresponding to the midpoint of the anterior border of the sternocleidomastoid muscle, recording the coordinates of this midpoint in the two-dimensional coordinate system; using the coordinates of the ellipse center as a reference, calculating the straight-line distance between the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the ellipse using the distance calculation method within the coordinate system; finally, setting this straight-line distance directly as the length of the semi-major axis of the ellipse.

[0072] Step 2.2: Based on the center and semi-major axis length of the ellipse, calculate the projection distance to the center of the ellipse according to the projection position of the center of the supraclavicular fossa in the short axis direction of the two-dimensional coordinate system. Combine the physiological parameters of the user's anteroposterior diameter of the neck to perform weighted correction on the projection distance, dynamically determine the length of the semi-major axis of the ellipse, and thus construct an elliptical detection area that completely covers the carotid artery and jugular vein region. Specifically, this includes: based on the determined center and semi-major axis length of the ellipse, finding the actual position of the center of the supraclavicular fossa according to the anatomical features of the neck, determining the projection point of the center in the short axis direction (y-axis) of the two-dimensional coordinate system, recording the coordinates of this projection point and calculating the distance between it and the center of the ellipse. The distance is used as the initial reference distance in the short axis direction. Next, the physiological parameters of the user's anteroposterior diameter of the neck are obtained. These parameters can be acquired through personal body data entered by the user before use, or measured in real time during wear by a miniature ranging sensor integrated on the flexible neckband. Then, the initial reference distance in the minor axis direction is weighted and corrected based on the value of the anteroposterior diameter of the neck. For example, the correction coefficient is appropriately increased to expand the minor axis length when the anteroposterior diameter of the neck is large, and the correction coefficient is appropriately decreased when the anteroposterior diameter of the neck is small. The final length of the minor semi-axis of the ellipse is determined through dynamic adjustment. Finally, by combining the center of the ellipse, the length of the major semi-axis, and the corrected length of the minor semi-axis, a complete elliptical detection area is constructed in a two-dimensional coordinate system to ensure that the area can fully cover the key areas of the carotid artery and jugular vein in the neck.

[0073] Step 2.3: Based on the constructed elliptical detection region, using the center of the ellipse as the origin, the elliptical region is divided into multiple rectangular sub-regions by drawing equally spaced orthogonal grid lines parallel to the major and minor axes of the ellipse. Specifically, this involves: using the center of the constructed elliptical detection region as the origin of the two-dimensional coordinate system, defining the direction of the major axis (x-axis) and the direction of the minor axis (y-axis) of the ellipse; then, based on the distribution density of the sensor probes in the multi-sensor array on the flexible neck strap, setting equally spaced grid line intervals to ensure that the grid line spacing matches the effective detection range of the sensor probes, so that each sub-region is covered by at least one sensor probe. Then, multiple equally spaced grid lines parallel to the major axis and multiple equally spaced grid lines parallel to the minor axis of the ellipse are drawn, with the orthogonal grid lines intersecting each other, dividing the entire elliptical detection region into multiple uniformly sized rectangular sub-regions.

[0074] In this embodiment of the invention, because a two-dimensional coordinate system is established, the projection point of the carotid artery on the body surface is determined as the center of an ellipse. The length of the major semi-axis of the ellipse is set by calculating the distance between the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the ellipse. The length of the minor semi-axis is determined by weighted correction of the projection distance of the center of the supraclavicular fossa in the minor axis direction and the physiological parameters of the user's anteroposterior diameter of the neck, thus constructing an elliptical detection area covering the arteries and veins. Then, with the center of the ellipse as the origin, the elliptical area is divided into multiple rectangular sub-regions by equally spaced orthogonal grid lines parallel to the major and minor axes. Therefore, this invention overcomes the technical problems in the prior art, such as incomplete arteriovenous coverage due to lack of anatomical basis in the division of the neck detection area, poor regional adaptability due to lack of consideration of individual differences in neck physiological size, and insufficient targeting of subsequent signal analysis due to disordered sub-region division and lack of clear spatial reference. As a result, the elliptical detection area can accurately fit the physiological structure of the neck and completely cover the key areas of the carotid artery and jugular vein, with orderly sub-region division and clear spatial positioning.

[0075] In a preferred embodiment of the present invention, step 3 above may include:

[0076] Step 3.1: Based on the divided sub-regions and their identifiers, extract the time-domain and frequency-domain features of the carotid artery and jugular vein oxygenation signals collected within each sub-region. Then, by comparing the signal-to-noise ratio and stability of the signals from each sub-region, determine the dominant sub-region for reference. Specifically, this includes: First, based on the previously divided rectangular sub-regions and their corresponding identifiers, map the carotid artery and jugular vein oxygenation signals collected in each sub-region to the identifiers one-to-one, ensuring accurate matching between the signals and their source regions. Next, extract the time-domain and frequency-domain features for the two oxygenation signals in each sub-region. The features include the average amplitude of the signal over a period of time, the amplitude fluctuation range, and the duration of the rising and falling edges of the signal. The frequency domain features include the main frequency components of the signal, the frequency distribution range, and the energy proportion of different frequency components. Then, the signal-to-noise ratio (SNR) of the two blood oxygenation signals in each sub-region is calculated. The SNR value is obtained by comparing the intensity of the effective signal component with the intensity of the noise component. At the same time, the fluctuation amplitude of the signal during the continuous acquisition period is observed to determine the signal stability. Finally, the signal SNR and stability of all sub-regions are quantitatively compared, and the sub-region with the highest SNR and the smallest fluctuation amplitude is selected as the dominant sub-region for subsequent benchmark reference.

[0077] Step 3.2: Based on the signal characteristics of the dominant sub-region, calculate the characteristic deviation of each sub-region signal relative to the dominant sub-region signal; in a two-dimensional coordinate system, calculate the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region, and calculate the geometric weight of each sub-region using the attenuation coefficient with the length of the major semi-axis of the elliptical region as a reference, through a predefined mathematical attenuation function; combine the geometric weight with the characteristic deviation to obtain the signal calibration parameters corresponding to each sub-region, specifically including: first, using the time-domain and frequency-domain characteristics of the carotid artery blood oxygen signal and the jugular vein blood oxygen signal of the dominant sub-region as reference features, comparing them one by one with the characteristics of the corresponding signals of each other sub-regions, and calculating the differences between the two in terms of amplitude, frequency components, fluctuation range, etc., the difference values ​​are the characteristic deviation of each sub-region signal; then, in the previously established two-dimensional coordinate system, determine the characteristic deviation of each sub-region signal. The coordinates of the center point of the region and the center point of the dominant sub-region are calculated using a distance calculation method within the coordinate system to obtain the straight-line distance between the two points, which is the Euclidean distance. Then, using the length of the semi-major axis of the elliptical region as a reference, a calculation standard for the attenuation coefficient is set. For example, when the Euclidean distance between the sub-region and the dominant sub-region is equal to the length of the semi-major axis, the attenuation coefficient is set to a fixed value. The closer the distance, the smaller the attenuation coefficient; the farther the distance, the larger the attenuation coefficient. Then, using a predefined mathematical attenuation function, the Euclidean distance and the attenuation coefficient are substituted into the calculation to obtain the geometric weight of each sub-region. Finally, the geometric weight of each sub-region is combined with the corresponding feature deviation for calculation. For example, the geometric weight is used to adjust the feature deviation. The closer the sub-region is to the dominant sub-region, the smaller the correction of the feature deviation by the geometric weight, and vice versa. Finally, the signal calibration parameters specific to each sub-region are obtained.

[0078] Step 3.3: Apply the signal calibration parameters to the corresponding sub-regions. Perform amplitude correction and timing alignment processing on the original carotid artery blood oxygenation signal and jugular vein blood oxygenation signal collected in each sub-region to obtain the calibrated blood oxygenation signal of each sub-region. Specifically, this includes: First, applying the obtained signal calibration parameters of each sub-region to the original signal of each sub-region according to the region identifier. For amplitude correction, the amplitudes of the original carotid artery and jugular vein oxygenation signals acquired in each sub-region are corrected according to the adjustment rules for signal amplitude in the calibration parameters. For example, if the signal amplitude of a certain sub-region is lower than the reference value, its amplitude is increased to a range that matches the reference characteristics according to the calibration parameters, ensuring that the amplitudes of the same type of signal in different sub-regions are at a consistent reference level. For timing alignment, the time delay caused by differences in sensor position or signal transmission speed during the acquisition process of signals in each sub-region is analyzed according to the compensation rules for signal acquisition time difference in the calibration parameters. The time axis of the signal is adjusted so that the values ​​of carotid artery oxygenation signals in all sub-regions correspond to each other at the same time point, and the jugular vein oxygenation signals are also synchronized in time. Finally, the carotid artery and jugular vein oxygenation signals of each sub-region are obtained after amplitude correction and timing alignment.

[0079] Step 3.4: Based on the calibrated blood oxygenation signals of each sub-region, spatial fusion processing is performed according to the region identifier to obtain the calibrated integrated carotid artery blood oxygenation signal and integrated jugular vein blood oxygenation signal, respectively. Specifically, this includes: First, collecting the calibrated carotid artery and jugular vein blood oxygenation signals from all sub-regions, and classifying and organizing the two types of signals according to the region identifier to ensure that signals of the same type are grouped together and traceable to their corresponding sub-regions; then, setting fusion weights based on the importance of each sub-region's location within the elliptical detection area. For example, sub-regions closer to the carotid artery surface projection point have a higher fusion weight for carotid artery blood oxygenation signals, while sub-regions closer to the main distribution area of ​​the jugular vein have a higher fusion weight for the jugular vein. The fusion weight of the blood oxygen signal is higher, and the signal-to-noise ratio and stability of each sub-region signal are taken into account. The sub-region with better signal quality has a higher fusion weight. Then, according to the set fusion weight, the carotid blood oxygen signals after calibration of all sub-regions are weighted and calculated. The values ​​of each sub-region signal at the same time point are superimposed according to the weight to obtain the comprehensive carotid blood oxygen signal. The carotid blood oxygen signals after calibration of all sub-regions are weighted and fused in the same way to obtain the comprehensive carotid blood oxygen signal. Finally, the integrity of the two comprehensive signals after fusion is checked to ensure that the signal is continuous and uninterrupted and that the value does not fluctuate abnormally. Finally, the calibrated comprehensive carotid blood oxygen signal and comprehensive carotid blood oxygen signal are output.

[0080] In this embodiment of the invention, because a division based on sub-regions and region identifiers is used, the time-domain and frequency-domain features of arterial and venous blood oxygenation signals of each sub-region are extracted, and the dominant sub-region is determined by comparing the signal-to-noise ratio and stability; the feature deviation of each sub-region is calculated based on the signal of the dominant sub-region; the geometric weight is calculated by combining the Euclidean distance between the center point of the sub-region and the center point of the dominant sub-region, the attenuation coefficient of the semi-major axis of the ellipse reference, and the predefined mathematical attenuation function; the geometric weight and feature deviation are fused to obtain the signal calibration parameters of each sub-region; the calibration parameters are applied to the corresponding sub-regions for amplitude correction and timing alignment, and then the signals are categorized by region. The technique of spatially fusing calibrated signals to generate a comprehensive arterial and venous blood oxygenation signal overcomes the technical problems in existing technologies, such as the lack of regional calibration of neck region acquisition signals leading to significant interference from physiological structural differences, inconsistent amplitude and timing of sub-region signals, and low reliability of the comprehensive signal due to the lack of signal fusion methods. As a result, the quality defects of blood oxygenation signals in each sub-region are accurately corrected, and the amplitude consistency and time synchronization of arterial and venous blood oxygenation signals are significantly improved. The final output comprehensive carotid artery and jugular vein blood oxygenation signal can more realistically and accurately reflect the physiological state of carotid artery and venous blood oxygenation.

[0081] In a preferred embodiment of the present invention, step 3.2 above may include:

[0082] Step 3.21: Extract the feature vector of the dominant sub-region signal as a benchmark. Based on the feature vector benchmark, calculate the difference between the feature vector of each sub-region signal and the benchmark vector to obtain the initial feature deviation for each sub-region. Specifically, this includes: First, for the determined dominant sub-region, extract the time-domain and frequency-domain features of its carotid artery blood oxygenation signal and jugular vein blood oxygenation signal. The time-domain features include the average amplitude of the signal during the continuous acquisition period, the difference between the maximum and minimum amplitude (i.e., the fluctuation range), the duration of the rising edge from the start to the peak, and the duration of the falling edge from the peak to the start. The frequency-domain features include the main frequency component with the highest energy proportion in the signal, the proportion of the main frequency component in the total signal energy, and the interval of the signal frequency distribution. The time-domain and frequency-domain features are integrated in a preset fixed order to form the reference feature vectors corresponding to the two blood oxygenation signals in the dominant sub-region. Then, according to the same feature types and arrangement order as the reference feature vectors, the features of the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal in each other sub-region are extracted to construct the comparison feature vectors corresponding to each sub-region. Then, for each blood oxygenation signal, the comparison feature vector of each sub-region is compared with the reference feature vector of the dominant sub-region dimension by dimension, and the numerical difference between the two in each feature dimension is calculated, such as the difference in the average amplitude, the difference in the main frequency, etc. The difference values ​​of all feature dimensions are integrated in the original feature order to finally form the initial feature deviation of the two blood oxygenation signals corresponding to each sub-region.

[0083] Step 3.22: Based on the identifiers of each sub-region, calculate the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region in a two-dimensional coordinate system. Then, calculate the calculated Euclidean distance and the attenuation coefficient determined based on the semi-major axis of the elliptical region. Process this attenuation using a predefined mathematical attenuation function to obtain the spatial geometric weight corresponding to each sub-region. Specifically, this includes: First, based on the region identifiers of each sub-region, find the spatial range corresponding to each sub-region in the previously established two-dimensional coordinate system. Calculate the geometric center coordinates of the spatial range of the sub-region to determine the center point coordinates of each sub-region. Simultaneously, determine the center point coordinates of the dominant sub-region. These coordinates can be obtained by calculating the geometric center of the spatial range of the dominant sub-region, or by directly using the elliptical center coordinates of the previously detected elliptical region. Next, for each sub-region, calculate the straight-line distance between the two points based on the center point coordinates and the center point coordinates of the dominant sub-region. This distance is the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region. Then, the attenuation coefficient is determined based on the length of the semi-major axis of the elliptical detection region: when the Euclidean distance between the sub-region and the center point of the dominant sub-region is 0 (i.e., the sub-region contains the center point of the dominant sub-region), the attenuation coefficient is 1; when the Euclidean distance is equal to the length of the semi-major axis, the attenuation coefficient is 0.5; and when the Euclidean distance is greater than the length of the semi-major axis, the attenuation coefficient is 0.3, forming an attenuation coefficient setting rule based on distance gradient. After that, a predefined mathematical attenuation function, such as a linear attenuation function, is called, and the Euclidean distance of each sub-region and the corresponding attenuation coefficient are substituted into the function for calculation. For example, when using a linear attenuation function, the product of the distance and the attenuation coefficient is used as the function input, and the output result is the spatial geometric weight of the sub-region. The weight value is controlled between 0.3 and 1 to ensure that the sub-region closer to the dominant sub-region has a larger spatial geometric weight.

[0084] Step 3.23 involves receiving the initial feature deviation and corresponding spatial geometric weight of each sub-region and performing weighted fusion calculation to obtain the signal calibration parameters of the original signal of each sub-region. Specifically, this includes: First, receiving the initial feature deviation and corresponding spatial geometric weight of each sub-region, establishing a one-to-one correspondence between the sub-region identifier, initial feature deviation, and spatial geometric weight to ensure data matching accuracy; Next, for each type of blood oxygen signal in each sub-region, processing the initial feature deviation and spatial geometric weight using a weighted fusion method: For each feature dimension difference value in the initial feature deviation, multiplying the corresponding difference value by the spatial geometric weight of the sub-region to obtain the weighted correction deviation value for each feature dimension. For example, if the spatial geometric weight of a sub-region is 0.8 and the initial difference value of a feature dimension is 5, then the weighted correction deviation value is 0.8 × 5 = 4; Then, integrating the weighted correction deviation values ​​of all feature dimensions according to the original feature order to form the corrected deviation set of the blood oxygen signal corresponding to the sub-region. Finally, based on the corrected deviation set and the signal calibration requirements (such as amplitude correction requiring reference to amplitude-related deviations and timing alignment requiring reference to time-related deviations), the corrected deviation set is transformed into specific signal calibration parameters. For example, the amplitude-related weighted correction deviation value is transformed into an amplitude compensation coefficient, and the time-related weighted correction deviation value is transformed into a time delay compensation amount. Ultimately, the complete signal calibration parameters corresponding to the original carotid artery blood oxygenation signal and jugular vein blood oxygenation signal for each sub-region are obtained.

[0085] In this embodiment of the invention, because the feature vector of the dominant sub-region signal is extracted as a reference, the initial feature deviation is obtained by calculating the difference between the feature vector of each sub-region signal and the reference vector; the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated in a two-dimensional coordinate system based on the sub-region identifier, and the spatial geometric weight is obtained by combining the attenuation coefficient based on the length of the semi-major axis of the elliptical region and processing by a predefined mathematical attenuation function; the initial feature deviation of each sub-region and the corresponding spatial geometric weight are weighted and fused to finally generate the signal calibration parameters of the original signal of each sub-region. Therefore, this invention overcomes the technical problem in the prior art where signal calibration only considers the difference in signal features and ignores the influence of the spatial position of the sub-region on the signal quality, resulting in insufficient targeting of the calibration parameters and inability to effectively offset the signal interference caused by the differences in the physiological structure of the neck, thus causing the signal to still have deviation after calibration. This invention achieves the technical effect that the signal calibration parameters can simultaneously take into account the signal feature deviation and spatial position characteristics of the sub-region, improve the accuracy and targeting of the calibration parameters, reduce the interference of the differences in the physiological structure of the neck on the signal quality, and ensure that the calibrated signal is more in line with the real physiological state of arterial and venous blood oxygenation.

[0086] In a preferred embodiment of the present invention, step 4 above may include:

[0087] Step 4.1: Based on the calibrated integrated carotid artery oxygenation signal and integrated jugular venous oxygenation signal, extract the real-time time series of carotid artery oxygen saturation values ​​and jugular venous oxygen saturation values, respectively. Specifically, this includes: First, acquiring the calibrated integrated carotid artery oxygenation signal and calibrated integrated jugular venous oxygenation signal after spatial fusion processing. These two signals have eliminated interference from sub-region signal differences and can stably reflect the true state of carotid artery and venous oxygenation. Next, based on the physiological characteristics and detection principle of oxygenation signals, extract the carotid artery oxygen saturation values ​​corresponding to different times from the integrated carotid artery oxygenation signal. This is done by identifying feature parameters in the signal that are directly related to oxygen saturation. For example, by analyzing the absorption amplitude of light at a specific wavelength and based on a preset mapping relationship between blood oxygen saturation and characteristic parameters (established based on a large amount of physiological experimental data), signal characteristics can be accurately converted into saturation values, yielding specific carotid artery blood oxygen saturation values ​​at each acquisition moment. Similarly, the corresponding carotid venous blood oxygen saturation values ​​at each moment are extracted from the comprehensive jugular venous blood oxygen signal in the same manner. Finally, the extracted carotid artery blood oxygen saturation values ​​are arranged sequentially according to the order of signal acquisition time to form a real-time time series of carotid artery blood oxygen saturation values. Likewise, the jugular venous blood oxygen saturation values ​​are arranged in chronological order to form a real-time time series of jugular venous blood oxygen saturation values.

[0088] Step 4.2 involves time alignment of the carotid artery and jugular vein oxygen saturation time series to eliminate timing deviations caused by different signal acquisition paths. This process includes: First, analyzing the possible causes of timing deviations between the two oxygen saturation time series. Due to the different anatomical locations of the carotid artery and jugular vein in the neck, there are subtle differences in the length of the corresponding sensor acquisition paths and the time it takes for the signal to be transmitted to the processing module. This can lead to inconsistent timestamps for the same actual moment's arterial and venous oxygen saturation values ​​recorded in the two time series. Next, a feature point matching method is used for time alignment: In the carotid artery oxygen saturation time series, multiple consecutive feature points with significant identifiability are selected, such as peak points, trough points, or abrupt change points of the saturation value, and the timestamp corresponding to each feature point is recorded; in the jugular vein... In the time series of arterial oxygen saturation values, corresponding feature points with physiological significance consistent with the characteristic points of the carotid artery sequence are identified, such as the peak saturation point within the same respiratory cycle, and their timestamps are recorded. Then, the time difference between each pair of corresponding feature points is calculated, and the average of multiple time differences is taken as the overall temporal deviation value between the two sequences. If the deviation value is positive, it indicates that the jugular vein sequence lags behind the carotid artery sequence, and the entire jugular vein oxygen saturation value time series is shifted forward by the time length corresponding to the deviation value. If the deviation value is negative, it indicates that the carotid artery sequence lags behind the jugular vein sequence, and the entire carotid artery oxygen saturation value time series is shifted forward by the corresponding time length. Finally, the shifted time series is verified, and the timestamps of the corresponding feature points are checked again to ensure that the saturation values ​​corresponding to the same timestamp in both sequences come from the carotid artery and vein oxygen detection data at the same actual time, thus completely eliminating the temporal deviation.

[0089] Step 4.3 involves subtracting the time-aligned carotid artery oxygen saturation value sequence from the jugular vein oxygen saturation value sequence point by point to calculate the instantaneous arteriovenous oxygen difference at each moment. Specifically, this includes: First, confirming that the time-aligned carotid artery oxygen saturation value time series and the jugular vein oxygen saturation value time series meet the condition that the same timestamp corresponds to the same detection time, and that the length and time interval of the two sequences are completely consistent, such as both containing 86,400 data points, corresponding to data collected once every 0.5 seconds within 24 hours; Next, matching the data in the two sequences one by one according to the timestamps. Starting from the first timestamp, extracting the carotid artery oxygen saturation value and the jugular vein oxygen saturation value corresponding to the timestamp, and subtracting the jugular vein oxygen saturation value from the carotid artery oxygen saturation value to obtain the instantaneous arteriovenous oxygen difference at that moment; Then, processing the second, third, and so on up to the last timestamp, repeating the above subtraction operation to ensure that an instantaneous arteriovenous oxygen difference is calculated for each timestamp.

[0090] In this embodiment of the invention, because the real-time time series of carotid artery oxygen saturation values ​​and jugular vein oxygen saturation values ​​are extracted from the calibrated integrated carotid artery and jugular vein blood oxygen signals respectively, and the two time series are time-aligned to eliminate the timing deviation caused by different acquisition paths, and then the two aligned sequences are subtracted point by point to calculate the instantaneous arteriovenous blood oxygen difference at each moment, the technical problems in the prior art of being unable to effectively extract arteriovenous blood oxygen saturation time series data from accurately calibrated neck region signals, the asynchrony of arteriovenous blood oxygen saturation sequences due to differences in signal acquisition paths, the difficulty in accurately calculating the instantaneous arteriovenous blood oxygen difference, and thus being unable to truly reflect the brain's oxygen metabolism state, are overcome. This achieves the ability to accurately obtain the real-time change time series of arteriovenous blood oxygen saturation, ensuring that the two sequences are completely synchronized in the time dimension, and the calculated instantaneous arteriovenous blood oxygen difference can accurately reflect the brain's oxygen supply and demand balance at each moment.

[0091] In a preferred embodiment of the present invention, step 5 above may include:

[0092] Step 5.1: Receive the arterial-venous oxygen difference value and simultaneously receive the user's current sleep stage information determined by analyzing heart rate variability and body movement signals. Specifically, this includes: First, receiving the calculated instantaneous arterial-venous oxygen difference value, which is continuously input as a real-time data stream. The data transmission frequency is consistent with the previous oxygen signal acquisition frequency to ensure that the oxygen metabolism state at each moment can be captured in a timely manner. Simultaneously, receiving the user's current sleep stage information, which is obtained by first extracting the heart rate signal collected by the biosensor integrated into the flexible neckband, and then parsing the heart rate variability-related indicators from the heart rate signal, such as the standard deviation of adjacent heartbeat intervals and the proportion of adjacent heartbeat intervals with a difference greater than 50 milliseconds. Then, combining this with the body movement signals collected by the miniature accelerometer built into the neckband, it is determined whether the user has turned over and the range of limb movement. Based on the heart rate variability indicators and body movement signals, the user is determined to be in a specific sleep stage, such as light sleep, deep sleep, or REM sleep.

[0093] Step 5.2: Based on the user's current sleep stage information, dynamically select the arterial-venous oxygen difference benchmark reference range and oxygen supply trigger threshold corresponding to the sleep stage. The benchmark reference range and oxygen supply trigger threshold are obtained through personalized modeling and updating using a machine learning model based on the user's historical sleep data. Specifically, this includes: First, retrieving the user's historical sleep database, which stores the user's sleep data for at least the past week, including daily arterial-venous oxygen difference variation curves for each sleep stage, the average oxygen difference for each stage, and the normal fluctuation range of the oxygen difference; then, starting the pre-trained machine learning model and applying the historical sleep data... The sleep stage in the data is used as the input feature, and the corresponding normal fluctuation range of arteriovenous oxygen difference is used as the output label. The model is trained in a personalized way so that it can learn the reasonable range of arteriovenous oxygen difference for the user in different sleep stages, i.e., the baseline reference range, and the critical value for initiating oxygen supply, i.e., the oxygen supply trigger threshold. After the model is trained, it will be updated regularly according to the user's latest sleep data, such as retraining every 7 days, to ensure that the parameters always fit the user's current physiological changes. When the current sleep stage information is obtained, the corresponding baseline reference range and oxygen supply trigger threshold will be called from the trained model based on that stage.

[0094] Step 5.3 compares the real-time acquired arteriovenous oxygen difference data stream with the oxygen supply trigger threshold for the current sleep stage, and performs joint analysis based on the statistical trend characteristics of the arteriovenous oxygen difference data stream within a preset time window to determine if the oxygen supply trigger condition is met. Specifically, this includes: first, setting a preset time window with a duration of 3 minutes; then, comparing the real-time acquired arteriovenous oxygen difference data stream with the oxygen supply trigger threshold corresponding to the current sleep stage hour by hour, recording the number of times and duration the oxygen difference value exceeds the trigger threshold within the 3-minute time window; simultaneously, analyzing the statistical trend characteristics of the oxygen difference data stream within this time window, calculating the average value, maximum value, and slope of the oxygen difference within the window, and determining whether the oxygen difference is continuously higher than the threshold, fluctuates above the threshold, or is briefly higher than the threshold. If the average value is higher than the threshold, the slope is positive or zero, and the duration of exceeding the threshold is more than 2 minutes, then the oxygen supply trigger condition is met; if it only exceeds the threshold once instantaneously, the average value is not higher than the threshold, or the slope is negative, then the trigger condition is not met.

[0095] Step 5.4: After determining that the oxygen supply triggering condition is met, based on the deviation of the real-time arterial-venous oxygen difference from the baseline reference range and the specific sleep stage, the required oxygen supply intensity level is determined through a predefined oxygen supply intensity mapping rule to obtain the final judgment result. Specifically, this includes: First, after determining that the oxygen supply triggering condition is met, calculating the deviation of the real-time arterial-venous oxygen difference from the baseline reference range for the current sleep stage. For example, if the current sleep stage is deep sleep and the upper limit of the baseline reference range is 8%, and the real-time oxygen difference is 10%, then the deviation is (10%-8%)÷8%×100%=25%; if the real-time oxygen difference is 9%, then the deviation is 12.5%. The system calls a predefined oxygen supply intensity mapping rule. The rule sets the intensity level based on both sleep stage and deviation degree. For example, during light sleep, a deviation degree of 10% to 20% corresponds to Level 1 oxygen supply intensity with an oxygen production rate of 1 to 2 liters per minute, while a deviation degree of more than 20% corresponds to Level 2 oxygen supply intensity with an oxygen production rate of 3 to 4 liters per minute. During deep sleep, a deviation degree of 10% to 20% corresponds to Level 2 oxygen supply intensity, while a deviation degree of more than 20% corresponds to Level 3 oxygen supply intensity with an oxygen production rate of 5 to 6 liters per minute. Finally, based on the current sleep stage and the calculated deviation degree, the system finds the corresponding oxygen supply intensity level in the mapping rule, combines this level with the command to start oxygen supply, and forms the final judgment result.

[0096] In this embodiment of the invention, the method involves receiving arteriovenous oxygen difference values ​​and simultaneously receiving information about the user's current sleep stage determined by analyzing heart rate variability and body movement signals. Based on the sleep stage, a reference range for arteriovenous oxygen difference and an oxygen supply trigger threshold are dynamically selected, personalized and updated using a machine learning model based on the user's historical sleep data. The real-time arteriovenous oxygen difference data stream is compared with the current stage threshold, and combined with statistical trend characteristics within a preset time window for joint analysis to determine the oxygen supply trigger condition. Finally, the oxygen supply intensity level is determined according to the degree of deviation of the real-time arteriovenous oxygen difference from the reference and the specific sleep stage, using predefined rules. This technique overcomes the technical problems of existing technologies, such as oxygen supply triggering relying on fixed thresholds, failure to combine sleep stages leading to a disconnect between triggering timing and actual brain oxygen demand, lack of personalized parameter settings resulting in poor adaptability, failure to consider signal trends leading to misjudgment, and lack of correlation between oxygen supply intensity and oxygen demand and sleep state, resulting in intensity mismatch. It achieves the following: oxygen supply trigger judgment can accurately match the brain oxygen demand of users in different sleep stages; personalized parameters improve adaptability to the physiological differences of different users; trend analysis reduces the probability of false triggering; and oxygen supply intensity can match the oxygen demand gap and sleep state as needed, avoiding over-intervention or under-intervention.

[0097] In a preferred embodiment of the present invention, step 6 above may include:

[0098] Step 6.1: Receive the final judgment result and parse the oxygen supply command and oxygen supply intensity level contained in the judgment result. Specifically, this includes: First, receiving the final judgment result, which is transmitted in the form of structured data, containing a command identifier for whether to start oxygen supply and a numerical identifier for the oxygen supply intensity level; Next, decomposing the structured data, first identifying the oxygen supply command field to confirm that the current oxygen supply operation needs to be started. If the judgment result is that oxygen supply is not required, this step is terminated; then, extracting the specific value of the oxygen supply intensity level field, and simultaneously retrieving the pre-stored intensity level and parameter correspondence table to clarify the basic parameters such as the oxygen production rate range and diffuser outlet power corresponding to the level, thus completing the complete parsing of the core information of the judgment result.

[0099] Step 6.2: Based on the analyzed oxygen supply intensity level, obtain the corresponding equipment control signal and release oxygen into the indoor environment according to the set initial oxygen supply intensity. Specifically, this includes: First, matching the corresponding equipment control parameters in the intensity level and parameter correspondence table based on the analyzed oxygen supply intensity level. For example, Level 1 intensity corresponds to an oxygen production rate of 1 to 2 liters per minute and a diffuser air output power of 30%, while Level 2 intensity corresponds to an oxygen production rate of 3 to 5 liters per minute and a diffuser air output power of 60%. Next, converting the parameters into electrical signals recognizable by the diffusion oxygen supply execution module, i.e., equipment control signals, which include specific control commands such as the driving voltage of the oxygen production unit and the speed command of the diffuser fan. Then, sending a control signal to the diffusion oxygen supply module to start the oxygen production unit, such as a silent molecular sieve oxygen production unit and diffuser, to release oxygen into the indoor environment according to the matched initial oxygen supply intensity. At the same time, activating the concentration monitoring sensor built into the module to provide real-time feedback on the current local indoor oxygen concentration, ensuring that the initial oxygen supply intensity meets the set requirements.

[0100] Step 6.3: After oxygen supply is initiated, continuously receive updated arteriovenous oxygen difference data streams and simultaneously monitor the user's real-time sleep stage changes. Specifically, this includes: First, after oxygen supply is initiated, maintain a real-time data connection with the signal acquisition and processing components, continuously receive updated arteriovenous oxygen difference data streams at a frequency of once per second, synchronously record the arteriovenous oxygen difference values ​​at each moment, forming a continuous monitoring curve to track changes in brain oxygen metabolism status in real time; simultaneously, continuously receive heart rate variability data and body movement signals, analyze the fluctuation characteristics of heart rate variability, such as the standard deviation changes of adjacent heartbeat intervals, analyze the activity amplitude of body movement signals, such as whether there are acceleration abrupt changes caused by turning over, and combine with the preset sleep stage algorithm to update the user's current sleep stage in real time, such as from light sleep to deep sleep, or from deep sleep to REM sleep, to ensure that oxygen supply intervention is always synchronized with the user's sleep state.

[0101] Step 6.4: Based on continuously monitored arteriovenous oxygen difference data streams and real-time sleep stages, dynamically evaluate the actual effect of oxygen supply intervention and determine whether the conditions for adjusting oxygen supply intensity or stopping oxygen supply are met to obtain the judgment results. Specifically, this includes: First, based on continuously monitored arteriovenous oxygen difference data streams and real-time sleep stages, directly analyze the actual effect of oxygen supply intervention. On the one hand, observe the changing trend of arteriovenous oxygen difference. If the value gradually decreases from above the threshold to the baseline reference range, it indicates that oxygen supply is effective; if the value remains above the threshold or shows an upward trend, it indicates that the current oxygen supply intensity is insufficient; if the value is below the baseline reference range, it indicates that oxygen supply may be excessive; on the other hand... In conjunction with changes in sleep stages, if a user transitions from deep sleep to REM sleep, the brain's oxygen demand decreases, requiring a reassessment of the necessity for oxygen supply. Next, based on the analysis results, it is determined whether the conditions for adjustment or cessation are met: Conditions for adjusting oxygen supply intensity include an arteriovenous oxygen difference consistently exceeding the baseline by 10% without a downward trend, in which case the intensity needs to be increased; or an arteriovenous oxygen difference dropping to near the lower limit of the baseline and still decreasing, in which case the intensity needs to be decreased. Conditions for stopping oxygen supply include an arteriovenous oxygen difference remaining stable within the baseline reference range for more than 2 minutes, or transitioning from non-REM sleep to REM sleep. The final judgment is derived through a combined assessment.

[0102] Step 6.5: Based on the judgment result, execute the corresponding operation. If the oxygen supply intensity adjustment conditions are met, then based on the current value and trend of the arteriovenous oxygen difference and the current sleep stage, generate and execute control instructions to upgrade or downgrade the oxygen supply intensity according to the predefined adjustment strategy. If the oxygen supply cessation conditions are met, generate and execute control instructions to gradually reduce the oxygen supply intensity until it is completely shut off. Specifically, this includes: First, if the judgment result indicates that the oxygen supply intensity adjustment conditions are met, formulate an adjustment strategy based on the current value and trend of the arteriovenous oxygen difference and the sleep stage. For example, if the user is in deep sleep, and the arteriovenous oxygen difference has decreased but is still 15% higher than the baseline and the trend is flat, then... The intensity is upgraded from Level 1 to Level 2. If the user is in a light sleep phase and the arteriovenous oxygen difference has dropped to the lower limit of the baseline by only 5% and is showing a downward trend, the intensity is downgraded from Level 2 to Level 1. A new device control signal is generated according to the strategy and sent to the diffusion oxygen supply component to perform intensity adjustment. If the judgment result is that the oxygen supply stop condition is met, a gradual stop procedure is initiated. First, the current oxygen supply intensity is reduced by one level, such as from Level 2 to Level 1, and this intensity is maintained for 1 minute to monitor whether the indoor oxygen concentration drops steadily. Then, it is reduced to the lowest intensity, such as 0.5 liters per minute, and maintained for 30 seconds. Finally, the oxygen generating unit and diffuser are completely shut down to avoid sudden changes in indoor oxygen concentration that may disturb sleep.

[0103] In this embodiment of the invention, by receiving and parsing the oxygen supply command and oxygen supply intensity level in the final judgment result, generating a device control signal based on the intensity level to set the initial oxygen supply intensity for releasing oxygen; continuously receiving updated arterial and venous blood oxygen difference data streams after oxygen supply is started and synchronously monitoring the user's real-time sleep stage changes, dynamically evaluating the oxygen supply effect based on the monitoring data, and determining whether the intensity adjustment or cessation conditions are met; and then executing oxygen supply intensity upgrade and downgrade commands according to a predefined strategy based on the judgment result, generating commands to gradually reduce the intensity until it is turned off, the invention overcomes the technical problems of fixed oxygen supply intensity that cannot adapt to the user's real-time brain oxygen demand changes, lack of continuous monitoring and dynamic adjustment leading to excessive or insufficient oxygen supply, lack of buffer when oxygen supply stops easily causing sudden changes in environmental oxygen concentration that interfere with sleep, and disconnection between oxygen supply intervention and real-time sleep stage and blood oxygen changes. This achieves the technical effect of dynamically adapting oxygen supply intensity to the user's brain oxygen demand and sleep stage, avoiding resource waste and excessive intervention, ensuring sleep continuity through the buffer process when oxygen supply stops, and forming a complete closed loop of oxygen supply, monitoring, evaluation, and adjustment, thereby improving the accuracy and safety of oxygen supply intervention and supporting improved sleep quality.

[0104] like Figure 2 As shown, embodiments of the present invention also provide a smart oxygenation sleep engine device based on bioelectrical signals, comprising:

[0105] The acquisition module is used to collect the carotid artery blood oxygenation signal and jugular vein blood oxygenation signal corresponding to the user's carotid artery surface projection point, the midpoint of the anterior border of the sternocleidomastoid muscle, and the center of the supraclavicular fossa, respectively.

[0106] The segmentation module is used to obtain an elliptical region of the neck region based on the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal, with the carotid artery surface projection point as the center of the ellipse, the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis of the ellipse, and the length of the minor axis of the ellipse adjusted according to the actual size of the neck region to obtain the elliptical region of the neck region. The elliptical region is then divided by the inscribed orthogonal lines of the ellipse to obtain multiple fan-shaped sub-regions.

[0107] The calibration module is used to generate calibration parameters based on the signal characteristics of each sub-region; the collected carotid artery blood oxygenation signal and jugular vein blood oxygenation signal are calibrated using the calibration parameters to obtain the calibrated carotid artery blood oxygenation signal and jugular vein blood oxygenation signal.

[0108] The calculation module is used to calculate the arteriovenous oxygen difference in real time based on the calibrated carotid artery blood oxygen signal and jugular vein blood oxygen signal;

[0109] The judgment module is used to determine whether the oxygen supply triggering conditions are met based on the arteriovenous blood oxygen difference and the user's current sleep stage, and to determine the oxygen supply intensity when the conditions are met, so as to obtain the judgment result;

[0110] The processing module is used to perform diffused oxygen supply according to the determined oxygen supply intensity when the oxygen supply triggering conditions are met, release oxygen into the indoor environment where the user is located, continuously monitor the changes in arterial and venous blood oxygen difference, and adjust the oxygen supply intensity or shut down the oxygen supply according to the changes.

[0111] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A smart oxygenation sleep engine device based on bioelectrical signals, characterized in that, include: The acquisition module is used to determine three key acquisition locations based on the anatomical features of the neck. Through a multi-sensor array on a flexible neck strap, the optical sensing probe and bioelectrical impedance electrode are precisely aligned in the corresponding areas. Multi-channel signal acquisition is started simultaneously to acquire arterial and venous signal sources respectively. The raw signals are then subjected to real-time quality assessment and parameter adaptive adjustment. Finally, the quality-verified and accurately positioned multi-channel carotid artery and jugular vein blood oxygenation signals are output. The segmentation module is used to obtain an elliptical region of the neck contour based on carotid artery and jugular venous blood oxygenation signals. The ellipse is centered on the carotid artery surface projection point, with the midpoint of the anterior border of the sternocleidomastoid muscle and the center of the supraclavicular fossa as the endpoints of the major axis. The minor axis length is adjusted according to the actual size of the neck region. This elliptical region is then divided into multiple fan-shaped sub-regions using inscribed orthogonal lines. These sub-regions include: Based on an established two-dimensional coordinate system, the carotid artery surface projection point is determined as the center of the ellipse. Using this center as a reference, the distance to the center of the ellipse is calculated based on the coordinates of the midpoint of the anterior border of the sternocleidomastoid muscle. The distance to the heart is set as the length of the major semi-axis of the ellipse. Based on the center of the ellipse and the length of the major semi-axis, the projection distance to the center of the ellipse is calculated according to the projection position of the center of the supraclavicular fossa on the minor axis of the two-dimensional coordinate system. The projection distance is then weighted and corrected in combination with the physiological parameters of the user's anteroposterior diameter of the neck, and the length of the minor semi-axis of the ellipse is dynamically determined, thereby constructing an elliptical detection area that fully covers the carotid artery and jugular vein regions. Based on the constructed elliptical detection area, with the center of the ellipse as the origin of the coordinate system, the elliptical region is divided into multiple rectangular sub-regions by equally spaced orthogonal grid lines parallel to the major and minor axes of the ellipse. The calibration module is used to generate calibration parameters based on the signal characteristics of each sub-region; the collected carotid artery blood oxygenation signal and jugular vein blood oxygenation signal are calibrated using the calibration parameters to obtain the calibrated carotid artery blood oxygenation signal and jugular vein blood oxygenation signal. The calculation module is used to calculate the arteriovenous oxygen difference in real time based on the calibrated carotid artery blood oxygen signal and jugular vein blood oxygen signal; The judgment module is used to determine whether the oxygen supply triggering conditions are met based on the arteriovenous blood oxygen difference and the user's current sleep stage, and to determine the oxygen supply intensity when the conditions are met, so as to obtain the judgment result; The processing module is used to perform diffused oxygen supply according to the determined oxygen supply intensity when the oxygen supply triggering conditions are met, release oxygen into the indoor environment where the user is located, continuously monitor the changes in arterial and venous blood oxygen difference, and adjust the oxygen supply intensity or shut down the oxygen supply according to the changes.

2. The intelligent oxygenation sleep engine device based on bioelectrical signals according to claim 1, characterized in that, Calibration parameters are generated based on the signal characteristics of each sub-region; the acquired carotid artery and jugular vein oxygenation signals are calibrated using these parameters to obtain calibrated carotid artery and jugular vein oxygenation signals, including: Based on the division of multiple sub-regions and region identifiers, the time-domain and frequency-domain features of the carotid artery blood oxygenation signal and the jugular vein blood oxygenation signal collected in each sub-region are extracted respectively. By comparing the signal-to-noise ratio and stability of the signals in each sub-region, the dominant sub-region used for benchmark reference is determined. Based on the signal characteristics of the dominant sub-region, the characteristic deviation of each sub-region signal relative to the dominant sub-region signal is calculated. In a two-dimensional coordinate system, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated. Using the length of the semi-major axis of the elliptical region as a reference, the attenuation coefficient is used to calculate the geometric weight of each sub-region through a predefined mathematical attenuation function. The geometric weight is combined with the characteristic deviation to obtain the signal calibration parameters corresponding to each sub-region. The signal calibration parameters are applied to the corresponding sub-regions, and the original carotid artery blood oxygenation signal and jugular vein blood oxygenation signal collected in each sub-region are subjected to amplitude correction and time alignment processing to obtain the calibrated blood oxygenation signal of each sub-region. Based on the calibrated blood oxygenation signals of each sub-region, spatial fusion processing is performed according to the region identifier to obtain the calibrated integrated carotid artery blood oxygenation signal and integrated jugular vein blood oxygenation signal, respectively.

3. The intelligent oxygenation sleep engine device based on bioelectrical signals according to claim 2, characterized in that, Based on the signal characteristics of the dominant sub-region, the characteristic deviation of each sub-region signal relative to the dominant sub-region signal is calculated. In the two-dimensional coordinate system, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated. The geometric weight of each sub-region is calculated using the attenuation coefficient with the length of the major semi-axis of the elliptical region as a reference and a predefined mathematical attenuation function. The geometric weights are combined with the feature bias to obtain the signal calibration parameters corresponding to each sub-region, including: The feature vector of the dominant sub-region signal is extracted as a reference. Based on the feature vector reference, the difference between the feature vector of each sub-region signal and the reference vector is calculated in turn to obtain the initial feature deviation of each sub-region. Based on the identifiers of each sub-region, the Euclidean distance between the center point of each sub-region and the center point of the dominant sub-region is calculated sequentially in a two-dimensional coordinate system. The calculated Euclidean distance is then used in conjunction with an attenuation coefficient determined based on the length of the semi-major axis of the elliptical region. This attenuation is then processed by a predefined mathematical attenuation function to obtain the spatial geometric weight corresponding to each sub-region. The initial feature deviation of each sub-region and the spatial geometric weight of the corresponding sub-region are received and weighted and fused to obtain the signal calibration parameters of the original signal of each sub-region.

4. The intelligent oxygenation sleep engine device based on bioelectrical signals according to claim 3, characterized in that, Based on the calibrated carotid artery and jugular venous oxygenation signals, the arteriovenous oxygen difference is calculated in real time, including: Based on the calibrated integrated carotid artery blood oxygenation signal and integrated jugular vein blood oxygenation signal, the real-time time series of carotid artery blood oxygen saturation value and jugular vein blood oxygen saturation value were extracted respectively. Time alignment processing was performed on the time series of carotid artery blood oxygen saturation values ​​and jugular vein blood oxygen saturation values ​​to eliminate the timing deviation caused by different signal acquisition paths, so as to obtain the result that eliminates the timing deviation that may be caused by different signal acquisition paths. The instantaneous arterial and venous oxygen difference at each moment is calculated by subtracting the time-aligned carotid artery oxygen saturation value sequence from the jugular vein oxygen saturation value sequence point by point.

5. The intelligent oxygenation sleep engine device based on bioelectrical signals according to claim 4, characterized in that, Based on the arteriovenous oxygen difference and the user's current sleep stage, it is determined whether the oxygen supply triggering conditions are met, and if the conditions are met, the oxygen supply intensity is determined to obtain the judgment result, including: It receives arterial and venous blood oxygen difference values ​​and simultaneously receives information on the user's current sleep stage determined by analyzing heart rate variability and body movement signals; Based on the user's current sleep stage information, the reference range of arterial and venous blood oxygen difference and the oxygen supply trigger threshold corresponding to the sleep stage are dynamically selected; the reference range and the oxygen supply trigger threshold are obtained by personalized modeling and updating based on the user's historical sleep data through machine learning model; The real-time acquired arteriovenous blood oxygen difference data stream is compared with the oxygen supply trigger threshold under the current sleep stage, and a joint analysis is performed based on the statistical trend characteristics of the arteriovenous blood oxygen difference data stream within a preset time window to obtain the oxygen supply trigger conditions. After determining that the oxygen supply triggering conditions are met, the required oxygen supply intensity level is determined based on the deviation of the real-time arterial and venous blood oxygen difference from the benchmark reference range and the specific sleep stage, through a predefined oxygen supply intensity mapping rule, in order to obtain the final judgment result.

6. The intelligent oxygenation sleep engine device based on bioelectrical signals according to claim 5, characterized in that, Based on the judgment results, when the oxygen supply triggering conditions are met, diffused oxygen supply is executed according to the determined oxygen supply intensity, releasing oxygen into the user's indoor environment. Changes in the arteriovenous blood oxygen difference are continuously monitored, and the oxygen supply intensity is adjusted or shut off based on these changes, including: Receive the final judgment result and parse the oxygen supply instructions and oxygen supply intensity level contained in the judgment result; Based on the oxygen supply intensity level obtained from the analysis, the corresponding equipment control signal is obtained, and oxygen is released into the indoor environment according to the set initial oxygen supply intensity. After oxygen supply is activated, it continuously receives updated arterial and venous blood oxygen difference data streams and simultaneously monitors the user's real-time sleep stage changes. Based on the continuous monitoring of arteriovenous blood oxygen difference data stream and real-time sleep stages, the actual effect of oxygen supply intervention is dynamically evaluated, and it is determined whether the conditions for adjusting oxygen supply intensity or stopping oxygen supply are met, so as to obtain the judgment result. Based on the judgment result, the corresponding operation is executed. If the conditions for adjusting the oxygen supply intensity are met, the control command for upgrading or downgrading the oxygen supply intensity is generated and executed according to the current value and trend of the arteriovenous blood oxygen difference and the current sleep stage, in accordance with the predefined adjustment strategy. If the conditions for stopping the oxygen supply are met, the control command for gradually reducing the oxygen supply intensity until it is completely shut down is generated and executed.

Images