A sports armband monitoring system and method

Abstract

The present invention discloses a sports armband monitoring system and method, comprising the following steps: setting an initial respiratory rate detection period based on the user's normal exercise intensity and physiological needs, establishing a standardized baseline to ensure routine monitoring without disturbing the user's normal activities; during the respiratory rate detection, the blood oxygen level is monitored in real time through a sensor. By setting a personalized initial respiratory rate detection period, monitoring according to the user's exercise intensity and physiological needs to ensure no interference with exercise, after quantifying the trend of blood oxygen change, using a support vector machine regression model to predict future trends, and dynamically adjusting the detection period through a clustering analysis algorithm to flexibly adapt to the user's physiological changes, it can not only accurately identify the trend of blood oxygen decline in real time, but also improve the system response sensitivity by intelligently adjusting the detection period, ensuring timely intervention when the user is at risk of hypoxia and effectively preventing health crises such as acute hypoxia.

Description

Technical Field

[0001] The present invention relates to the technical field of sports armband monitoring, and particularly to a sports armband monitoring system and method. Background Art

[0002] A sports armband monitoring system is a wearable device, usually worn on the upper arm, for real-time monitoring of users' physiological data, motion states, and health indicators. Such systems usually integrate sensors, such as heart rate sensors, accelerometers, gyroscopes, etc., which can collect key data during exercise, such as heart rate, step frequency, step amplitude, exercise intensity, calorie consumption, etc. These data are transmitted to a smartphone or other terminal devices via wireless connection (such as Bluetooth) for analysis and storage. By analyzing users' exercise data, the sports armband monitoring system can provide personalized health advice, give real-time feedback on exercise effects, help optimize exercise plans, and improve exercise effects. In addition, some high-end systems also have functions such as real-time positioning, fatigue monitoring, and environmental adaptability, enhancing their diversity and richness of application scenarios.

[0003] The function of a sports armband monitoring system for monitoring blood oxygen saturation (SpO2) is to track and evaluate in real time the oxygen content in the user's blood during exercise, helping to determine whether the body can maintain sufficient oxygen supply during exercise. Blood oxygen saturation is an important indicator reflecting the health of the heart, lungs, and blood vessels. Especially in high-intensity exercise or high-altitude environments, monitoring blood oxygen levels can effectively prevent hypoxemia and overexertion, ensuring that the body's oxygen needs are met. If the blood oxygen saturation is too low, it may mean insufficient oxygen supply, which may lead to a decline in exercise performance, shortness of breath, increased fatigue, or symptoms such as dizziness. Therefore, the sports armband system can monitor blood oxygen saturation in real time, provide early warning information, help users adjust exercise intensity or take rest measures, thereby improving exercise safety and effects and avoiding the risks brought by overtraining.

[0004] The existing technologies have the following deficiencies:

[0005] When a sports armband monitoring system monitors blood oxygen saturation, it usually cannot intelligently adjust the detection period of the breathing frequency according to the trend of blood oxygen change. When the blood oxygen saturation drops, the breathing frequency will change rapidly as an immediate reaction of the body to the low-oxygen state. If the detection period of the breathing frequency is too long, the system may not be able to capture these changes in time, resulting in the inability to identify the low-oxygen risk in time. In this case, the system cannot adjust the exercise intensity in time or recommend rest, and users may continue to exercise in a continuous low-oxygen state, further exacerbating the insufficient oxygen supply, and ultimately may trigger acute hypoxia, seriously endangering health and even life.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide a sports armband monitoring system and method. By setting a personalized initial breathing frequency detection period, monitoring according to the user's exercise intensity and physiological needs to ensure no interference with exercise, obtaining blood oxygen saturation data in real-time synchronization, and eliminating interference through data preprocessing technology to ensure data accuracy. After quantifying the blood oxygen change trend, using a support vector machine regression model to predict the future trend, and dynamically adjusting the detection period through a clustering analysis algorithm to flexibly adapt to the user's physiological changes. This method can not only accurately identify the blood oxygen decline trend in real-time, but also improve the system response sensitivity by intelligently adjusting the detection period, ensuring timely intervention when the user is at risk of hypoxia, and effectively preventing health crises such as acute hypoxia, so as to solve the problems in the above background art.

[0008] To achieve the above object, the present invention provides the following technical solution: A sports armband monitoring method, including the following steps:

[0009] Set an initial breathing frequency detection period based on the user's normal exercise intensity and physiological needs, and establish a standardized baseline to ensure routine monitoring without interfering with the user's normal activities;

[0010] During the breathing frequency detection, real-time monitor the blood oxygen level through a sensor, synchronously obtain the user's blood oxygen saturation data, so as to capture the subtle changes in blood oxygen saturation;

[0011] Preprocess the obtained blood oxygen saturation data to eliminate measurement errors and external environmental interference, and ensure data accuracy;

[0012] After the data preprocessing is completed, extract the features reflecting the decline of blood oxygen saturation from it, input the extracted blood oxygen saturation change features into the monitoring window for real-time analysis, and quantify the decline trend of blood oxygen saturation;

[0013] After quantifying the decline trend of blood oxygen saturation, input the quantified features into a pre-trained support vector machine regression model, and predict the future trend of blood oxygen saturation decline through the support vector machine regression model;

[0014] According to the future trend of blood oxygen saturation decline predicted by the support vector machine regression model, adopt a clustering analysis algorithm to intelligently adjust the actual breathing frequency detection period, flexibly adapt to different physiological changes, and then efficiently capture the dynamic changes of the breathing frequency.

[0015] Preferably, the specific steps for setting the initial breathing frequency detection period based on the user's normal exercise intensity and physiological needs are as follows:

[0016] First, collect the user's physiological data and normal exercise intensity;

[0017] Next, combine the collected data and determine a suitable initial detection period for the user by analyzing the breathing frequency change trend of the user at different exercise intensities;

[0018] Finally, make personalized adjustments according to the initial detection period.

[0019] Preferably, the specific steps for real-time monitoring of the blood oxygen level by a sensor and synchronously obtaining the user's blood oxygen saturation data are as follows:

[0020] First, accurately install the sensor on the user's skin surface to ensure that the contact between the sensor and the user's blood vessels meets the requirements;

[0021] Then, measure the oxygen content in the blood of the blood vessels through the sensor and calculate the blood oxygen saturation in real time.

[0022] Preferably, extract the features reflecting the decrease in blood oxygen saturation. Among them, the extracted features include the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation attenuation rate. After obtaining them, input the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation attenuation rate into the monitoring window for real-time analysis, and generate a blood oxygen expected deviation reference value and a blood oxygen attenuation reference value respectively, and quantify the decrease trend of the blood oxygen saturation through the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value.

[0023] Preferably, the specific steps for inputting the deviation between the current blood oxygen saturation and the expected value into the monitoring window for real-time analysis to generate a blood oxygen expected deviation reference value are as follows:

[0024] Calculate the absolute deviation value according to the deviation between the current blood oxygen saturation and the expected blood oxygen saturation. The calculation expression is:

[0025]

[0026] Where: is the absolute deviation value between the blood oxygen saturation at time t and the expected blood oxygen saturation, is the blood oxygen saturation monitored by the user at time t, is the expected blood oxygen saturation value set at time t;

[0027] To quantify the downward trend of blood oxygen saturation, the absolute deviation values at each moment within the monitoring window will be weighted and summed to generate a reference value for the expected deviation of blood oxygen. The weighted summation process takes into account the influence of historical deviations on the current deviation and dynamically adjusts the weights of historical data by introducing an exponentially weighted decay coefficient. The calculation expression is as follows:

[0028]

[0029] Where: is the reference value for the expected deviation of blood oxygen, representing the quantified downward trend of blood oxygen within the monitoring window, is the reference value for the expected deviation of blood oxygen at the previous moment, reflecting the influence of historical blood oxygen changes, is the size of the monitoring window, representing the time range of concern, is within the monitoring window i is the absolute deviation value between the blood oxygen saturation at time and the expected value, is the exponentially weighted decay factor, where

[0030] Preferably, under the monitoring window, the specific steps for analyzing the decay rate of blood oxygen saturation to generate a reference value for blood oxygen decay are as follows:

[0031] Real-time monitor the blood oxygen saturation data in the monitoring window and calculate its change rate within each time step. The decay rate of blood oxygen saturation is defined as the instantaneous decline rate of blood oxygen saturation. The calculation expression is:

[0032]

[0033] Where, represents the decay rate of blood oxygen saturation at time t, and are the blood oxygen saturation data at time t and time respectively, is the time interval, and the change trend of the decay rate is captured by calculating the logarithmic change of blood oxygen saturation, reflecting the rapid decline of blood oxygen saturation;

[0034] To more accurately quantify the downward trend of blood oxygen saturation, a weighted exponential decay model is introduced. Based on the instantaneous change rate and the decay weighting factor at the decay moment, a reference value for blood oxygen decay is generated. The expression for generating the reference value for blood oxygen decay is:

[0035]

[0036] Where, is the reference value for blood oxygen decay, is within the monitoring window The instantaneous change rate within indicates the offset from the current time t to the past by time steps, where n is the decay parameter used to control the weight of past decay information, and

[0037] Preferably, after quantitatively analyzing the trend of oxygen saturation decline, the quantified oxygen saturation expected deviation reference value and the oxygen saturation decay reference value are input into a pre-trained support vector machine regression model. Based on the model output of the decline trend index, the future trend of oxygen saturation decline is predicted through the decline trend index.

[0038] Preferably, according to the future trend of oxygen saturation decline predicted by the support vector machine regression model, a clustering analysis algorithm is used to intelligently adjust the actual respiratory rate detection period. The specific steps are as follows:

[0039] After obtaining the decline trend index, a clustering analysis algorithm is used to classify the trend of oxygen saturation decline. The clustering divides the data into several categories according to the rate of oxygen saturation decline, and each category represents a different oxygen saturation change state. The specific classification formula is as follows:

[0040]

[0041] where represents the k th cluster, is the range of the decline trend index of the k th cluster, is the decline trend index at time t, K is the total number;

[0042] Set the first decline trend index threshold and the second decline trend index threshold to divide the decline trend index into three categories, where the first decline trend index is greater than the second decline trend index threshold, and the decline trend index , and the specific division steps are

[0043] If the decline trend index is greater than the first decline trend index , then the trend of oxygen saturation decline is classified as ;

[0044] If the decline trend index is greater than the second decline trend index threshold and less than the first decline trend index , then the trend of oxygen saturation decline is classified as ;

[0045] If the decline trend index is less than the second decline trend index threshold, the trend of blood oxygen saturation decline is classified as ;

[0046] After obtaining the blood oxygen change state, the actual respiratory rate detection period is intelligently adjusted according to different blood oxygen attenuation trends. The adjustment process is as follows:

[0047] If the blood oxygen drops rapidly, shorten the respiratory rate detection period to more frequently monitor the change of the user's respiratory rate. The adjustment formula is:

[0048]

[0049] where is the initial respiratory rate detection period, is the adjustment coefficient, is the minimum respiratory rate detection period, is the adjusted actual respiratory rate detection period;

[0050] If the blood oxygen drops slowly, slightly extend the detection period. The adjustment formula for the actual respiratory rate detection period is:

[0051]

[0052] where is the extension coefficient;

[0053] If the blood oxygen is in a stable state, maintain the initial respiratory rate detection period. The formula is: .

[0054] A sports armband monitoring system includes an initial detection period setting module, a blood oxygen data real-time monitoring module, a data preprocessing and error elimination module, a blood oxygen change feature extraction and analysis module, a blood oxygen decline trend prediction module, and a respiratory rate detection period adjustment module;

[0055] The initial detection period setting module sets an initial respiratory rate detection period based on the user's normal exercise intensity and physiological needs, and establishes a standardized baseline to ensure routine monitoring without disturbing the user's normal activities;

[0056] The blood oxygen data real-time monitoring module, during the respiratory rate detection, real-time monitors the blood oxygen level through a sensor and synchronously obtains the user's blood oxygen saturation data to be able to capture the subtle changes in blood oxygen saturation;

[0057] The data preprocessing and error elimination module preprocesses the obtained blood oxygen saturation data, eliminates measurement errors and external environmental interferences, and ensures the accuracy of the data;

[0058] The blood oxygen change feature extraction and analysis module extracts features reflecting the decrease in blood oxygen saturation after the data preprocessing is completed, inputs the extracted blood oxygen saturation change features into the monitoring window for real-time analysis, and quantifies the decreasing trend of blood oxygen saturation.

[0059] The blood oxygen decreasing trend prediction module, after quantitatively analyzing the decreasing trend of blood oxygen saturation, inputs the quantified features into a pre-trained support vector machine regression model, and predicts the future trend of the decrease in blood oxygen saturation through the support vector machine regression model.

[0060] The respiratory rate detection period adjustment module, according to the future trend of the decrease in blood oxygen saturation predicted by the support vector machine regression model, uses the clustering analysis algorithm to intelligently adjust the actual respiratory rate detection period, flexibly adapts to different physiological changes, and then efficiently captures the dynamic changes of the respiratory rate.

[0061] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0062] The present invention first sets an initial respiratory rate detection period according to the user's normal exercise intensity and physiological needs, establishes a personalized monitoring baseline, and ensures effective monitoring without disturbing the user's normal exercise. Secondly, during the monitoring process, the system synchronously obtains blood oxygen saturation data in real time and combines data preprocessing technology to eliminate external interference and measurement errors, thereby ensuring the high precision and reliability of the data. Based on the quantitative analysis of the blood oxygen change trend, a pre-trained support vector machine regression model is used to predict the future blood oxygen change trend, and further, the clustering analysis algorithm is used to dynamically adjust the respiratory rate detection period, enabling the system to flexibly adapt to the user's physiological changes and timely capture the dynamic changes of the respiratory rate. It can not only identify the decreasing trend of blood oxygen saturation in real time and accurately, but also improve the reaction sensitivity of the system by intelligently adjusting the detection period, ensure that the user can obtain intervention suggestions in time when the low oxygen risk appears, and then effectively prevent health crises such as acute hypoxia, and ensure that the user always exercises in a safe physiological state. Description of the Drawings

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings according to these drawings.

[0064] Figure 1 It is a method flow chart of a motion armband monitoring method of the present invention.

[0065] Figure 2 This is a schematic diagram of the modules of a motion armband monitoring system according to the present invention. Specific embodiments

[0066] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that this disclosure will be more complete and thorough, and will fully convey the concept of the example embodiments to those skilled in the art.

[0067] The present invention provides a Figure 1 motion armband monitoring method as shown, including the following steps:

[0068] Based on the user's normal exercise intensity and physiological needs, set an initial respiratory rate detection period, establish a standardized baseline, and ensure routine monitoring without disturbing the user's normal activities.

[0069] The specific steps for setting the initial respiratory rate detection period based on the user's normal exercise intensity and physiological needs are as follows:

[0070] First, collect the user's basic physiological data (such as age, gender, weight, health status, etc.) and normal exercise intensity (such as daily activity level, exercise type and intensity). Then, combine these data and determine an initial detection period suitable for the user by analyzing the trend of respiratory rate changes at different exercise intensities. This period should be able to effectively reflect the respiratory rate changes of the user during normal activities while avoiding overly frequent measurements from interfering with the user's daily exercise. Finally, according to the initial detection period, make personalized adjustments. If the user's exercise intensity or physiological state changes (such as increased exercise intensity or physical condition changes), fine-tuning can be performed through subsequent feedback mechanisms to ensure the continuous accuracy and adaptability of the monitoring. The core of this process is to ensure that the initial detection period can provide sufficient monitoring data without affecting the user's normal activities.

[0071] During the respiratory rate detection period, the blood oxygen level is monitored in real time through a sensor, and the user's blood oxygen saturation data is synchronously obtained to be able to capture the subtle changes in blood oxygen saturation.

[0072] These data will be used for analysis and further processing so that intelligent responses can be made according to blood oxygen changes during the detection process. Real-time acquisition of blood oxygen saturation data is the core for the system to evaluate the user's physiological state. Blood oxygen saturation directly reflects the oxygen supply state in the user's body. Especially in strenuous exercise or high-altitude environments, changes in blood oxygen saturation have a profound impact on health. Through real-time data collection, the system can capture rapidly changing blood oxygen values and timely determine whether it is necessary to adjust the detection frequency or intervene in the exercise intensity.

[0073] The specific steps to continuously monitor the blood oxygen level through a sensor and synchronously obtain the user's blood oxygen saturation data are as follows:

[0074] First, select a suitable sensor (such as a pulse oximeter sensor or an optical sensor) and accurately install it on the user's skin surface (such as fingers, earlobes, or wrists) to ensure good contact between the sensor and the user's blood vessels. Then, the sensor starts to measure the oxygen content in the blood passing through the blood vessels through optical principles (such as the differences in reflection and absorption of infrared light and visible light), and calculates the blood oxygen saturation (SpO2) in real time. During the entire monitoring process, the sensor continuously collects data and transmits these real-time blood oxygen saturation data to the monitoring system through wireless or wired communication methods. The system continuously monitors and analyzes these data, captures the subtle changes in blood oxygen saturation, and uses them for subsequent health assessments and risk predictions. This process ensures the real-time and accuracy of blood oxygen saturation data, providing a key basis for the user's health monitoring.

[0075] Preprocess the obtained blood oxygen saturation data to eliminate measurement errors and external environmental interferences, ensuring the accuracy of the data;

[0076] The preprocessing steps include denoising, smoothing the data, removing outliers, and normalization processing, etc. The purpose is to eliminate measurement errors or external environmental interferences and ensure the accuracy of the data. The preprocessed data will be more suitable for feature extraction and subsequent analysis. Data preprocessing is a key step to ensure the analysis accuracy of the system. The data without preprocessing may contain noise or errors, affecting the accuracy of subsequent feature extraction. By cleaning and normalizing the blood oxygen data, stable and reliable data can be obtained, ensuring that the analysis results and predictions are more credible.

[0077] After the data preprocessing is completed, extract the features reflecting the decrease in blood oxygen saturation, and input the extracted blood oxygen saturation change features into the monitoring window for real-time analysis to quantify the decreasing trend of blood oxygen saturation;

[0078] Extract the features reflecting the decrease in blood oxygen saturation. Among them, the extracted features include the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation attenuation rate. After obtaining them, input the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation attenuation rate into the monitoring window for real-time analysis, and generate the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value respectively. Quantify the decreasing trend of blood oxygen saturation through the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value.

[0079] When the deviation between the current blood oxygen saturation and the expected value is large, it usually indicates a downward trend in the user's blood oxygen saturation. The expected value is typically set based on the user's health status, exercise intensity, and physiological needs, representing a reasonable range of blood oxygen saturation. When the actual blood oxygen saturation deviates significantly from this expected value, it means that the oxygen supply in the blood fails to reach the expected level, which may be caused by insufficient oxygen uptake, decreased respiratory efficiency, or other health problems. A large deviation is often an indication signal of a decrease in blood oxygen saturation, especially when the user is engaged in strenuous exercise or under high-intensity load, and this deviation is more significant. By real-time monitoring the deviation between the blood oxygen saturation and the expected value, the trend of decreasing blood oxygen can be identified early, and the exercise intensity can be adjusted in a timely manner or rest suggestions can be provided.

[0080] The specific steps for inputting the deviation between the current blood oxygen saturation and the expected value into the monitoring window for real-time analysis to generate a reference value for the blood oxygen expected deviation are as follows:

[0081] Calculate the absolute deviation value based on the deviation between the current blood oxygen saturation and the expected blood oxygen saturation. The calculation expression is:

[0082]

[0083] Where: is the absolute deviation value between the blood oxygen saturation and the expected blood oxygen saturation at time t, is the blood oxygen saturation monitored by the user at time t, is the expected value of blood oxygen saturation set according to the user's normal exercise intensity, health status, and other factors at time t;

[0084] This absolute deviation value reflects the degree of deviation of the current blood oxygen saturation from the normal or expected value and is a key indicator for measuring the trend of hypoxia. The larger the absolute deviation value, the lower the blood oxygen saturation and the greater the risk.

[0085] To quantify the downward trend of blood oxygen saturation, the absolute deviation values at each moment will be weighted and summed within the monitoring window to generate a reference value for the blood oxygen expected deviation. The weighted summation process will consider the influence of historical deviations on the current deviation and dynamically adjust the weights of historical data by introducing an exponential weighted decay coefficient. The calculation expression is:

[0086]

[0087] Where: is the reference value for the blood oxygen expected deviation, indicating the quantified downward trend of blood oxygen within the monitoring window, is the reference value for the blood oxygen expected deviation at the previous moment (i.e., the reference value for the blood oxygen expected deviation quantified within the previous monitoring window), reflecting the influence of historical blood oxygen changes, is the size of the monitoring window, representing the time range that the system focuses on. For within the monitoring window i is the absolute deviation value of the blood oxygen saturation at a certain moment from the expected value. is the exponentially weighted decay factor, where is the decay coefficient, controlling the influence degree of the deviation on the current blood oxygen expected deviation reference value over time;

[0088] By weighted summation and decay adjustment of the deviation at each moment within the monitoring window, a dynamically updated blood oxygen expected deviation reference value is generated, thereby realizing the quantitative evaluation of the blood oxygen decline trend. Through the exponentially weighted method, the system can give higher weight to the blood oxygen changes at the most recent moment, ensuring timely response to the blood oxygen decline.

[0089] From the blood oxygen expected deviation reference value, it can be seen that under the monitoring window, the larger the performance value of the blood oxygen expected deviation reference value generated after analyzing the deviation between the current blood oxygen saturation and the expected value, the greater the deviation between the current blood oxygen saturation and the expected value, reflecting a more significant trend of blood oxygen saturation decline. This is because the blood oxygen expected deviation reference value quantifies the blood oxygen deviation at each moment within the monitoring window and is dynamically updated through weighted summation and decay. When the difference between the blood oxygen saturation and the expected value is large, it indicates that the user's blood oxygen level has significantly decreased relative to the expected state, and the deviation value is large, indicating that the risk of hypoxic state is increasing. While when the deviation is small, it means that the blood oxygen saturation is close to the expected value, the decline trend is not obvious, and the hypoxic risk is small. Therefore, the blood oxygen expected deviation reference value can effectively reflect the change trend of blood oxygen saturation, thus providing timely decision-making basis for the monitoring system.

[0090] When the blood oxygen saturation decay rate is large, it usually means that there is an obvious downward trend in the user's blood oxygen saturation. The blood oxygen saturation decay rate refers to the speed or acceleration of the blood oxygen saturation decline. If the decay rate is large, it indicates that the process of blood oxygen reduction is rapid and continuous. This situation usually reflects that the user may be in a hypoxic state and the body is losing the effective ability to absorb oxygen, which may be caused by excessive exercise, respiratory system abnormalities, environmental factors or health problems. The rapid decay of blood oxygen saturation may lead to insufficient oxygen supply, and then trigger emergency reactions such as shortness of breath and increased heart rate. Therefore, when the blood oxygen saturation decay rate is large, it is an important indicator, indicating that timely measures need to be taken.

[0091] Under the monitoring window, the specific steps for analyzing the blood oxygen saturation decay rate to generate the blood oxygen decay reference value are as follows:

[0092] Real-time monitor the blood oxygen saturation data in the monitoring window and calculate its change rate within each time step. The blood oxygen saturation decay rate is defined as the instantaneous decline rate of blood oxygen saturation, and the calculation expression is:

[0093]

[0094] Among them, represents the blood oxygen saturation decay rate at time t, and are the blood oxygen saturation data at time t and time respectively, is the time interval. By calculating the logarithmic change of blood oxygen saturation, the change trend of the decay rate is captured, reflecting the rapid decline of blood oxygen saturation;

[0095] The function of this step is to capture the rapid change of blood oxygen saturation in real time, which serves as the basis for calculating the blood oxygen decay reference value.

[0096] To more precisely quantify the decline trend of blood oxygen saturation, a weighted exponential decay model is introduced. Based on the instantaneous change rate and the decay weighting factor at the decay moment (the weighting factor is determined by the decay situation in the past period), the blood oxygen decay reference value is generated. The expression for generating the blood oxygen decay reference value is:

[0097]

[0098] Among them, is the blood oxygen decay reference value, is the instantaneous change rate within the monitoring window , represents the offset from the current time t to the past time steps, is the decay parameter, which is used to control the weight of past decay information, n is the length of the historical monitoring window considered. The weighted decay model makes the contribution of earlier blood oxygen changes to the current index smaller through exponential decay, which can effectively capture the recent decay trend and reflect the acceleration or deceleration changes in the blood oxygen decline process;

[0099] The function of this step is to quantify the overall trend of blood oxygen decline through the weighted model and adjust the influence of historical information according to the time decay factor.

[0100] From the oxygen saturation attenuation reference value, it can be seen that under the monitoring window, the larger the performance value of the oxygen saturation attenuation rate analysis for generating the oxygen saturation attenuation reference value, the more significant the downward trend of the user's oxygen saturation. Vice versa. The oxygen saturation attenuation reference value quantifies the decline process of oxygen saturation by real-time monitoring the attenuation rate of oxygen saturation and combining it with a weighted exponential decay model. When the value of the oxygen saturation attenuation reference value is large, it means that within the monitoring window, the oxygen saturation has decreased rapidly and significantly, which usually indicates that the user may be in a hypoxic state and there is a serious shortage of oxygen supply. On the contrary, if the oxygen saturation attenuation reference value is small, it means that the decline process of oxygen saturation is relatively gentle, the downward trend is not obvious, the user's oxygen supply is relatively sufficient, and the hypoxic risk is low. Through this quantitative analysis, the oxygen saturation attenuation reference value can help to evaluate the trend of oxygen saturation changes in real time and provide a timely basis for system intervention.

[0101] After quantitatively analyzing the downward trend of oxygen saturation, the quantified features are input into a pre-trained support vector machine regression model to predict the future trend of oxygen saturation decline through the support vector machine regression model;

[0102] After quantitatively analyzing the downward trend of oxygen saturation, the quantified oxygen saturation expected deviation reference value and oxygen saturation attenuation reference value are input into a pre-trained support vector machine regression model. Based on the model output of the decline trend index, the future trend of oxygen saturation decline is predicted through the decline trend index.

[0103] The pre-trained support vector machine regression model refers to the support vector machine (SVM) regression model trained based on historical data (such as parameters like oxygen saturation, oxygen saturation attenuation reference value, oxygen saturation expected deviation reference value, etc.) in the construction of the motion armband monitoring system. In this model, by inputting specific physiological data features (such as oxygen saturation, attenuation rate, etc.) and corresponding labels (such as the trend or specific value of oxygen saturation decline), the system can learn to identify the pattern of oxygen saturation changes from the data and then be used to predict the future trend of oxygen saturation changes. SVM regression is essentially an extension of the support vector machine, aiming to find an optimal hyperplane in the multi-dimensional space to fit the non-linear relationship between the input features and the output target by minimizing the regression error and maximizing the margin of the model (i.e., the distance between the hyperplane and the support vectors).

[0104] Specifically, the training process of the SVM regression model usually includes several steps. First, key features that affect the change in blood oxygen saturation are extracted from historical data. These features may include the blood oxygen attenuation rate, the expected deviation index, etc., and these features are mapped to a high-dimensional space through a suitable kernel function (such as the radial basis kernel function). In the high-dimensional space, the SVM model can handle these non-linear relationships and find an optimal regression function through an optimization process. This function can maintain a low error while avoiding overfitting as much as possible. After training, the SVM regression model can predict new input data, that is, based on the input blood oxygen saturation data and its attenuation trend, it outputs a predicted decline trend index. The decline trend index can reflect the future change trend of blood oxygen saturation, help the system identify hypoxic risks in advance, and adjust the exercise intensity or rest suggestions according to the prediction results to optimize the user's health monitoring and intervention strategies.

[0105] The important advantage of this model is that it can not only handle data features with non-linearity and high dimensions, but also automatically discover the hidden laws in the data through training. In practical applications, the support vector machine regression model usually requires a large amount of historical data for training, and adjusts the model parameters through techniques such as cross-validation to ensure its accuracy and robustness in predicting the change of blood oxygen saturation. By using the SVM regression model, the motion armband monitoring system can effectively predict the future trend based on the real-time detected blood oxygen attenuation situation and historical data, and provide users with more intelligent and real-time health feedback and suggestions.

[0106] The support vector machine regression model is not specifically limited here, as long as it can realize the comprehensive analysis of the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value to generate the decline trend index is acceptable. To implement the technical solution of the present invention, the present invention provides a specific implementation method; the calculation formula for generating the decline trend index is as follows:

[0107]

[0108] In the formula, and are respectively the preset proportionality coefficients of the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value , and and are both greater than 0.

[0109] From the downward trend index, it can be seen that under the monitoring window, the larger the performance value of the blood oxygen expected deviation reference value generated after analyzing the deviation between the current blood oxygen saturation and the expected value, the larger the performance value of the blood oxygen expected deviation reference value generated after analyzing the deviation between the current blood oxygen saturation and the expected value. That is, when predicting the future trend of the decrease in blood oxygen saturation through a pre-trained support vector machine regression model, the larger the performance value of the downward trend index generated, the more significant the downward trend of blood oxygen saturation. On the contrary, it indicates that the downward trend of blood oxygen saturation is less significant.

[0110] According to the future trend of the decrease in blood oxygen saturation predicted by the support vector machine regression model, the clustering analysis algorithm is used to intelligently adjust the actual respiratory rate detection period, flexibly adapt to different physiological changes, and then efficiently capture the dynamic changes of the respiratory rate.

[0111] The specific adjustment process of the actual respiratory rate detection period is as follows: If it is predicted that the blood oxygen saturation decreases rapidly, the actual respiratory rate detection period is reduced to more frequently monitor the changes in the user's respiratory rate; if the blood oxygen changes slowly, the actual respiratory rate detection period is maintained or appropriately extended.

[0112] According to the future trend of the decrease in blood oxygen saturation predicted by the support vector machine regression model, the clustering analysis algorithm is used to intelligently adjust the actual respiratory rate detection period. The specific steps are as follows:

[0113] After obtaining the downward trend index, the clustering analysis algorithm (such as K-means, DBSCAN, etc.) is used to classify the trend of the decrease in blood oxygen saturation. The purpose is to dynamically adjust the respiratory rate detection period based on the speed and pattern of blood oxygen changes. The clustering divides the data into several categories according to the rate of decrease in blood oxygen saturation, and each category represents a different blood oxygen change state. The specific classification formula is as follows:

[0114]

[0115] Among them, represents the k th cluster, is the range of the downward trend index of the k th cluster, is the downward trend index at time t, K is the total number of

[0116] The downward trend index is divided into three categories. Among them, The first downward trend index is greater than the second downward trend index threshold, The downward trend index , the specific division steps are as follows

[0117] If the decline trend index is greater than the first decline trend index , then the trend of the decline in blood oxygen saturation is divided into ;

[0118] If the decline trend index is greater than the second decline trend index threshold and less than the first decline trend index , then the trend of the decline in blood oxygen saturation is divided into ;

[0119] If the decline trend index is less than the second decline trend index threshold, then the trend of the decline in blood oxygen saturation is divided into .

[0120] ;

[0121] ;

[0122] .

[0123] After obtaining the blood oxygen change state, the system intelligently adjusts the actual respiratory rate detection period according to different blood oxygen attenuation trends. The adjustment process is as follows:

[0124] If the blood oxygen is rapidly declining, shorten the respiratory rate detection period to more frequently monitor the change of the user's respiratory rate. The adjustment formula is:

[0125]

[0126] Among them, is the initial respiratory rate detection period, is the adjustment coefficient, is the minimum respiratory rate detection period, is the adjusted actual respiratory rate detection period;

[0127] This formula speeds up the monitoring frequency by reducing the actual respiratory rate detection period to ensure that the change of the respiratory rate can be captured in time when the blood oxygen drops rapidly;

[0128] If the blood oxygen is slowly declining, slightly extend the detection period. The adjustment formula for the actual respiratory rate detection period is:

[0129]

[0130] Among them, is the extension coefficient, which is applicable to the situation where the blood oxygen changes slowly. It can appropriately increase the detection period to avoid unnecessary frequent detections;

[0131] If the blood oxygen is in a stable state, maintain the initial respiratory rate detection period. The formula is: ;

[0132] The function of this adjustment step is to flexibly adjust the respiratory rate detection period according to the downward trend of blood oxygen saturation. Based on the results of cluster analysis, the system can adaptively shorten or extend the monitoring period to respond to changes in blood oxygen status in real time. This helps to more efficiently capture the dynamic changes in respiratory rate, improve the monitoring accuracy of blood oxygen level changes, and ensure that the health risks of users are identified and responded to in a timely manner.

[0133] Through the above-mentioned motion armband monitoring method, it is possible to achieve real-time monitoring of the future trend of blood oxygen saturation decline and intelligent adjustment of the respiratory rate detection period, thereby significantly improving the response speed and accuracy of the system and avoiding the opportunity of missing the timely identification of hypoxic risks due to too long detection periods. First, set the initial detection period based on the user's normal exercise intensity and physiological needs to provide a personalized monitoring baseline to ensure effective monitoring without interfering with normal exercise. Second, obtain blood oxygen saturation data in real-time synchronization and combine data preprocessing techniques to eliminate external interference and measurement errors to ensure high-precision data. After quantifying the blood oxygen change trend, use the trained support vector machine regression model to predict the future blood oxygen change trend, and further dynamically adjust the respiratory rate detection period through the cluster analysis algorithm, enabling the system to flexibly adapt to the physiological changes of the user, capture the dynamic changes in respiratory rate, and avoid the inability to timely identify the downward trend of blood oxygen due to too long detection periods. This solution can not only identify the downward trend of blood oxygen saturation in real time and accurately, but also improve the reaction sensitivity of the motion monitoring system by adjusting the detection period, ensuring that users can receive timely intervention suggestions when hypoxic risks occur, thereby effectively preventing health crises such as acute hypoxia and ensuring that users exercise in a safe physiological state.

[0134] A motion armband monitoring system includes an initial detection period setting module, a blood oxygen data real-time monitoring module, a data preprocessing and error elimination module, a blood oxygen change feature extraction and analysis module, a blood oxygen decline trend prediction module, and a respiratory rate detection period adjustment module;

[0135] The initial detection period setting module sets an initial respiratory rate detection period based on the user's normal exercise intensity and physiological needs, and establishes a standardized baseline to ensure routine monitoring without interfering with the normal activities of the user;

[0136] The blood oxygen data real-time monitoring module monitors the blood oxygen level in real time through a sensor during the respiratory rate detection, synchronously obtaining the user's blood oxygen saturation data so as to be able to capture the subtle changes in blood oxygen saturation;

[0137] The data preprocessing and error elimination module preprocesses the obtained blood oxygen saturation data, eliminates measurement errors and the interference of the external environment, and ensures the accuracy of the data;

[0138] The blood oxygen change feature extraction and analysis module, after the data preprocessing is completed, extracts the features reflecting the decrease in blood oxygen saturation therefrom, inputs the extracted blood oxygen saturation change features into the monitoring window for real-time analysis, and quantifies the decreasing trend of blood oxygen saturation;

[0139] The blood oxygen decreasing trend prediction module, after quantitatively analyzing the blood oxygen saturation decreasing trend, inputs the quantified features into a pre-trained support vector machine regression model, and predicts the future trend of the blood oxygen saturation decrease through the support vector machine regression model;

[0140] The respiratory rate detection period adjustment module, according to the future trend of the blood oxygen saturation decrease predicted by the support vector machine regression model, uses the clustering analysis algorithm to intelligently adjust the actual respiratory rate detection period, flexibly adapts to different physiological changes, and then efficiently captures the dynamic changes of the respiratory rate;

[0141] A method for monitoring a sports armband provided by an embodiment of the present invention is implemented through the above-mentioned sports armband monitoring system. The specific method and process of a sports armband monitoring system are detailed in the embodiment of the above-mentioned sports armband monitoring method, which will not be elaborated here.

[0142] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to obtain a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0143] As mentioned above, only the specific implementation manners of the present application are described, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

[0144] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. A sports armband monitoring method, characterized in that: The following steps are involved: Set an initial respiratory rate detection cycle based on the user's normal exercise intensity and physiological needs to establish a standardized baseline to ensure regular monitoring without interfering with the user's normal activities; During the respiratory rate detection, the sensor monitors the blood oxygen level in real time and simultaneously obtains the user's blood oxygen saturation data so as to capture subtle changes in blood oxygen saturation; Pre-process the acquired blood oxygen saturation data to eliminate measurement errors and interference from the external environment to ensure data accuracy; After the data preprocessing is completed, the features reflecting the decrease in blood oxygen saturation are extracted, and the extracted blood oxygen saturation change features are input into the monitoring window for real-time analysis to quantify the decreasing trend of blood oxygen saturation; After quantitatively analyzing the trend of blood oxygen saturation decrease, the quantified features are input into the pre-trained support vector machine regression model, and the future trend of blood oxygen saturation decrease is predicted by the support vector machine regression model; According to the future trend of decreased blood oxygen saturation predicted by the support vector machine regression model, a cluster analysis algorithm is used to intelligently adjust the actual respiratory rate detection period to flexibly adapt to different physiological changes, thereby efficiently capturing the dynamic changes of respiratory rate.

2. A sports armband monitoring method according to claim 1, characterized in that: The specific steps for setting the initial breathing rate detection cycle based on the user's normal exercise intensity and physiological needs are as follows: First, collect the user's physiological data and normal exercise intensity; Next, based on the collected data, an initial detection cycle suitable for the user is determined by analyzing the changing trend of the user's breathing rate under different exercise intensities; Finally, make personalized adjustments based on the initial detection cycle.

3. A sports armband monitoring method according to claim 1, characterized in that: The specific steps for monitoring blood oxygen levels in real time through sensors and synchronously obtaining the user's blood oxygen saturation data are as follows: First, the sensor is accurately installed on the user's skin surface to ensure that the sensor meets the requirements for contact with the user's blood vessels; Then, the oxygen content in the blood in the blood vessels is measured through sensors, and the blood oxygen saturation is calculated in real time.

4. A sports armband monitoring method according to claim 1, characterized in that: Features reflecting the decrease in blood oxygen saturation are extracted, wherein the extracted features include the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation decay rate. After acquisition, the deviation between the current blood oxygen saturation and the expected value and the blood oxygen saturation decay rate are input into the monitoring window for real-time analysis, and a blood oxygen expected deviation reference value and a blood oxygen attenuation reference value are generated respectively. The decreasing trend of blood oxygen saturation is quantified by the blood oxygen expected deviation reference value and the blood oxygen attenuation reference value.

5. A sports armband monitoring method according to claim 4, characterized in that: The specific steps of inputting the deviation between the current blood oxygen saturation and the expected value into the monitoring window for real-time analysis to generate the blood oxygen expected deviation reference value are as follows: The absolute deviation value is calculated based on the deviation between the current blood oxygen saturation and the expected blood oxygen saturation. The calculation expression is: in: is the absolute deviation between the blood oxygen saturation at time t and the expected blood oxygen saturation, is the blood oxygen saturation monitored by the user at time t, The expected value of blood oxygen saturation set at time t; In order to quantify the downward trend of blood oxygen saturation, the absolute deviation value at each moment in the monitoring window is weighted and summed to generate the expected deviation reference value of blood oxygen. The weighted summation process considers the impact of historical deviation on the current deviation, and dynamically adjusts the weight of historical data by introducing an exponential weighted attenuation coefficient. The calculation expression is: in: is the blood oxygen expected deviation reference value, which indicates the quantitative blood oxygen decline trend within the monitoring window. It is the expected deviation reference value of blood oxygen at the previous moment, reflecting the impact of historical blood oxygen changes. is the size of the monitoring window, indicating the time range of interest, For monitoring window i The absolute deviation between the blood oxygen saturation at the moment and the expected value, is an exponentially weighted decay factor, where is the attenuation coefficient, which controls the influence of the deviation on the current blood oxygen expected deviation reference value over time.

6. A sports armband monitoring method according to claim 5, characterized in that: In the monitoring window, the specific steps for analyzing the attenuation rate of blood oxygen saturation to generate a blood oxygen attenuation reference value are as follows: The blood oxygen saturation data is monitored in real time in the monitoring window, and its rate of change in each time step is calculated. The blood oxygen saturation decay rate is defined as the instantaneous rate of decrease of blood oxygen saturation. The calculation expression is: in, represents the blood oxygen saturation decay rate at time t, and are time t and Blood oxygen saturation data at all times, is the time interval, which captures the changing trend of the attenuation rate by calculating the logarithmic change of blood oxygen saturation, reflecting the rapid decrease of blood oxygen saturation; In order to more accurately quantify the downward trend of blood oxygen saturation, a weighted exponential attenuation model is introduced to generate a blood oxygen attenuation reference value based on the instantaneous change rate and the attenuation weighting factor at the attenuation moment. The expression for generating the blood oxygen attenuation reference value is: in, is the blood oxygen attenuation reference value, In the monitoring window The instantaneous rate of change within It means from the current time t to the past The offset of the time step, is the decay parameter, which is used to control the weight of past decay information. n is the length of the historical monitoring window considered.

7. A sports armband monitoring method according to claim 6, characterized in that: After quantitative analysis of the downward trend of blood oxygen saturation, the quantified blood oxygen expected deviation reference value and blood oxygen attenuation reference value are input into the pre-trained support vector machine regression model. Based on the downward trend index output by the model, the future trend of the downward trend of blood oxygen saturation is predicted by the downward trend index.

8. A sports armband monitoring method according to claim 7, characterized in that: According to the future trend of decreased blood oxygen saturation predicted by the support vector machine regression model, the cluster analysis algorithm is used to intelligently adjust the actual respiratory rate detection cycle. The specific steps are as follows: After obtaining the decline trend index, the cluster analysis algorithm is used to classify the future trend of blood oxygen saturation decline, and the data is divided into several categories. Each category represents a different blood oxygen change state. The specific classification formula is as follows: in, Indicates k clusters, It is k The range of the decline index of the clusters is is the downward trend index at time t, K for Total number of; The first downward trend index threshold and the second downward trend index threshold are set to divide the downward trend index into three categories, wherein First Downward Trend Index is greater than the second downward trend index threshold, the downward trend index The specific division steps are: If the downward trend index is greater than the first downward trend index , the trend of decreased blood oxygen saturation is divided into ; If the downward trend index is greater than the second downward trend index threshold and less than the first downward trend index , the trend of decreased blood oxygen saturation is divided into ; If the downward trend index is less than the second downward trend index threshold, the trend of blood oxygen saturation decrease is divided into ; After obtaining the blood oxygen change state, the actual respiratory rate detection cycle is intelligently adjusted according to different blood oxygen attenuation trends. The adjustment process is as follows: If the blood oxygen level drops rapidly, the respiratory rate detection cycle is shortened to monitor the user's respiratory rate changes more frequently. The adjustment formula is: in, is the initial respiratory rate detection cycle, is the adjustment factor, is the minimum breathing rate detection cycle, is the actual respiratory rate detection period after adjustment; If the blood oxygen level decreases slowly, the detection period is slightly extended, and the formula for adjusting the actual respiratory rate detection period is: in, is the elongation factor; If the blood oxygen is in a stable state, the initial respiratory rate detection cycle is maintained, and the formula is: .

9. A sports armband monitoring system, used to implement the sports armband monitoring method according to any one of claims 1 to 8, characterized in that: It includes an initial detection cycle setting module, a blood oxygen data real-time monitoring module, a data preprocessing and error elimination module, a blood oxygen change feature extraction and analysis module, a blood oxygen decline trend prediction module, and a respiratory rate detection cycle adjustment module; The initial detection cycle setting module sets an initial respiratory rate detection cycle based on the user's normal exercise intensity and physiological needs, establishes a standardized baseline, and ensures routine monitoring without interfering with the user's normal activities; The blood oxygen data real-time monitoring module monitors the blood oxygen level in real time through sensors during respiratory rate detection, and simultaneously obtains the user's blood oxygen saturation data so as to capture subtle changes in blood oxygen saturation; The data preprocessing and error elimination module preprocesses the acquired blood oxygen saturation data to eliminate measurement errors and interference from the external environment to ensure data accuracy; The blood oxygen change feature extraction and analysis module extracts the features reflecting the decrease of blood oxygen saturation after data preprocessing is completed, inputs the extracted blood oxygen saturation change features into the monitoring window for real-time analysis, and quantifies the decreasing trend of blood oxygen saturation; The blood oxygen decline trend prediction module quantitatively analyzes the decline trend of blood oxygen saturation, inputs the quantified features into the pre-trained support vector machine regression model, and predicts the future trend of blood oxygen saturation decline through the support vector machine regression model; The respiratory rate detection cycle adjustment module uses a cluster analysis algorithm to intelligently adjust the actual respiratory rate detection cycle according to the future trend of decreased blood oxygen saturation predicted by the support vector machine regression model, flexibly adapting to different physiological changes, and thus efficiently capturing the dynamic changes of respiratory rate.

Images