Children expiration flow monitoring method based on respiratory system symptoms

By establishing an individualized lung function database and cluster analysis, combined with real-time expiratory pressure data evaluation, the problem of difficulty in determining respiratory inflammation in children after exercise is solved, and accurate assessment and early warning of the risk of respiratory inflammation in children is achieved.

CN118924279AInactive Publication Date: 2024-11-12SHANGHAI EAST HOSPITAL EAST HOSPITAL TONGJI UNIV SCHOOL OF MEDICINE
View PDF 0 Cites 5 Cited by

Patent Information

Application Number
CN202411219314.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The prior art is difficult to accurately judge the respiratory inflammation of children, especially the correlation between the trend of PEF value changes and the degree of inflammation at different time points after exercise is unclear, and the individual physiological states vary greatly, making it difficult to grasp the change pattern of PEF value.

Method used

Establish an individualized basic pulmonary function value database, obtain the PEF baseline value of children under normal state, draw the PEF recovery curve after exercise, analyze the recovery mode through clustering algorithm, establish the correspondence between the PEF recovery mode and the diagnosis results of respiratory tract inflammation, and collect it in real time Expiration pressure change data, determine whether the child meets the force-exhaling standard, and evaluates the measurement posture, and judges the risk of inflammation based on the characteristics of the recovery curve and the pattern matching.

Benefits of technology

Through individualized analysis and recovery pattern matching, the early warning ability of children with respiratory inflammation is improved, providing accurate reference for clinical diagnosis and treatment, and optimizing prediction models.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118924279A_ABST
    Figure CN118924279A_ABST
Patent Text Reader

Abstract

The invention provides a children expiratory flow monitoring method based on respiratory system symptoms, which comprises the following steps: establishing an individualized lung function basic value database according to physiological parameters such as age, height and weight of children, and acquiring a PEF baseline value of the children in a normal state as a reference standard for judging abnormal PEF recovery; acquiring PEF measured values of a plurality of children at different time points after exercise, drawing a PEF recovery curve to obtain slope and curvature shape characteristic parameters of the PEF recovery curve, and calculating a PEF recovery rate and a recovery time index; the PEF recovery curve of each child is compared with the PEF baseline value of the child, and the deviation degree of the recovery curve relative to the baseline is calculated and serves as one of indexes for measuring PEF recovery abnormity.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for monitoring children's expiratory flow based on respiratory system symptoms. Background Art

[0002] At present, the method of measuring peak expiratory flow rate (PEF) is often used to assist in the diagnosis of respiratory inflammation in children. However, changes in PEF values ​​may not only be caused by pathological factors, but also by physiological factors, such as exercise. Therefore, it is difficult to accurately determine whether a child has respiratory inflammation based on a single measurement of PEF values. In order to improve the accuracy of PEF values ​​in judging respiratory inflammation, multiple measurements are required, and data at different time points need to be comprehensively analyzed. However, there is currently a lack of a scientific solution to integrate and analyze the PEF values ​​measured at different time points after exercise, and to determine whether there is inflammation or not. Specifically, the changing trend of PEF values ​​at different time points after exercise may be related to the degree of respiratory inflammation. However, the physiological states of different individuals vary greatly, making it difficult to grasp the changing pattern of PEF values. In addition, the specific changing characteristics of PEF values ​​at different time points after exercise under inflammatory and non-inflammatory conditions are currently unclear. Summary of the invention

[0003] The present invention provides a method for monitoring expiratory flow of children based on respiratory system symptoms, which mainly includes:

[0004] According to the physiological parameters of children's age, height, and weight, an individualized database of basic lung function values ​​was established to obtain the PEF baseline value of children in normal state as a reference standard for judging abnormal PEF recovery;

[0005] Obtain PEF measurement values ​​of multiple children at different time points after exercise, draw a PEF recovery curve, obtain the slope and curvature shape characteristic parameters of the PEF recovery curve, and calculate the PEF recovery rate and recovery time indicators;

[0006] The PEF recovery curve of each child was compared with his / her personal PEF baseline value, and the deviation of the recovery curve from the baseline was calculated as one of the indicators to measure the abnormal PEF recovery;

[0007] Analyze the PEF recovery curves of multiple children, and obtain different PEF recovery pattern categories according to the steepness and volatility characteristics of the curves. Match each child's PEF recovery curve with the PEF recovery pattern obtained by clustering to determine the recovery pattern to which it belongs.

[0008] Obtain clinical inflammation diagnosis results corresponding to the existing children's PEF recovery curve, establish the corresponding relationship between PEF recovery mode and respiratory inflammation diagnosis results, and obtain the probability distribution of inflammation under different PEF recovery modes;

[0009] When children are measuring their PEF, the pressure sensor collects the data of their expiratory pressure changes in real time, and determines whether the children have reached the standard of forceful exhalation based on the preset pressure threshold. If the children have not exerted enough force, the sensor guides the children to increase their force until the measurement requirements are met. The sensor also evaluates the children's measurement posture in real time. If an improper posture is detected, the sensor gives a correction prompt to obtain multiple PEF measurement values ​​after exercise.

[0010] Based on multiple PEF measurement values ​​after exercise, the shape characteristics of the children's PEF recovery curve and the degree of deviation from the baseline are obtained, and matched with the established PEF recovery pattern. According to the inflammation probability distribution corresponding to the matched recovery pattern, the risk of respiratory inflammation in children is judged. If the risk probability exceeds the preset threshold, the recovery pattern is considered to be related to inflammation, and the clinical evaluation results are output.

[0011] The technical solution provided by the embodiment of the present invention may include the following beneficial effects: The present invention discloses a method for monitoring expiratory flow of children based on respiratory system symptoms. The present invention establishes an individualized database of basic values ​​of children's lung function, obtains the PEF baseline value under normal conditions as a reference standard, collects PEF measurement values ​​at different time points after children exercise, and draws a PEF recovery curve. The PEF recovery curves of multiple children are analyzed by a clustering algorithm to obtain different recovery mode categories, and a corresponding relationship is established with the clinical inflammation diagnosis results. During the application process, the present invention collects the expiratory pressure change data of children through a pressure sensor, guides the children to reach the forced exhalation standard, and evaluates their measurement posture. According to the matching of the PEF recovery curve characteristics of the children with the established recovery mode, it is judged that they have the risk of respiratory inflammation, and clinical evaluation suggestions are given to high-risk children, and their diagnosis and treatment are tracked to verify and optimize the prediction model. The present invention can provide early warning of the risk of respiratory inflammation in children and provide a reference for clinical diagnosis and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 The present invention is a flow chart of a method for monitoring expiratory flow of children based on respiratory system symptoms. DETAILED DESCRIPTION

[0013] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0014] like Figure 1 In this embodiment, a method for monitoring expiratory flow of children based on respiratory system symptoms may specifically include:

[0015] Step S101, establish an individualized lung function basic value database based on the child's age, height, and weight physiological parameters, and obtain the child's PEF baseline value under normal conditions as a reference standard for judging abnormal PEF recovery.

[0016] Children were divided into three groups according to their age: preschool, school age and adolescent, and the physiological parameters of height, weight and chest circumference were obtained for each group. A correlation model between the physiological parameters of children in each age group and the baseline values ​​of lung function was established to obtain an individualized prediction formula for the basic value of lung function. For children of different age groups, peak expiratory flow data were collected, the correlation between PEF and age, height and weight factors was analyzed, a PEF prediction model was constructed, and the normal PEF reference value range for children of each age group was calculated. The respiratory rate and tidal volume of children at rest were measured, and the effects of respiratory rate, tidal volume and chest movement amplitude on lung function were evaluated in combination with the data of chest movement amplitude. The PEF change data of children at different activity intensities were collected, and the autoregressive integral moving average model was used to analyze the change trend of PEF over time, and a PEF recovery curve model for children after exercise was established to determine whether there were abnormalities in the PEF recovery process.

[0017] Specifically, children were divided into three groups according to their age: preschoolers aged 3-6, school-age children aged 7-12, and adolescent children aged 13-18. Physiological data such as height, weight, and chest circumference were collected for each group. A correlation model between the physiological parameters of children of each age group and the baseline values ​​of lung function was established through polynomial regression to obtain an individualized prediction formula for the basic value of lung function. Peak expiratory flow (PEF) data of children of different age groups were collected, and the correlation between PEF and age, height, weight and other factors of boys and girls was analyzed using the polynomial regression method. A PEF prediction model was constructed to calculate the normal PEF reference value range for children of different age groups. For children of different genders and body shape characteristics, their respiratory rate and tidal volume at rest were measured. Combined with the data of chest movement amplitude, the principal component analysis method was used to comprehensively evaluate the effects of respiratory rate, tidal volume and chest movement amplitude on lung function, and an individualized lung function assessment benchmark was obtained. From the resting state to the exercise state, the changes in children's PEF under different activity intensities were analyzed. Collect data on children's PEF changes at different activity intensities, use the autoregressive integrated moving average model to analyze the trend of PEF changes over time, establish a PEF recovery curve model for children after exercise, and determine whether there are abnormalities in the PEF recovery process. Integrate all collected data into a unified database, including PEF values ​​at rest and in exercise, individual baseline values, and related physiological parameters. Based on the deviation between the individual PEF baseline value in the database and the actual measured value, determine whether the PEF recovery is abnormal. In the process of establishing a database of basic lung function values ​​for children, first divide children aged 3-6, 7-12, and 13-18 into three age groups, and collect height, weight, and chest circumference data from 100 children in each group. Use polynomial regression to establish an association model. For example, for the 7-12 year old group, the vital capacity prediction formula may be: ;

[0018] FVC stands for vital capacity, H stands for height, W stands for weight, C stands for chest circumference, a, b, c, d, e, f, g are regression coefficients, a is a constant term, b, c, d are the linear coefficients of height, weight and chest circumference respectively, and e, f, g are the quadratic coefficients of height, weight and chest circumference respectively. PEF data were collected for boys and girls respectively, with at least 200 samples in each age group. Polynomial regression was used to analyze the nonlinear relationship between PEF and age, height and weight, such as PEF=k1age+k2height+k3weight+k4age^2+k5height^2+k6weight^2+k7, where k1 to k7 are the corresponding regression coefficients. The reference value range of PEF for each age group was calculated, such as the normal range of PEF for 7-year-old boys is 180-220L / min. For children with different body shape characteristics, measure the respiratory rate and tidal volume at rest, such as 8-year-old girls with a resting respiratory rate of 18-22 times / minute and a tidal volume of 250-300mL. Combined with the chest movement amplitude data, principal component analysis is used to evaluate the impact of these factors on lung function. Principal component analysis may find that the first principal component explains 70% of the variance, mainly composed of respiratory rate and tidal volume. Transition from rest to exercise, analyze the changes in children's PEF at different activity intensities, such as mild exercise with a heart rate of 120-140 beats / minute, moderate exercise with a heart rate of 140-160 beats / minute, and heavy exercise with a heart rate of 160-180 beats / minute. Collect PEF recovery data after exercise, such as measuring once every 30 seconds for 10 minutes. Use an autoregressive integrated moving average model to analyze the trend of PEF over time. The model may be ARIMA(1,1,1), that is, first-order autoregression, first-order difference, and first-order moving average. According to the PEF recovery curve predicted by the model, it is judged whether the actual PEF recovery process is abnormal. If the deviation between the actual value and the predicted value exceeds 20%, it is considered abnormal. All data are integrated into a unified database, including PEF values ​​at rest and in motion, individual baseline values ​​and related physiological parameters. The database stores each child's unique identification code, age, gender, height, weight, resting PEF, PEF values ​​at different exercise intensities and recovery curves. According to the deviation between the individual PEF baseline value in the database and the actual measured value, if the deviation exceeds a preset threshold, such as 15%, the PEF recovery is judged to be abnormal.

[0019] Step S102, obtaining PEF measurement values ​​of multiple children at different time points after exercise, drawing a PEF recovery curve, obtaining the slope and curvature shape characteristic parameters of the PEF recovery curve, and calculating the PEF recovery rate and recovery time index.

[0020] The PEF measurement value sequence of the child after exercise is obtained, and the PEF measurement value sequence is smoothed by using a sliding window median filter to obtain a smoothed PEF data sequence; according to the smoothed PEF data sequence and the corresponding time point, a quadratic polynomial is used to fit a PEF recovery curve, and the function form of the PEF recovery curve is: ;

[0021] Where y is the PEF value, x is time, a is the quadratic term coefficient, which indicates the influence of the square of time x on the PEF value, b is the linear term coefficient, which indicates the linear influence of time x on the PEF value, and c is the constant term, which indicates the basic level of the PEF value when x=0; the slope of the PEF recovery curve at each time point is calculated by the central difference method, and the slope value reflects the change rate of the PEF value; the curvature of the PEF recovery curve at different time points is calculated, and the curvature calculation formula is: ;

[0022] k represents curvature, a and b are respectively the quadratic coefficient and the linear coefficient of the curve equation, and x is the horizontal coordinate corresponding to the point on the curve; the mean and standard deviation of the curvature are calculated, and the mean and standard deviation are used as parameters to describe the shape characteristics of the PEF recovery curve; according to the PEF recovery curve, the time for the PEF value to recover from the peak value after exercise to the resting baseline value is calculated to obtain the PEF recovery time index, and the resting baseline value is the average value of the PEF measured a preset number of times before exercise; the PEF recovery rate is obtained by calculating the average slope of the PEF recovery curve.

[0023] Specifically, the PEF measurement values ​​of multiple children after exercise were obtained, once every 30 seconds for 10 minutes, and a total of 20 data points were obtained. The original PEF data was smoothed using a sliding window median filter, and the window size was set to 3 to remove noise interference during the measurement process and obtain a smoothed PEF data sequence. According to the smoothed PEF data sequence and the corresponding time points, a quadratic polynomial was used to fit the PEF recovery curve. The function form was y=ax^2+bx+c, where y is the PEF value and x is time. The slope of the fitting curve at each time point was calculated by the central difference method, and the slope value reflected the rate of change of the PEF value. Using the quadratic function obtained by fitting, the curvature of the PEF recovery curve at different time points was calculated. The curvature calculation formula is: ;

[0024] k represents curvature, a and b are coefficients of the curve equation, and x is the horizontal coordinate corresponding to the point on the curve. k is the curvature. The larger the curvature, the more curved the curve at that point, and the smaller the curvature, the straighter the curve at that point. a is the coefficient of the quadratic term, which determines the opening direction and width of the curve, b is the coefficient of the linear term, which determines the degree of inclination of the curve, and x is the independent variable, that is, the time point, where it represents the horizontal coordinate of a specific point on the curve. The mean and standard deviation of the curvature are calculated as parameters to describe the shape characteristics of the PEF recovery curve. Based on the fitted PEF recovery curve, the time for the PEF value to recover from the peak value after exercise to the resting baseline value is calculated to obtain the PEF recovery time index. The resting baseline value is the average value of the PEF measured three times before exercise. The PEF recovery rate is obtained by taking the average slope of the PEF recovery curve. Combined with the mean and standard deviation of the curvature, it is judged whether the PEF recovery is stable. If the standard deviation of the curvature is less than the preset threshold, the recovery process is considered to be stable. In the process of measuring PEF after children's exercise, the automated measuring device records the PEF value every 30 seconds for 10 minutes, and a total of 20 data points are obtained. Assume that the original PEF data sequence of a child is [350,345,355,340,348,352,347,351,349,353,350,348,352,351,349,350,348,351,350,349] L / min. Apply sliding window median filtering with a window size of 3 to obtain the smoothed PEF data sequence [350,350,348,348,348,351,350,351,350,350,350,350,351,351,349,350,350,350,350,349] L / min. The least squares method was used to fit a quadratic polynomial, and the function y=-0.02x^2+0.5x+349 was obtained, where y is the PEF value (L / min) and x is the time (minutes). The slope was calculated by the central difference method, such as the slope at 5 minutes was 0.3L / min / minute. The curvature was calculated to obtain a mean of 0.04 and a standard deviation of 0.01. The average PEF value of 340L / min measured three times before exercise was taken as the resting baseline value. From the fitting curve, it was determined that the time for the PEF value to recover from the peak value of 355L / min to the resting baseline value was 8.5 minutes, which is the PEF recovery time. The average recovery rate was calculated to be (355-340) / 8.5=1.76L / min / minute. Since the curvature standard deviation of 0.01 is less than the preset threshold of 0.05, it is determined that the PEF recovery process is stable.

[0025] Step S103, comparing the PEF recovery curve of each child with his / her personal PEF baseline value, and calculating the deviation degree of the recovery curve relative to the baseline as one of the indicators for measuring abnormal PEF recovery.

[0026] The PEF recovery curve data and personal PEF baseline value of each child in the database are obtained, wherein the personal PEF baseline value is the average value of PEF measured continuously for multiple times when the child is in a resting state; cubic spline interpolation is performed on the PEF recovery curve data to obtain a fitting function f(t), wherein t is time and f(t) is the PEF value; according to the fitting function f(t), a weighted least squares method is used to calculate a deviation index D between the PEF recovery curve and the personal PEF baseline value; a sliding window method is used to calculate the local deviation of the PEF recovery curve in different time periods, wherein the size of the sliding window is set to a preset value, and the sliding step is set to a preset value; with respect to the local deviation, an abnormal change point in the PEF recovery process is detected, and if the cumulative sum exceeds a preset control limit, it is determined to be an abnormal change point; it is determined whether the deviation index D exceeds a preset threshold or whether the number of abnormal change points exceeds a preset number of times, and if so, it is determined that the PEF recovery process of the child is abnormal.

[0027] Specifically, the PEF recovery curve data and personal PEF baseline value of each child were obtained from the database. The personal PEF baseline value was the average value of the PEF measured three times in a resting state. The PEF recovery curve data was interpolated with cubic spline to obtain the fitting function f(t), where t is time and f(t) is the PEF value, to improve the fitting accuracy of the complex PEF recovery curve. The weighted least squares method was used to calculate the deviation index D between the PEF recovery curve and the personal PEF baseline value of PEF_baseline. ;

[0028] n represents the total number of sampling points, ti represents the i-th sampling time point, f(ti) is the PEF recovery curve value at time point ti, PEFbaseline is the individual PEF baseline value, wi is the weight of the i-th time point, and this weight decreases over time, indicating that the deviation in the early stage of recovery is more important. The calculated deviation index D is standardized to facilitate comparison between different children. The local deviation of the PEF recovery curve in different time periods is calculated using the sliding window method. The window size is set to 60 seconds and the sliding step is 30 seconds to obtain the local deviation sequence. The local deviation sequence is normalized to eliminate the dimensional difference of the deviation in different time periods. According to the normalized local deviation sequence, the cumulative sum (CUSUM) algorithm is used to detect abnormal change points in the PEF recovery process. The control limit of the CUSUM algorithm is set. When the cumulative sum exceeds the control limit, it is determined to be an abnormal change point. Combined with the overall deviation index D and the number of abnormal change points detected, if the D value exceeds the preset threshold or the number of abnormal change points exceeds the preset number, it is determined that the PEF recovery process of the child is abnormal. Assume that the PEF recovery curve data of a 10-year-old child is obtained from the database. The measurement time is 10 minutes, with one data point every 30 seconds, for a total of 21 data points. The child's personal PEF baseline value is 250L / min, which is the average of three consecutive measurements at rest. The PEF recovery curve data is interpolated with cubic splines to obtain a continuous fitting function f(t). Then the weighted deviation index D is calculated, and the weight is set to wi=1 / (1+0.1*i), where i is the data point number, and D=15.3L / min is calculated. The D value is divided by the baseline value of 250L / min to obtain a standardized deviation of 5.12%. Then, the local deviation sequence is calculated using the sliding window method with a 60-second window and a 30-second step size, and 19 local deviation values ​​are obtained. These values ​​are normalized to obtain a normalized local deviation sequence ranging from 0 to 1. The CUSUM algorithm is applied to detect abnormal change points, and the control limit is set to the mean plus 3 times the standard deviation. The CUSUM algorithm detected two abnormal change points, located at the 3rd and 7th minutes. Finally, based on the preset standardized deviation threshold of 4% and the threshold of the number of abnormal change points of 1, it was determined that the child's PEF recovery process was abnormal, because the standardized deviation of 5.12% exceeded the threshold of 4%, and the two abnormal change points detected also exceeded the preset threshold of 1. The entire analysis process was completed by automated software without manual intervention, ensuring the consistency and objectivity of data processing.

[0029] Step S104, analyzing the PEF recovery curves of multiple children, and obtaining different PEF recovery mode categories according to the steepness and volatility characteristics of the curves. The PEF recovery curve of each child is matched with the PEF recovery mode obtained by clustering to determine the recovery mode to which it belongs.

[0030] The PEF recovery curve data of multiple children are obtained, and the PEF recovery curve data are subjected to minimum and maximum normalization processing to obtain normalized PEF recovery curve data; characteristic indicators of the PEF recovery curve are calculated according to the normalized PEF recovery curve data, and the characteristic indicators include average slope, standard deviation, maximum curvature, peak position and recovery time; the characteristic indicators are subjected to dimension reduction processing by principal component analysis to obtain a characteristic vector after dimension reduction; the characteristic vector after dimension reduction is subjected to cluster analysis by K-means clustering algorithm to obtain a PEF recovery mode category; for a new child PEF recovery curve, characteristic indicators of the new child PEF recovery curve are calculated, and the characteristic indicators are projected into the principal component space to obtain a projected characteristic vector; the Euclidean distance between the projected characteristic vector and each cluster center is calculated to determine the PEF recovery mode to which the new child PEF recovery curve belongs; if the Euclidean distances all exceed a preset threshold, the new child PEF recovery curve is determined to be a new recovery mode.

[0031] Specifically, the PEF recovery curve data of multiple children were obtained from the database, and each curve was normalized to the minimum and maximum values, and the numerical range was unified to between 0 and 1 to eliminate the baseline level differences between individuals. The characteristic indicators of the curve were calculated, including the average slope (steepness), standard deviation (volatility), maximum curvature, peak position and recovery time. For curves of different lengths, the interpolation method was used to unify them to the same number of time points. The principal component analysis method was used to reduce the dimension of the calculated characteristic indicators. First, the characteristic data was standardized to eliminate the dimension effect. Through variance contribution analysis, the principal components with a cumulative contribution rate of more than 80% were selected as clustering features to obtain the reduced-dimensional feature vector. For possible outliers, the quartile method was used to detect and eliminate them to ensure data quality. The K-means clustering algorithm was used to cluster the reduced-dimensional feature vector, and the optimal value of the number of clusters was determined by the silhouette coefficient method. The K-means++ algorithm was used to select the initial cluster center to improve the clustering stability and convergence speed. Through iterative optimization, different PEF recovery mode categories were obtained, and each category was represented by its cluster center. For each new child's PEF recovery curve, its characteristic index is calculated and projected into the principal component space, and then the Euclidean distance from each cluster center is calculated. A distance threshold is set. If the distance from all cluster centers exceeds the threshold, it is determined to be a new recovery mode. Otherwise, it is divided into the cluster category with the closest distance to determine which PEF recovery mode it belongs to. For samples determined to be new recovery modes, the cluster analysis is recorded and re-performed regularly to update the recovery mode category. In the process of implementing the PEF recovery curve analysis, the PEF recovery curve data of 100 children were first extracted from the database, and each curve contained the measurement values ​​of 20 time points. Each curve was normalized to the minimum and maximum value to unify the PEF value range to between 0 and 1. When calculating the characteristic index, the average slope was obtained by linear regression, the standard deviation was calculated directly, the maximum curvature was estimated by the three-point method, the peak position was the time point corresponding to the maximum PEF value, and the recovery time was defined as the time required for the PEF value to recover to 90% of the baseline. For the 19-point curve, linear interpolation was used to supplement it to 20 points. In principal component analysis, the eigenvalues ​​of the five standardized features were calculated to be [2.8, 1.2, 0.6, 0.3, 0.1]. The first three principal components were selected, and the cumulative contribution rate was 92%. The quartile method was used to detect outliers, and 1.5 times the interquartile range was set as the threshold to remove three abnormal samples. In K-means clustering, the optimal number of clusters was determined to be 4 by calculating the silhouette coefficient, and the K-means++ algorithm selected the initial center point. After 50 iterations, four PEF recovery modes were obtained, and the cluster centers were [-0.8, 0.5, 0.2], [0.1, -0.7, 0.4], [0.6, 0.3, -0.5], and [-0.2, -0.3, -0.6].When classifying new samples, calculate the Euclidean distance between them and the cluster center, and set the threshold to 1.5. If the minimum distance is greater than 1.5, it is considered a new pattern, otherwise it is classified into the category with the smallest distance. After accumulating 50 new samples, re-clustering is automatically triggered, and the PEF recovery pattern category is updated to ensure the accuracy and adaptability of classification.

[0032] Step S105, obtaining the clinical inflammation diagnosis result corresponding to the existing children's PEF recovery curve, establishing the corresponding relationship between the PEF recovery mode and the respiratory inflammation diagnosis result, and obtaining the probability distribution of the presence of inflammation under different PEF recovery modes.

[0033] Obtain children's PEF recovery curve data and corresponding respiratory inflammation diagnosis results from the clinical database, perform feature extraction on the PEF recovery curve, and obtain feature data and label data. Perform data cleaning on the feature data and label data, remove outliers and missing data, and balance the samples. Construct a classification model between the PEF recovery mode and the respiratory inflammation diagnosis results based on the processed data set. Optimize the classification model parameters by grid search method, evaluate the performance of the classification model by cross validation, and screen out the optimal model. For the optimal model, calculate the probability distribution of inflammation at various levels under different PEF recovery modes, estimate the confidence interval of the probability, and obtain the probability and confidence interval of no inflammation, mild inflammation, moderate inflammation and severe inflammation corresponding to each PEF recovery mode.

[0034] Specifically, the PEF recovery curve data of children and the corresponding respiratory inflammation diagnosis results were obtained from the clinical database, and the PEF recovery curve was feature extracted, including the calculation of the average slope, maximum slope, peak value, recovery time and area under the curve. The inflammation diagnosis results were divided into four levels: no inflammation, mild inflammation, moderate inflammation and severe inflammation, and labeled. The extracted feature data and label data were cleaned to remove outliers and missing data beyond 3 times the standard deviation. The SMOTE algorithm was used to balance the samples, setting the k-nearest neighbor parameter to 5 and the oversampling ratio to 200% to solve the class imbalance problem. The processed data set was randomly divided into a training set and a test set at a ratio of 8:2. The random forest algorithm and the gradient boosting tree algorithm were used to construct the classification model between the PEF recovery pattern and the respiratory inflammation diagnosis results, respectively. The model parameters were optimized by the grid search method, and the number of trees in the random forest was searched between 100 and 1000, and the tree depth was searched between 5 and 20; the learning rate of the gradient boosting tree was searched between 0.01 and 0.1, and the subsampling ratio was searched between 0.5 and 1. The performance of the model was evaluated by 10-fold cross validation, and the model with the highest F1 score was selected. Based on the trained optimal model, the probability distribution of inflammation at different levels under different PEF recovery modes was calculated. The confidence interval of the probability was estimated by the bootstrap method, and 1000 resamplings were performed. Each time, 80% of the samples were extracted from the original data with replacement for probability estimation. The probabilities of no inflammation, mild inflammation, moderate inflammation, and severe inflammation corresponding to each PEF recovery mode and their 95% confidence intervals were obtained. In the implementation of the association analysis between PEF recovery curve and inflammation diagnosis, the PEF recovery curve data and inflammation diagnosis results of 1000 children were first extracted from the clinical database. Five features were calculated for each PEF curve: average slope 0.5 L / min / s, maximum slope 2.3 L / min / s, peak 380 L / min, recovery time 120 s, and area under the curve 36000 L·s. The inflammation diagnosis results were divided into four levels: no inflammation (0), mild (1), moderate (2), and severe (3). After data cleaning, 950 valid data remained, including 400 without inflammation, 300 mild, 200 moderate, and 50 severe. After SMOTE algorithm processing, the number of samples in each category was 800. The optimization result of the random forest model parameters was 500 trees with a maximum depth of 15; the optimal parameters of the gradient boosting tree were a learning rate of 0.05 and a subsampling ratio of 0.8. After 10-fold cross-validation, the average F1 score of the random forest was 0.85, and that of the gradient boosting tree was 0.83. The random forest was selected as the final model.For a certain PEF recovery mode, such as the rapid recovery type, 1000 bootstrap samplings were performed, and the probability distribution of inflammation was obtained as follows: no inflammation 0.6 [0.55, 0.65], mild 0.25 [0.21, 0.29], moderate 0.12 [0.09, 0.15], severe 0.03 [0.01, 0.05], with 95% confidence intervals in square brackets.

[0035] Step S106, when the child is performing PEF measurement, the expiratory pressure change data is collected in real time through the pressure sensor, and it is determined whether the child has reached the standard of forceful exhalation based on the preset pressure threshold. If the force is insufficient, the child is guided to increase the force until the measurement requirement is met, and the child's measurement posture is evaluated in real time. If an improper posture is detected, a correction prompt is given to obtain multiple PEF measurement values ​​after exercise.

[0036] Acquire the child's expiratory pressure data collected by the pressure sensor, and smooth the expiratory pressure data to eliminate instantaneous fluctuations; determine whether the preset pressure threshold is reached based on the smoothed expiratory pressure data, and if the expiratory pressure data is lower than the minimum pressure threshold, trigger the speech synthesis module to generate a prompt sound; obtain the image data of the child during measurement through a camera, and process the image data to evaluate the measurement posture; if an improper posture is detected, generate a correction prompt; record the measurement value corresponding to the expiratory pressure data, determine whether the measurement value deviates from the preset percentage of the previous valid measurement value, and if not, save the measurement value to the database; perform multiple valid measurements continuously, calculate the average value of the measurement values ​​of the multiple valid measurements, and obtain the final PEF measurement result.

[0037] Specifically, the pressure sensor is used to collect the data of the child's exhalation pressure change in real time, and the sampling frequency is set to 100Hz. The collected pressure data is converted into a digital signal and stored in the buffer area. The sliding window size is set to 50ms, and the data is smoothed to eliminate instantaneous fluctuations. At the same time, the camera is used to collect the image data of the child during measurement for posture evaluation. According to the preset pressure threshold, it is judged whether the child meets the standard of forceful exhalation. The minimum pressure threshold is set to 2kPa and the maximum pressure threshold is set to 5kPa. If the detected exhalation pressure is lower than 2kPa, the speech synthesis module is triggered to generate a prompt sound of "Please exhale forcefully"; if it is higher than 5kPa, a prompt sound of "Do not use excessive force" is generated. If the pressure value is between 2kPa and 5kPa, it is determined to be a valid measurement. The image processing method based on template matching is used to evaluate the child's measurement posture in real time. The template image of the standard measurement posture is stored in advance, and the posture is judged to be correct by calculating the similarity between the current image and the template image. If an improper posture is detected, the speech synthesis module is triggered to generate corresponding correction prompts, such as "Please keep your head upright" or "Please straighten your back". After the child completes a PEF measurement, the measurement value is recorded and saved in the database. If the measurement value deviates from the previous valid measurement value by more than 20%, it is judged as an invalid measurement and prompted to re-measure. Start the timer and prompt for the next measurement every 1 minute. Perform 5 valid measurements in succession to obtain multiple PEF measurement values ​​after exercise. Calculate the average and standard deviation of these measurement values ​​as the final PEF measurement result. In actual applications, the PEF measurement device uses a high-precision pressure sensor with a sampling frequency set to 100Hz, and collects expiratory pressure data every 10ms. The collected raw data is converted by a 12-bit ADC and stored in the buffer in the form of a digital signal. At the same time, a 50ms sliding window is used to smooth the data and eliminate high-frequency noise. The device has a built-in 5-megapixel camera that captures 30 frames of images per second for posture assessment. The pressure threshold is dynamically adjusted according to the child's age and height. For example, the minimum threshold for an 8-year-old child is set to 2.2kPa and the maximum threshold is 4.8kPa. When the exhalation pressure is detected to be 1.8kPa, the device immediately plays the pre-recorded "Please exhale forcefully" prompt tone; when the pressure reaches 5.1kPa, the "Do not use excessive force" prompt tone is played. Posture assessment uses a template matching algorithm, pre-stores 5 standard posture templates, and calculates the normalized correlation coefficient between the current image and the template in real time, with the threshold set to 0.85. If the coefficient is lower than the threshold, the corresponding correction prompt is played. For example, when the head deviation is detected to be more than 15 degrees, the "Please keep your head upright" prompt is played. After each measurement is completed, the data is compared with the previous valid value, with a 20% deviation allowed. If a certain measurement value is 280L / min, and the previous valid value is 380L / min, it is judged to be invalid and automatically prompted to re-measure. The device prompts for the next measurement every 60 seconds until 5 valid measurements are completed.Finally, the average value and standard deviation of the five measurements were calculated, such as an average value of 350 L / min and a standard deviation of 15 L / min, as the final PEF measurement result.

[0038] Step S107, based on multiple PEF measurement values ​​after exercise, the shape characteristics of the child's PEF recovery curve and the degree of deviation from the baseline are obtained, and matched with the established PEF recovery pattern, and the child is judged to have a risk of respiratory inflammation based on the inflammation probability distribution corresponding to the matched recovery pattern. If the risk probability exceeds the preset threshold, the recovery pattern is considered to be related to inflammation, and the clinical evaluation result is output.

[0039] The multiple PEF measurement values ​​obtained after exercise are fitted using the cubic spline interpolation method to obtain the PEF recovery curve function f(t), where t is time and f(t) is the PEF value. The relative deviation from the child's personal PEF baseline value is calculated based on the PEF recovery curve function f(t), and the weighted root mean square error is used as a quantitative indicator of the degree of deviation. The shape features extracted from the PEF recovery curve function f(t) and the deviation are standardized to eliminate the dimension effect to obtain standardized features. The standardized features are matched with the pre-established PEF recovery pattern library using the K nearest neighbor algorithm, and the optimal K value is determined by cross-validation. If the standardized features match a recovery pattern in the PEF recovery pattern library, the inflammation probability distribution corresponding to the recovery pattern is obtained. According to the inflammation probability distribution, the conditional probability of the child having respiratory inflammation is calculated to obtain the risk probability. If the risk probability exceeds the preset threshold, it is judged that the recovery pattern is related to inflammation; if the risk probability falls within the preset confidence interval, it is marked as requiring further observation.

[0040] Specifically, the multiple PEF measurements obtained after exercise were fitted using the cubic spline interpolation method to obtain the PEF recovery curve function f(t), where t is time and f(t) is the PEF value. The shape characteristics of the recovery curve were extracted by calculating parameters such as the slope, curvature, area, and peak position of the curve. For the case where the measurement time intervals are uneven, the interpolation method is used to unify the data points into an equally spaced time series. The relative deviation between the PEF recovery curve and the child’s personal PEF baseline value is calculated, and the weighted root mean square error (WRMSE) is used as a quantitative indicator of the degree of deviation. The WRMSE is calculated using the following formula, ;

[0041] n represents the number of measurements, ti represents the time point of the i-th measurement, f(ti) represents the PEF measurement value at time ti, PEFbaseline represents the child's personal PEF baseline value, and wi represents the time weight of the i-th measurement. This weight decreases over time, highlighting the importance of the initial recovery period. The extracted shape features and deviations are standardized to eliminate the dimension effect. The processed features are matched with the pre-established PEF recovery pattern library using the K nearest neighbor algorithm, and the optimal K value is determined by 5-fold cross validation. The K nearest neighbors with the smallest Euclidean distance are selected, and the most matching recovery pattern and its corresponding inflammation probability distribution are obtained by weighted average. According to the inflammation probability distribution corresponding to the matched recovery pattern, the conditional probability of the child having respiratory inflammation is calculated using the Bayesian theorem as the final risk probability. If the risk probability exceeds the preset threshold, such as 0.6, the recovery pattern is judged to be related to inflammation. For uncertain matching results, a confidence interval is set. If the risk probability falls within the confidence interval, it is marked as requiring further observation. The clinical assessment recommendation report is automatically generated, including information such as the PEF recovery curve, the matching recovery pattern, the risk probability and its confidence interval. In practical application, the PEF recovery data of a 10-year-old child were analyzed, and the PEF values ​​at five time points were measured to be 280, 310, 330, 345, and 350 L / min, respectively. The data points were fitted using cubic spline interpolation to obtain a continuous PEF recovery curve function. Features were extracted from the curve, including the initial slope of 2.5 L / min / s, the maximum curvature of 0.05 s^-2, the area under the curve of 17500 L·s, and the peak position of 240 s. The child's personal PEF baseline value was 360 L / min, and the weighted WRMSE was calculated to be 15.3 L / min. The weight was set to wi=1 / (1+0.1*i), where i was the measurement number. The features and WRMSE were z-score standardized to obtain the standardized feature vector [-0.8, 1.2, 0.5, -0.3, 0.7]. The K nearest neighbor algorithm was used for pattern matching, and the optimal K value was determined to be 4 through 5-fold cross validation. The matching results showed that the recovery pattern belonged to the "slow recovery type", and the corresponding inflammation probability distribution was [0.2, 0.4, 0.3, 0.1], which respectively represented the probabilities of no, mild, moderate and severe inflammation. Applying the Bayesian theorem, the risk probability of respiratory inflammation in the child was calculated to be 0.65. Since 0.65 exceeded the preset threshold of 0.6, the system automatically generated a clinical assessment recommendation report, including the PEF recovery curve, the "slow recovery type" pattern description, the risk probability of 0.65 and its 95% confidence interval [0.58, 0.72].

[0042] The above is only a preferred embodiment of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and supplements without departing from the principle of the present invention. These improvements and supplements should also be regarded as the scope of protection of the present invention.

Claims

1. A method for monitoring expiratory flow of children based on respiratory system symptoms, characterized in that: The method comprises: According to the physiological parameters of children's age, height, and weight, an individualized pulmonary function basic value database was established to obtain the PEF baseline value of children in normal state as a reference standard for judging abnormal PEF recovery; Obtain PEF measurement values ​​of multiple children at different time points after exercise, draw a PEF recovery curve, obtain the slope and curvature shape characteristic parameters of the PEF recovery curve, and calculate the PEF recovery rate and recovery time indicators; The PEF recovery curve of each child was compared with his / her personal PEF baseline value, and the deviation of the recovery curve from the baseline was calculated as one of the indicators to measure the abnormal PEF recovery; Analyze the PEF recovery curves of multiple children, and obtain different PEF recovery pattern categories according to the steepness and volatility characteristics of the curves. Match each child's PEF recovery curve with the PEF recovery pattern obtained by clustering to determine the recovery pattern to which it belongs. Obtain clinical inflammation diagnosis results corresponding to the existing children's PEF recovery curve, establish the corresponding relationship between PEF recovery mode and respiratory inflammation diagnosis results, and obtain the probability distribution of inflammation under different PEF recovery modes; When children are measuring their PEF, the pressure sensor collects the data of their expiratory pressure changes in real time, and determines whether the children have reached the standard of forceful exhalation based on the preset pressure threshold. If the children have not exerted enough force, the sensor guides the children to increase their force until the measurement requirements are met. The sensor also evaluates the children's measurement posture in real time. If an improper posture is detected, the sensor gives a correction prompt to obtain multiple PEF measurement values ​​after exercise. Based on multiple PEF measurement values ​​after exercise, the shape characteristics of the children's PEF recovery curve and the degree of deviation from the baseline are obtained, and matched with the established PEF recovery pattern. According to the inflammation probability distribution corresponding to the matched recovery pattern, the risk of respiratory inflammation in children is judged. If the risk probability exceeds the preset threshold, the recovery pattern is considered to be related to inflammation, and the clinical evaluation results are output.

2. A method for monitoring pediatric expiratory flow based on respiratory system symptoms according to claim 1, characterized in that: The method of establishing an individualized lung function basic value database based on the physiological parameters of the child's age, height, and weight, and obtaining the PEF baseline value of the child in a normal state as a reference standard for judging abnormal PEF recovery includes: Children were divided into three groups according to their age: preschool, school age, and adolescent, and the physiological parameters of height, weight, and chest circumference were obtained for each group; Establish the correlation model between physiological parameters and lung function baseline values ​​of children of different age groups, and obtain the individualized lung function baseline value prediction formula; For children of different age groups, we collected peak expiratory flow data, analyzed the correlation between PEF and age, height and weight, built a PEF prediction model, and calculated the normal PEF reference value range for children of different age groups. Measure the children's respiratory rate and tidal volume at rest, and combine the chest movement data to evaluate the effects of respiratory rate, tidal volume and chest movement on lung function; The PEF change data of children under different activity intensities were collected, and the autoregressive integrated moving average model was used to analyze the change trend of PEF over time. A PEF recovery curve model for children after exercise was established to determine whether there were abnormalities in the PEF recovery process.

3. A method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: The method of obtaining PEF measurement values ​​of multiple children at different time points after exercise, drawing a PEF recovery curve, obtaining the slope and curvature shape characteristic parameters of the PEF recovery curve, and calculating the PEF recovery rate and recovery time index includes: Obtaining a PEF measurement value sequence of the child after exercise, and smoothing the PEF measurement value sequence using a sliding window median filter to obtain a smoothed PEF data sequence; According to the smoothed PEF data sequence and the corresponding time point, a quadratic polynomial is used to fit the PEF recovery curve, and the functional form of the PEF recovery curve is: ; Where y is the PEF value, x is time, a is the quadratic term coefficient, which indicates the influence of the square of time x on the PEF value, b is the linear term coefficient, which indicates the linear influence of time x on the PEF value, and c is the constant term, which indicates the basic level of the PEF value when x=0; The slope of the PEF recovery curve at each time point is calculated by the central difference method, and the slope value reflects the change rate of the PEF value; Calculate the curvature of the PEF recovery curve at different time points. The curvature calculation formula is: ; k represents the curvature, a and b are the coefficients of the quadratic term and the linear term of the curve equation respectively, and x is the horizontal coordinate corresponding to the point on the curve; Calculating the mean and standard deviation of the curvature, wherein the mean and standard deviation are used as parameters to describe the shape characteristics of the PEF recovery curve; According to the PEF recovery curve, the time for the PEF value to recover from the peak value after exercise to the resting baseline value is calculated to obtain a PEF recovery time index, wherein the resting baseline value is the average value of the PEF measured a preset number of times before exercise; The PEF recovery rate is obtained by calculating the average slope of the PEF recovery curve.

4. A method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: The PEF recovery curve of each child is compared with his / her personal PEF baseline value, and the deviation of the recovery curve from the baseline is calculated as one of the indicators to measure the abnormal PEF recovery, including: Obtaining the PEF recovery curve data and personal PEF baseline value of each child in the database, wherein the personal PEF baseline value is the average value of the PEF measured continuously for multiple times when the child is in a resting state; Performing cubic spline interpolation on the PEF recovery curve data to obtain a fitting function f(t), wherein t is time and f(t) is the PEF value; According to the fitting function f(t), a deviation index D between the PEF recovery curve and the personal PEF baseline value is calculated using a weighted least squares method; The local deviation of the PEF recovery curve in different time periods is calculated using a sliding window method, the size of the sliding window is set to a preset value, and the sliding step is set to a preset value; For the local deviation, detect the abnormal change point in the PEF recovery process, and if the cumulative sum exceeds the preset control limit, it is determined to be an abnormal change point; It is determined whether the deviation index D exceeds a preset threshold or whether the number of abnormal change points exceeds a preset number of times. If so, it is determined that there is an abnormality in the PEF recovery process of the child.

5. The method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: The PEF recovery curves of multiple children are analyzed to obtain different PEF recovery mode categories according to the steepness and volatility characteristics of the curves, and the PEF recovery curve of each child is matched with the PEF recovery mode obtained by clustering to determine the recovery mode to which the child belongs, including: Acquiring PEF recovery curve data of a plurality of children, performing minimum and maximum normalization processing on the PEF recovery curve data, and obtaining normalized PEF recovery curve data; Calculating characteristic indices of the PEF recovery curve according to the normalized PEF recovery curve data, wherein the characteristic indices include average slope, standard deviation, maximum curvature, peak position and recovery time; The principal component analysis method is used to perform dimension reduction processing on the characteristic index to obtain a characteristic vector after dimension reduction; Using K-means clustering algorithm to perform cluster analysis on the feature vector after dimension reduction to obtain PEF recovery mode category; For a new children's PEF recovery curve, calculating characteristic indicators of the new children's PEF recovery curve, projecting the characteristic indicators into a principal component space, and obtaining a projected characteristic vector; Calculating the Euclidean distance between the projected eigenvector and each cluster center to determine the PEF recovery mode to which the new child PEF recovery curve belongs; If the Euclidean distances all exceed the preset threshold, the new children's PEF recovery curve is determined to be a new recovery mode.

6. A method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: The method of obtaining the clinical inflammation diagnosis result corresponding to the existing children's PEF recovery curve, establishing the corresponding relationship between the PEF recovery mode and the respiratory inflammation diagnosis result, and obtaining the probability distribution of inflammation under different PEF recovery modes includes: Obtain children's PEF recovery curve data and corresponding respiratory inflammation diagnosis results in a clinical database, perform feature extraction on the PEF recovery curve, and obtain feature data and label data; Performing data cleaning on the feature data and label data, removing outliers and missing data, and performing balanced processing on the samples; A classification model between PEF recovery pattern and respiratory inflammation diagnosis results was constructed based on the processed data set; Optimize the classification model parameters by grid search method, evaluate the classification model performance by cross validation, and select the optimal model; For the optimal model, the probability distribution of inflammation at various levels under different PEF recovery modes was calculated, and the confidence interval of the probability was estimated to obtain the probabilities and confidence intervals of no inflammation, mild inflammation, moderate inflammation and severe inflammation corresponding to each PEF recovery mode.

7. A method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: When the child is performing PEF measurement, the pressure sensor collects the expiratory pressure change data in real time, and determines whether the child has reached the standard of forceful exhalation according to the preset pressure threshold. If the child has insufficient force, the child is guided to increase the force until the measurement requirement is met, and the child's measurement posture is evaluated in real time. If an improper posture is detected, a correction prompt is given to obtain multiple PEF measurement values ​​after exercise, including: Acquire the child's exhalation pressure data collected by the pressure sensor, and perform smoothing on the exhalation pressure data to eliminate instantaneous fluctuations; Determining whether the preset pressure threshold is reached according to the smoothed exhalation pressure data, and if the exhalation pressure data is lower than the minimum pressure threshold, triggering the speech synthesis module to generate a prompt tone; Acquiring image data of the child during measurement through a camera, and processing the image data to evaluate the measurement posture; If improper posture is detected, correction prompts are generated; Recording the measured value corresponding to the expiratory pressure data, determining whether the measured value deviates from a preset percentage of the previous valid measured value, and if not, saving the measured value to a database; Multiple valid measurements are performed continuously, and the average value of the measurement values ​​of the multiple valid measurements is calculated to obtain the final PEF measurement result.

8. A method for monitoring expiratory flow of children based on respiratory system symptoms according to claim 1, characterized in that: The shape characteristics of the child's PEF recovery curve and the degree of deviation from the baseline are obtained based on multiple PEF measurement values ​​after exercise, and matched with the established PEF recovery pattern. According to the inflammation probability distribution corresponding to the matched recovery pattern, the child is judged to have a risk of respiratory inflammation. If the risk probability exceeds a preset threshold, the recovery pattern is considered to be related to inflammation, and the clinical evaluation results are output, including: The cubic spline interpolation method was used to fit multiple PEF measurement values ​​obtained after exercise to obtain the PEF recovery curve function f(t), where t is time and f(t) is the PEF value; The relative deviation from the child's personal PEF baseline value is calculated according to the PEF recovery curve function f(t), and the weighted root mean square error is used as a quantitative indicator of the degree of deviation; The shape feature extracted from the PEF recovery curve function f(t) and the deviation are normalized to eliminate the dimension effect, thereby obtaining a normalized feature; The standardized features are matched with a pre-established PEF recovery pattern library using a K-nearest neighbor algorithm, and the optimal K value is determined by cross-validation; If the standardized feature matches a recovery pattern in the PEF recovery pattern library, then obtaining the inflammation probability distribution corresponding to the recovery pattern; According to the inflammation probability distribution, the conditional probability that the child has respiratory inflammation is calculated to obtain the risk probability; If the risk probability exceeds a preset threshold, the recovery pattern is judged to be related to inflammation; If the risk probability falls within the preset confidence interval, it is marked as requiring further observation.

Citation Information

Cited By

  • Intelligent diagnosis auxiliary system and method for children respiratory diseases

    CN119517384A

  • RER estimation method, device and equipment based on PetCO2 and storage medium

    CN119587003A

  • Monitoring system of newborn breathing machine

    CN119857199A

  • Quality control method and device for lung function examination report, computer equipment and medium

    CN120105319A

  • Pelvic floor muscle pressure detection method and system for pelvic floor muscle function screening

    CN120345902A