Children asthma patient health management system combined with multi-level supervision

Abstract

The invention relates to the technical field of medical health monitoring and management, and particularly discloses a pediatric asthma patient health management system combined with multistage supervision, which comprises a data acquisition module, a respiratory stability evaluation module, a respiratory mechanics sufficiency evaluation module, an asthma state comprehensive evaluation module and an early warning intervention module, a wearable device is used for collecting airway resistance and elastic retraction force physiological signals of a child patient in real time, an airway reaction differential frequency characteristic value and a respiratory mechanics difference characteristic value are extracted respectively, and comprehensive evaluation is conducted from the two dimensions of respiratory tract dynamic stability and lung recruitment capacity. The feature fusion is further input into an asthma severity judgment model constructed based on a random forest algorithm, intelligent grading of asthma states is achieved, a three-level early warning mechanism is triggered according to a grading result, and remote notification and personalized intervention are supported.

Description

Technical Field

[0001] The present invention relates to the technical field of medical health monitoring and management, and particularly relates to a health management system for pediatric asthma patients combined with multi-level supervision. Background Art

[0002] In the clinical management of childhood asthma, the dynamic monitoring of respiratory function status and the accurate assessment of disease severity are the keys to achieving early intervention and personalized treatment. Traditional methods mainly rely on subjective symptom scoring, pulmonary function testing, and doctor experience judgment, which have problems such as being insensitive to the real-time physiological changes of children, the evaluation results being easily affected by subjective factors, and being unable to continuously monitor. In recent years, with the development of wearable physiological monitoring devices and artificial intelligence algorithms, intelligent health management systems based on multi-source physiological signal fusion analysis have gradually been applied to the field of asthma management. For example, existing research has attempted to extract features using parameters such as airway resistance, respiratory rate, and blood oxygen saturation, and combine machine learning models to achieve disease grading and early warning. However, existing technologies are still difficult to comprehensively reflect the complex respiratory mechanics characteristics and their dynamic change trends of asthmatic children, and there is an urgent need for a more refined and intelligent comprehensive evaluation system to improve the accuracy and real-time nature of disease management.

[0003] The existing technologies have the following deficiencies:

[0004] The existing technologies lack multi-dimensional fusion analysis means that can simultaneously quantitatively evaluate the dynamic stability of the respiratory tract and the lung recruitment ability. Existing systems often rely only on a single indicator (such as the mean expiratory flow or airway pressure) for judgment, failing to fully explore the correlation between potential pathological mechanisms such as airway hyperresponsiveness and respiratory muscle fatigue, and it is also difficult to capture the complex features in non-linear and non-stationary respiratory signals. In addition, most models do not consider individual differences, the early warning mechanism is fixed and has poor adaptability, resulting in a high false alarm rate and limited clinical practicability. Therefore, how to construct an asthma health management system with multi-scale feature extraction, adaptive classification judgment, and grading early warning capabilities through advanced signal processing and intelligent modeling technologies has become a technical bottleneck that urgently needs to be broken through in the field of intelligent monitoring of pediatric respiratory diseases. Summary of the Invention

[0005] The purpose of the present invention is to provide a health management system for pediatric asthma patients combined with multi-level supervision to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A health management system for pediatric asthma patients combined with multi-level supervision includes:

[0008] A data acquisition module, which is used to collect the physiological signals of airway resistance and elastic recoil force of the patient in real time;

[0009] A respiratory stability assessment module, which performs time-domain differential processing on the airway resistance signal, calculates the airway response differential frequency eigenvalue according to the fluctuation degree of the resistance change within adjacent time windows, and is used to evaluate the dynamic stability of the respiratory tract of pediatric asthma patients;

[0010] A respiratory mechanics sufficiency assessment module, which analyzes the change amplitude and recovery rate of the elastic recoil force, calculates the respiratory mechanics difference eigenvalue, and is used to evaluate the lung recruitment ability and the respiratory muscle function state;

[0011] An asthma state comprehensive assessment module, which inputs the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue into an asthma severity judgment model for analysis, and outputs the current asthma state level of the child according to the analysis result;

[0012] A warning and intervention module, which triggers a hierarchical warning mechanism based on the asthma state level result and sends intervention measures to the guardians or medical institutions.

[0013] As a further solution of the present invention: the acquisition process of the airway response differential frequency eigenvalue is as follows:

[0014] Real-time collect the airway resistance signal of the patient, perform time-domain differential processing on the airway resistance signal, calculate the airway response differential frequency eigenvalue according to the fluctuation degree of the resistance change within adjacent time windows, and judge whether the airway response differential frequency eigenvalue is greater than or equal to a preset threshold. If so, the respiratory tract of the pediatric asthma patient is unstable; if not, the respiratory tract of the pediatric asthma patient is stable.

[0015] As a further solution of the present invention: the acquisition process of the airway response differential frequency eigenvalue is as follows:

[0016] Perform empirical mode decomposition on the real-time collected airway resistance signal to obtain a plurality of intrinsic mode functions and a residual term;

[0017] Perform Hilbert transform on each intrinsic mode function to obtain the corresponding instantaneous phase and instantaneous frequency;

[0018] Within a set time window, calculate the average value of the absolute value of the instantaneous frequency difference between adjacent time periods to obtain the airway response differential frequency eigenvalue.

[0019] As a further solution of the present invention: the evaluation of the lung recruitment ability and the respiratory muscle function state specifically includes:

[0020] Collect the physiological signal of the patient's elastic recoil force in real time, analyze the change amplitude and recovery rate of the elastic recoil force, calculate the eigenvalue of respiratory mechanics difference, and determine whether the eigenvalue of respiratory mechanics difference is greater than or equal to the preset threshold. If so, the pulmonary recruitment ability and the respiratory muscle function state are abnormal. If not, the pulmonary recruitment ability and the respiratory muscle function state are normal.

[0021] As a further solution of the present invention: the process of obtaining the eigenvalue of respiratory mechanics difference is as follows:

[0022] Preprocess the physiological signal of the patient's elastic recoil force collected in real time. The preprocessing includes baseline drift removal, high-frequency noise filtering, and normalization processing to obtain a normalized signal.

[0023] Apply the discrete Haar wavelet transform to the normalized signal for multi-layer decomposition to obtain the approximation coefficients and detail coefficients of each layer.

[0024] Calculate the corresponding energy for the detail coefficients of each layer.

[0025] Based on the energies of each layer, calculate the ratio of the energy of each layer to the total energy of all layers as the energy ratio.

[0026] Calculate the mean value of the energy ratios at all scales to obtain the eigenvalue of respiratory mechanics difference.

[0027] As a further solution of the present invention: inputting the airway response differential frequency eigenvalue and the eigenvalue of respiratory mechanics difference into the asthma severity judgment model for analysis specifically includes:

[0028] Obtain the airway response differential frequency eigenvalue and the eigenvalue of respiratory mechanics difference of the patient, construct the airway response differential frequency eigenvalue and the eigenvalue of respiratory mechanics difference into a comprehensive feature vector as the input of the asthma severity judgment model, use minimizing the error between the predicted asthma state score and the actual asthma state score as the training objective of the model, and output the asthma state score of the patient according to the trained model. The asthma severity judgment model is a random forest model.

[0029] As a further solution of the present invention: the training process of the asthma severity judgment model is as follows:

[0030] Construct an asthma severity judgment model with the random forest algorithm as the core. The model consists of multiple decision trees. Each decision tree is split and trained based on the sample data corresponding to the input comprehensive feature vector. The CART algorithm is used to optimize the node division, and the generalization ability of the model is improved through the bootstrap sampling and feature random selection mechanism. During the training process, the mean square error of minimizing the predicted asthma status score and the actual asthma status score is used as the objective function. Finally, the trained asthma severity judgment model can automatically output the predicted value of the patient's asthma status score according to the newly input comprehensive feature vector.

[0031] As a further solution of the present invention: according to the analysis result, output the current asthma status level of the child, specifically including:

[0032] Judge whether the asthma status score of each pediatric patient is greater than or equal to a preset first threshold. If so, the asthma status of the corresponding pediatric patient is at a severe level. If not, judge whether the asthma status score of each pediatric patient is less than or equal to a preset second threshold. If so, the asthma status of the corresponding pediatric patient is at a mild level. If not, the asthma status of the corresponding pediatric patient is at a general level.

[0033] As a further solution of the present invention: based on the asthma status level result, trigger a grading warning mechanism, specifically including:

[0034] When it is determined that the asthma status score of a child is at a severe level, the system automatically activates a first-level red warning, generates an emergency notification warning including the child's current asthma score and medical intervention measures, and pushes the notification to the guardian's terminal device and the remote monitoring platform of the designated medical institution in real time through the wireless communication module;

[0035] When it is determined that the asthma status score of a child is at a general level, the system activates a second-level yellow warning, sends a mild abnormality warning to the guardian, and strengthens observation and environmental regulation. If the asthma status score is at a mild level, no active warning is triggered, and only the evaluation result of this time is recorded in the system background and a periodic trend analysis is performed.

[0036] The beneficial effects of the present invention:

[0037] (1) By systematically collecting key physiological signals such as airway resistance and elastic recoil force of children with asthma, the present invention constructs a multi-dimensional evaluation system that integrates respiratory dynamic stability and respiratory mechanics sufficiency, comprehensively reflecting the functional state of the children's respiratory tract and the lung recruitment ability. In terms of respiratory stability evaluation, the empirical mode decomposition combined with the Hilbert transform is used to perform high-time-frequency resolution analysis on the airway resistance signal, extract the instantaneous frequency change information, and further calculate the mean value of the frequency fluctuations in adjacent time periods to form the "airway response differential frequency eigenvalue" with high sensitivity, thus realizing the accurate identification of the airway hyperresponsiveness state and providing a reliable basis for the early warning of asthma attacks. In terms of respiratory mechanics evaluation, the discrete Haar wavelet transform is introduced to perform multi-scale decomposition on the elastic recoil force signal, obtain the energy distribution characteristics in different frequency bands, and extract the "respiratory mechanics difference eigenvalue" based on the energy ratio method to quantify the elastic recovery ability of the lung tissue and the functional state of the respiratory muscles, which has good interpretability and clinical applicability. The above two types of eigenvalues are fused to form a comprehensive feature vector, which is sent as an input variable into the asthma severity judgment model trained based on the random forest algorithm. By minimizing the error between the predicted score and the actual score, the intelligent grading judgment of the asthma state is realized. This evaluation mechanism breaks through the limitations of the traditional method that relies on a single indicator or subjective experience to judge the condition, significantly improves the objectivity, accuracy and individual adaptability of the evaluation results, constructs a complete closed-loop system from physiological signal acquisition, feature extraction, intelligent modeling to clinical decision support, and has important clinical application value and broad popularization prospects.

[0038] (2) The present invention innovatively introduces an asthma severity judgment model based on the random forest algorithm, constructs a data-driven, highly interpretable intelligent evaluation and early warning mechanism, and significantly improves the scientificity and clinical practicality of childhood asthma management. Specifically, the system integrates the "airway response differential frequency eigenvalue" extracted from the airway resistance signal with the "respiratory mechanics differential eigenvalue" obtained from the elastic recoil force signal to form a comprehensive feature vector with multi-dimensional physiological significance, and uses this as the model input. Through large-scale labeled sample training optimization, the asthma status score output by the model can be highly close to the actual scoring result of the clinician. On this basis, the system automatically divides the asthma severity level according to the scoring results, including severe level, general level and mild level, and triggers a three-level early warning mechanism accordingly: when it is judged to be a severe state, the system activates a first-level red warning, pushes an emergency notification to the guardian and medical institution through the wireless communication module, and links the wearable device to vibrate and sound and light prompts; for the general level, the second-level yellow warning is activated to remind to strengthen observation and intervention; and the mild level only records the trend change, and does not actively alarm, so as to achieve accurate graded early warning and avoid excessive interference. This mechanism not only effectively improves the ability to identify the risk of asthma attacks early, but also realizes data-driven automated clinical decision support, significantly enhancing the real-time response of the system and the targeted nature of intervention. It provides strong technical support for remote monitoring, dynamic management and personalized intervention of childhood asthma, and has good clinical transformation potential and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The present invention will be further described below in conjunction with the accompanying drawings.

[0040] Figure 1 It is a flowchart of a health management system for pediatric asthma patients combined with multi-level supervision of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] See also Figure 1 As shown, the present invention is a health management system for pediatric asthma patients combined with multi-level supervision, including:

[0043] A data acquisition module, wherein the data acquisition module is used to collect the patient's airway resistance and elastic recoil force physiological signals in real time;

[0044] Respiratory stability assessment module, which performs time-domain differential processing on the airway resistance signal, calculates the airway response differential frequency eigenvalue according to the fluctuation degree of resistance change within adjacent time windows, and is used to evaluate the dynamic stability of the respiratory tract of pediatric asthma patients;

[0045] Respiratory mechanics sufficiency assessment module, which analyzes the change amplitude and recovery rate of elastic recoil force, calculates the respiratory mechanics difference eigenvalue, and is used to evaluate the lung recruitment ability and the functional state of respiratory muscles;

[0046] Asthma state comprehensive assessment module, which inputs the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue into the asthma severity judgment model for analysis, and outputs the current asthma state level of the child according to the analysis result;

[0047] Warning and intervention module, which triggers a hierarchical warning mechanism based on the asthma state level result and sends intervention measures to guardians or medical institutions.

[0048] In the data acquisition module, the data acquisition module is used to collect the physiological signals of the patient's airway resistance and elastic recoil force in real time, specifically including:

[0049] Continuously monitor the respiratory state of the child during daily activities or treatment through a wearable respiratory monitoring device, and collect physiological change information related to airway patency; this device is equipped with a highly sensitive pressure sensing unit and an airflow detection component, which can accurately obtain the dynamic changes in the airway during inhalation and exhalation of the child, so as to obtain the original physiological signal reflecting the airway resistance level.

[0050] At the same time, the system also collects physiological signals related to the lung elastic recovery ability through a flexible sensor attached to the chest or abdomen, and records the amplitude and rhythm of chest expansion and retraction during the breathing process; the collected signals are processed through filtering, amplification and analog-to-digital conversion to form digital signals available for subsequent analysis; all data are continuously collected at set time intervals and uploaded to the local processing terminal or the cloud health management system synchronously, providing basic data support for respiratory stability assessment and asthma severity judgment.

[0051] In the respiratory stability assessment module, the respiratory stability assessment module performs time-domain differential processing on the airway resistance signal, calculates the airway response differential frequency eigenvalue according to the fluctuation degree of resistance change within adjacent time windows, and is used to evaluate the dynamic stability of the respiratory tract of pediatric asthma patients, specifically including:

[0052] Collect the airway resistance signal of the patient in real time, perform time-domain difference processing on the airway resistance signal, calculate the airway response differential frequency eigenvalue according to the fluctuation degree of the resistance change within adjacent time windows, and determine whether the airway response differential frequency eigenvalue is greater than or equal to the preset threshold. If so, the respiratory tract of the pediatric asthma patient is unstable; if not, the respiratory tract of the pediatric asthma patient is stable.

[0053] The process of obtaining the airway response differential frequency eigenvalue is as follows:

[0054] Perform empirical mode decomposition on the airway resistance signal collected in real time to obtain multiple intrinsic mode functions and a residue term. The calculation expression is: ;

[0055] In the formula, represents the value of the airway resistance signal at time , represents the time acquisition point, represents the number of intrinsic mode functions, represents the total number of intrinsic mode functions, represents the th intrinsic mode function at time , represents the value of the residue term remaining after times of empirical mode decomposition at time ;

[0056] Perform Hilbert transform on each intrinsic mode function to obtain the corresponding instantaneous phase and instantaneous frequency;

[0057] The calculation expression of the instantaneous phase is: ;

[0058] In the formula, represents the th intrinsic mode function at time instantaneous phase, represents the arctangent function, represents the th intrinsic mode function at time Hilbert transform result;

[0059] The calculation expression of the instantaneous frequency is: ;

[0060] In the formula, represents the th intrinsic mode function at time instantaneous frequency, represents the th intrinsic mode function at time instantaneous phase with respect to time The derivative of represents a constant;

[0061] Within a set time window, calculate the average value of the absolute value of the instantaneous frequency difference between adjacent time periods to obtain the airway response differential frequency eigenvalue. The calculation expression is: ;

[0062] In the formula, represents the airway response differential frequency eigenvalue, represents the number of sampling points within the time window, represents the time interval, represents the th sampling point.

[0063] It should be noted that: Through the respiratory stability evaluation module, the invention deeply analyzes the collected airway resistance signal, which can effectively reflect the dynamic stability changes of the respiratory tract in pediatric asthma patients. This module uses a method combining empirical mode decomposition and Hilbert transform to extract the instantaneous frequency characteristics of the airway resistance signal, and obtains the airway response differential frequency eigenvalue by calculating the fluctuation degree of the frequency difference within adjacent time periods. This eigenvalue can sensitively capture the abnormal changes that occur in the airway in a short time, thereby realizing the accurate assessment of the respiratory tract stable state of the child and providing a reliable basis for the early identification of the risk of asthma attacks.

[0064] In the respiratory mechanics sufficiency evaluation module, the respiratory mechanics sufficiency evaluation module analyzes the change amplitude and recovery rate of the elastic recoil force, calculates the respiratory mechanics difference eigenvalue, and is used to evaluate the lung recruitment ability and the respiratory muscle function state. Specifically, it includes:

[0065] Real-time collect the physiological signal of the patient's elastic recoil force, analyze the change amplitude and recovery rate of the elastic recoil force, calculate the respiratory mechanics difference eigenvalue, and judge whether the respiratory mechanics difference eigenvalue is greater than or equal to a preset threshold. If so, the lung recruitment ability and the respiratory muscle function state are abnormal. If not, the lung recruitment ability and the respiratory muscle function state are normal.

[0066] The acquisition process of the respiratory mechanics difference eigenvalue is as follows:

[0067] Perform preprocessing on the real-time collected physiological signal of the patient's elastic recoil force. The preprocessing includes baseline drift removal, high-frequency noise filtering, and normalization processing to obtain a normalized signal;

[0068] Apply discrete Haar wavelet transform to the normalized signal for multi-layer decomposition to obtain the approximation coefficients and detail coefficients of each layer;

[0069] Calculate the corresponding energy for the detail coefficients of each layer. The calculation expression is: ;

[0070] Among them, represents the energy of the wavelet coefficients of the th layer, represents the number of layers of Haar wavelet decomposition, represents the total number of wavelet coefficients of the th layer, represents the number of wavelet coefficients, represents the th layer and the th wavelet coefficient;

[0071] Based on the energy of each layer, calculate the ratio of the energy of each layer to the total energy of all layers as the energy ratio;

[0072] Calculate the mean value of the energy ratios at all scales to obtain the respiratory mechanics difference eigenvalue.

[0073] It should be noted that: through the respiratory mechanics sufficiency evaluation module, the present invention performs multi-scale feature extraction and quantitative analysis on the physiological signal of the patient's elastic recoil force, and can effectively evaluate the lung recruitment ability and the respiratory muscle function state. This module first performs systematic preprocessing on the collected original signal to remove interference factors and improve the signal quality; then introduces the discrete Haar wavelet transform method to decompose the signal into multiple time-frequency scales, calculates the energy distribution at each scale, and further obtains the average value of the energy ratio as the respiratory mechanics difference eigenvalue. This eigenvalue can comprehensively reflect the dynamic change trend of the elastic recoil force and its recovery ability during the breathing process, so as to realize the objective evaluation of the lung function state of children. The innovation of the present invention lies in the first application of the combination of the Haar wavelet transform and the energy ratio method in the respiratory mechanics evaluation of children with asthma, breaking through the limitations of traditional single-index evaluation, having good time-frequency localization characteristics, and being applicable to the analysis of non-stationary physiological signals; through the comprehensive statistics of the signal energy distribution at different scales, the stability and sensitivity of the evaluation results are improved, which helps to early detect potential problems such as insufficient lung recruitment or respiratory muscle fatigue, providing a new and quantifiable auxiliary diagnosis means for clinical practice, and having significant practical value and promotion prospects.

[0074] In the asthma state comprehensive evaluation module, the asthma state comprehensive evaluation module inputs the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue into the asthma severity judgment model for analysis, and outputs the current asthma state level of the child according to the analysis result, specifically including:

[0075] Obtain the differential frequency eigenvalue of the patient's airway response and the differential eigenvalue of respiratory mechanics. Construct the differential frequency eigenvalue of the airway response and the differential eigenvalue of respiratory mechanics into a comprehensive feature vector as the input of the asthma severity judgment model. Take minimizing the error between the predicted asthma status score and the actual asthma status score as the training objective of the model. According to the trained model, output the asthma status score of the patient. The asthma severity judgment model is a random forest model.

[0076] The training process of the asthma severity judgment model is as follows:

[0077] Construct an asthma severity judgment model with the random forest algorithm as the core. The model consists of multiple decision trees. Each decision tree is split and trained based on the sample data corresponding to the input comprehensive feature vector. The CART algorithm is used to optimize the node division, and the generalization ability of the model is improved through the bootstrap sampling and feature random selection mechanism. During the training process, take minimizing the mean square error between the predicted asthma status score and the actual asthma status score as the objective function. Finally, the trained asthma severity judgment model can automatically output the predicted value of the asthma status score of the patient according to the newly input comprehensive feature vector.

[0078] Judge whether the asthma status score of each pediatric patient is greater than or equal to the preset first threshold. If so, the asthma status of the corresponding pediatric patient is at the severe level. If not, judge whether the asthma status score of each pediatric patient is less than or equal to the preset second threshold. If so, the asthma status of the corresponding pediatric patient is at the mild level. If not, the asthma status of the corresponding pediatric patient is at the general level.

[0079] It should be noted that: Through the asthma status comprehensive evaluation module, the present invention fuses and analyzes the differential frequency eigenvalue of the airway response and the differential eigenvalue of respiratory mechanics, constructs a comprehensive feature vector as the input, introduces an asthma severity judgment model based on the random forest algorithm, and realizes the intelligent and quantitative evaluation of the asthma status of children. During the training process of the model, the bootstrap sampling and feature random selection mechanism are adopted to enhance the generalization ability of the model, and the mean square error between the predicted score and the actual score is minimized as the objective function to ensure that the output result has high accuracy and stability. By setting multiple score thresholds, the system can automatically divide the asthma status levels into severe level, general level and mild level, so as to provide a basis for clinical grading early warning and intervention, and significantly improve the scientificity and efficiency of asthma management. The innovation of the present invention lies in that for the first time, a variety of respiratory physiological characteristic parameters are fused and then input into the random forest model for asthma status modeling and classification, breaking through the traditional evaluation method that relies on a single index or subjective score; this method can effectively capture the complexity and individual differences of asthma conditions, has good robustness and interpretability, is applicable to long-term dynamic monitoring and the deployment of remote health management systems, and has important clinical application value and promotion prospects.

[0080] In the early warning and intervention module, based on the results of the asthma status level, the early warning and intervention module triggers a hierarchical early warning mechanism and sends intervention measures to the guardians or medical institutions, specifically including:

[0081] When it is determined that the asthma status score of a child is at a severe level, the system automatically activates a first-level red early warning, generates an emergency notification message containing the child's current asthma score and medical intervention measures, and pushes the notification to the guardian's terminal device and the remote monitoring platform of the designated medical institution in real time through the wireless communication module; the system controls the wearable monitoring device to emit vibration and sound and light prompts to remind the child and the caregiver to immediately take emergency response measures;

[0082] When it is determined that the asthma status score of a child is at a general level, the system activates a second-level yellow early warning, sends a mild abnormality reminder to the guardian, and strengthens observation and environmental regulation; if the asthma status score is at a mild level, no active early warning is triggered, and only the results of this assessment are recorded in the system background and a periodic trend analysis is performed; the early warning and intervention module also supports dynamically adjusting the threshold parameters according to historical early warning data to implement personalized early warning strategy configuration, thereby improving the adaptability and clinical practicality of the system.

[0083] The working principle of the present invention: The present invention aims to realize real-time monitoring, intelligent assessment and hierarchical intervention of the respiratory status of children with asthma. The system includes a data acquisition module, a respiratory stability assessment module, a respiratory mechanics sufficiency assessment module, an asthma status comprehensive assessment module, and an early warning and intervention module. The airway resistance and elastic recoil force signals of the child are continuously collected through a wearable physiological monitoring device, and the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue are respectively extracted for quantitatively evaluating the dynamic stability of the respiratory tract and the pulmonary function status. Among them, the airway response differential frequency eigenvalue is based on empirical mode decomposition and Hilbert transform technology to capture the instantaneous frequency fluctuations of the airway and reflect the changes in airway reactivity; the respiratory mechanics difference eigenvalue uses Haar wavelet transform and energy ratio method to analyze the multi-scale energy distribution of the elastic recoil force, reflecting the pulmonary recruitment ability and respiratory muscle function. The two eigenvalues are fused to construct a comprehensive feature vector, which is input into the asthma severity judgment model trained based on the random forest algorithm. By minimizing the error between the predicted score and the actual score, accurate hierarchical judgment of the asthma status is realized. The system automatically divides the asthma severity level according to the scoring results and triggers a three-level early warning mechanism, supporting remote notification, emergency prompt and personalized intervention strategy adjustment. The present invention realizes the whole-process closed-loop management from physiological signal acquisition, feature extraction, intelligent modeling to clinical decision support, has high sensitivity, strong adaptability and good clinical practicality, is applicable to the long-term monitoring of children with asthma and the deployment of a multi-level medical supervision system, and significantly improves the scientific and intelligent level of disease management.

[0084] The above has described in detail an embodiment of the present invention, but the above content is only a preferred embodiment of the present invention and cannot be considered as defining the scope of implementation of the present invention. All equivalent changes and improvements made in accordance with the scope of the application of the present invention shall still fall within the scope covered by the patent of the present invention.

Claims

1. A health management system for children with asthma combined with multi-level supervision, characterized in that, Including: A data acquisition module, which is used to collect the physiological signals of the airway resistance and elastic recoil force of the patient in real time; A respiratory stability assessment module, which performs time-domain differential processing on the airway resistance signal, calculates the airway response differential frequency eigenvalue according to the fluctuation degree of the resistance change within adjacent time windows, and is used to evaluate the dynamic stability of the respiratory tract of pediatric asthma patients; A respiratory mechanics sufficiency assessment module, which analyzes the change amplitude and recovery rate of the elastic recoil force, calculates the respiratory mechanics difference eigenvalue, and is used to evaluate the lung recruitment ability and the respiratory muscle function status; An asthma state comprehensive assessment module, which inputs the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue into an asthma severity judgment model for analysis, and outputs the current asthma state grade of the child according to the analysis result; An early warning intervention module, which triggers a hierarchical early warning mechanism based on the asthma state grade result and sends intervention measures to the guardian or medical institution.

2. The health management system for pediatric asthma patients integrating multi-level supervision according to claim 1, wherein, The acquisition process of the airway response differential frequency eigenvalue is as follows: Collect the airway resistance signal of the patient in real time, perform time-domain differential processing on the airway resistance signal, calculate the airway response differential frequency eigenvalue according to the fluctuation degree of the resistance change within adjacent time windows, and judge whether the airway response differential frequency eigenvalue is greater than or equal to a preset threshold. If so, the respiratory tract of the pediatric asthma patient is unstable; if not, the respiratory tract of the pediatric asthma patient is stable.

3. The health management system for pediatric asthma patients integrating multi-level supervision according to claim 2, characterized in that, The acquisition process of the airway response differential frequency eigenvalue is as follows: Perform empirical mode decomposition on the real-time collected airway resistance signal to obtain a plurality of intrinsic mode functions and a residual term; Perform Hilbert transform on each intrinsic mode function to obtain the corresponding instantaneous phase and instantaneous frequency; Within a set time window, calculate the average value of the absolute value of the instantaneous frequency difference between adjacent time periods to obtain the airway response differential frequency eigenvalue.

4. A health management system for pediatric asthma patients integrating multi-level supervision according to claim 1, characterized in that, The evaluation of the lung recruitment ability and the respiratory muscle function status specifically includes: Collect the elastic recoil force physiological signal of the patient in real time, analyze the change amplitude and recovery rate of the elastic recoil force, calculate the respiratory mechanics difference eigenvalue, and judge whether the respiratory mechanics difference eigenvalue is greater than or equal to a preset threshold. If so, the lung recruitment ability and the respiratory muscle function status are abnormal; if not, the lung recruitment ability and the respiratory muscle function status are normal.

5. The health management system for pediatric asthma patients integrating multi-level supervision according to claim 4, characterized in that, The acquisition process of the respiratory mechanics difference eigenvalue is as follows: Perform preprocessing on the real-time collected elastic recoil force physiological signal of the patient. The preprocessing includes baseline drift removal, high-frequency noise filtering, and normalization processing to obtain a normalized signal; Apply discrete Haar wavelet transform to the normalized signal for multi-layer decomposition to obtain the approximation coefficient and detail coefficient of each layer; Calculate the corresponding energy for the detail coefficient of each layer; Based on the energy of each layer, calculate the ratio of the energy of each layer to the total energy of all layers as the energy ratio; Calculate the mean value of the energy ratios at all scales to obtain the respiratory mechanics difference eigenvalue.

6. The health management system for pediatric asthma patients combining multi-level supervision according to claim 1, characterized in that, The inputting of the airway response differential frequency eigenvalue and the respiratory mechanics difference eigenvalue into the asthma severity judgment model for analysis specifically includes: Obtain the differential frequency eigenvalue of the patient's airway response and the differential eigenvalue of respiratory mechanics, construct the differential frequency eigenvalue of the airway response and the differential eigenvalue of respiratory mechanics into a comprehensive feature vector, which is used as the input of the asthma severity judgment model. Take minimizing the error between the predicted asthma status score and the actual asthma status score as the training objective of the model, and output the asthma status score of the patient according to the trained model. The asthma severity judgment model is a random forest model.

7. A health management system for pediatric asthma patients integrating multi-level supervision according to claim 6, characterized in that, The training process of the asthma severity judgment model is as follows: Construct an asthma severity judgment model with the random forest algorithm as the core. The model consists of multiple decision trees. Each decision tree is split and trained based on the sample data corresponding to the input comprehensive feature vector. The CART algorithm is used to optimize the node division, and the generalization ability of the model is improved through the bootstrap sampling and feature random selection mechanism. During the training process, the mean square error between the predicted asthma status score and the actual asthma status score is used as the objective function. Finally, the trained asthma severity judgment model can automatically output the predicted value of the asthma status score of the patient according to the newly input comprehensive feature vector.

8. A health management system for pediatric asthma patients integrating multi-level supervision according to claim 1, characterized in that, According to the analysis results, output the current asthma status level of the child, specifically including: Judge whether the asthma status score of each pediatric patient is greater than or equal to a preset first threshold. If so, the asthma status of the corresponding pediatric patient is at a severe level. If not, judge whether the asthma status score of each pediatric patient is less than or equal to a preset second threshold. If so, the asthma status of the corresponding pediatric patient is at a mild level. If not, the asthma status of the corresponding pediatric patient is at a general level.

9. A health management system for pediatric asthma patients incorporating multi-level supervision according to claim 1, characterized in that, Based on the asthma status level result, trigger a hierarchical warning mechanism, specifically including: If it is determined that the asthma status score of a child is at a severe level, the system automatically activates a first-level red warning, generates an emergency notice warning containing the current asthma score of the child and medical intervention measures, and pushes the notice to the guardian's terminal device and the remote monitoring platform of the designated medical institution in real time through the wireless communication module; If it is determined that the asthma status score of a child is at a general level, the system activates a second-level yellow warning, sends a mild abnormality warning to the guardian, and strengthens observation and environmental control. If the asthma status score is at a mild level, no active warning is triggered, and only the evaluation result of this time is recorded in the system background and a periodic trend analysis is carried out.

Images