Monitoring system of newborn breathing machine

Abstract

The invention relates to a monitoring system of a newborn breathing machine. Comprising the following steps: by monitoring airway pressure, tidal volume and respiratory frequency parameters of a newborn in real time and combining individual physiological characteristics of the newborn, constructing a respiratory mode model based on physiological characteristics; the respirator is adjusted by adopting a self-adaptive control algorithm based on real-time data feedback, and parameters of the respirator are dynamically adjusted; establishing a pressure and volume relation model by combining the airway resistance and lung elasticity of the newborn; calculating data of different development stages of the lungs of the newborn, and obtaining a relational data value between pressure and volume by adopting a nonlinear regression method; the respirator needs to calculate the current airway pressure and tidal volume, and also needs to calculate the balance between optimized pressure and volume according to dynamic characteristics of lung rebound resilience and airway ventilation; based on the change of respiratory frequency, airway pressure fluctuation and tidal volume, the respiratory driving force and rhythm of the newborn are monitored in real time, and whether the newborn enters an autonomous respiration state or not is judged.

Description

Technical Field

[0001] The invention relates to a monitoring system for a neonatal ventilator. Background Art

[0002] Although the current monitoring system for neonatal ventilators can provide effective support to a certain extent, it still has some significant deficiencies and drawbacks in actual clinical applications, especially in real-time monitoring, accurate diagnosis and individualized treatment. According to the description of the intelligent neonatal respiratory monitoring system proposed by the Chinese invention patent (patent number: 202410502308X), its goal is to make up for the inaccuracy of traditional monitoring systems in discovering problems and adjusting the timing of treatment. The deficiencies of the existing technology are mainly reflected in the following aspects.

[0003] First, traditional neonatal ventilator monitoring systems usually use fixed monitoring standards and evaluation criteria. This standardized method often does not show its due accuracy when dealing with individual differences between different newborns. Each newborn has obvious differences in physiological conditions, lung development, breathing habits, airway characteristics, etc. Therefore, unified evaluation standards often cannot fully adapt to these differences. The existing monitoring system does not fully consider individual differences in the respiratory characteristics of newborns, resulting in untimely and inaccurate numerical evaluation. Due to the different developmental states of lung function in each newborn, standardized parameters are prone to miss the best treatment time or fail to effectively respond to emergencies. For example, for premature newborns, their lungs are not fully developed and their respiratory systems are unstable, while other healthy newborns may have completely different physiological needs. If the ventilator cannot be flexibly adjusted according to these individual differences, overventilation or underventilation may occur, causing damage to the newborn or delaying the treatment time. Secondly, the monitoring technology of the traditional system lacks real-time dynamic feedback and adaptive adjustment mechanisms, making it difficult to cope with the dynamic changes in the developmental stages of the newborn's lungs. Physiological characteristics of newborns, such as airway resistance and lung elasticity, change over time, especially in premature infants or newborns with lung diseases. Most traditional ventilator monitoring systems rely on fixed parameters or overly simplified models, and cannot effectively adjust treatment parameters such as airway pressure and tidal volume in real time, resulting in the inability to adjust treatment plans in time according to changes in the lung condition of the newborn. Even if some devices are equipped with basic pressure-volume adjustment functions, this adjustment is usually based on a preset linear or approximate linear model, and fails to deeply consider the nonlinear characteristics of the newborn's lungs and their dynamic needs at different stages of development. Therefore, if the resilience of the lungs changes suddenly and the airway resistance changes, the traditional monitoring system may not be able to respond in time, resulting in the ventilator being unable to accurately match the current physiological needs, thereby affecting the treatment effect and the life safety of the newborn.

[0004] Third, existing technologies generally lack intelligent prediction and early warning mechanisms. Traditional monitoring systems often rely on real-time physiological data collection and analysis and feedback based on simple algorithms, but do not have good predictive capabilities for future respiratory needs. Many traditional systems only sound alarms when obvious physiological abnormalities occur, and when the respiratory characteristics of newborns begin to change slightly, the system may miss the best time to intervene. The diagnostic process of traditional ventilator monitoring is usually lagging, relying on real-time data input and subsequent processing, and lacks learning and prediction capabilities based on historical data, which makes it impossible for the ventilator to predict possible sudden situations such as respiratory arrest and suffocation in advance. Therefore, traditional systems do not have strong foresight and predictability in disease management, and are prone to miss some early risk signals, resulting in untimely or inaccurate treatment. Summary of the invention

[0005] The purpose of the present invention is to provide a monitoring system for a neonatal ventilator, thereby solving some of the drawbacks and deficiencies pointed out in the background technology.

[0006] The present invention solves the above-mentioned technical problems by adopting the following technical solution, which includes the following steps:

[0007] S1. Real-time physiological dynamic adaptation mechanism of neonatal ventilator:

[0008] S1.1. By real-time monitoring of the airway pressure, tidal volume, and respiratory rate parameters of the newborn, combined with the individual physiological characteristics of the newborn, a breathing pattern model based on physiological characteristics is constructed;

[0009] S1.2. The ventilator is regulated by an adaptive control algorithm based on real-time data feedback to dynamically adjust the ventilator parameters;

[0010] S2. Nonlinear modeling and dynamic optimization of the relationship between ventilator pressure and volume:

[0011] S2.1. Establish a pressure-volume relationship model by combining the airway resistance and lung elasticity of the newborn; calculate the data of different development stages of the newborn's lungs, and use nonlinear regression method to obtain the relationship data value between pressure and volume;

[0012] S2.2, the ventilator shall calculate the current airway pressure and tidal volume, and shall also calculate the balance between pressure and volume based on the dynamic characteristics of lung resilience and airway ventilation;

[0013] S3, Intelligent switching mechanism between ventilator-assisted and spontaneous breathing:

[0014] S3.1. Based on the changes in respiratory rate, airway pressure fluctuations and tidal volume, monitor the respiratory drive and rhythm of the newborn in real time to determine whether the newborn has entered the state of spontaneous breathing; by analyzing the airway pressure waveform and respiratory cycle characteristics, identify the timing of the newborn's spontaneous breathing;

[0015] S4. Adaptive optimization of ventilator parameters and individualized treatment pathways:

[0016] S4.1. Establish a treatment model based on individual differences by continuously collecting the heart rate, blood oxygen saturation, and airway pressure data of the newborn, combined with birth weight, gestational age, and disease background factors; the treatment model guides the individualized setting of the ventilator;

[0017] S4.2. Use machine learning algorithms to dynamically optimize individual ventilator parameters based on historical data and real-time monitoring results; adjust each output value of the ventilator through the algorithm.

[0018] Furthermore, the method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator includes:

[0019] A dynamic equation is used to describe the relationship between airway pressure and time; the dynamic equation calculates the current physiological state of the newborn and also makes adaptive adjustments based on birth weight and gestational age factors; the equation is as follows:

[0020]

[0021] in:

[0022] P(t) represents the airway pressure at time t; A, B, C, D, and E are constant coefficients, representing parameters calculated based on individual physiological characteristics of newborns including weight, gestational age, and lung function; exp(D·t) is an exponential function, representing the nonlinear change of airway pressure over time; the exponential function is used to simulate the rapid change of airway pressure during breathing; t is a time variable, representing the change of airway pressure monitored at a specific time point; and E is a constant bias term used to adjust the final airway pressure.

[0023] Furthermore, the method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator includes:

[0024] Use convolutional neural network CNN to predict future respiratory needs based on the real-time physiological data of the newborn and adjust the ventilator parameters in advance; introduce the prediction model, and the output is the predicted value of future respiratory needs. The formula is as follows:

[0025]

[0026] in:

[0027] represents the predicted respiratory demand at time t; X i (t) represents the value of the ith physiological parameter at time t; physiological parameters include airway pressure, tidal volume, blood oxygen saturation, heart rate, etc.; each parameter X i (t) reflects the immediate physiological condition of the newborn; W i is the weight coefficient obtained during the deep learning model training process; each weight W i It indicates the influence of the physiological parameter on the final prediction result; b is the bias term; f is the activation function that determines the nonlinear characteristics of the model output; N is the dimension of the input physiological data, indicating how many orders of magnitude of physiological parameters the model uses for prediction.

[0028] Furthermore, the method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator includes:

[0029] The multi-parameter data fusion technology is introduced to reflect the real needs of newborns by jointly analyzing multiple physiological data; the multi-dimensional physiological data are integrated and the data are synthesized using a weighted average method to obtain a comprehensive physiological state map; the formula is as follows:

[0030]

[0031] in:

[0032] Ψ(t) represents the comprehensive physiological state spectrum at time t; X k (t) represents the value of the kth physiological parameter at time t; physiological parameters include airway pressure, tidal volume, respiratory rate, and blood oxygen saturation; α k is the weighting coefficient, which indicates the contribution of each physiological parameter to the comprehensive physiological state; β is the bias term, which is used to adjust the output of the atlas; M represents the total number of physiological parameters involved in the fusion.

[0033] Furthermore, the nonlinear modeling and dynamic optimization method of the relationship between the pressure and volume of the ventilator:

[0034] Through individualized physiological data, combined with the airway resistance and lung elasticity of the newborn, a nonlinear dynamic pressure-volume relationship model is constructed, and the balance of pressure and volume is adjusted in real time; according to the different stages of lung development and physiological state of the newborn, the output of the ventilator is optimized; the airway resistance (R aw ) and lung elasticity (C l ), combined with real-time monitoring of neonatal respiratory data, to dynamically calculate the balance between airway pressure and tidal volume:

[0035]

[0036] in:

[0037] Paw (t) is the airway pressure of the newborn at the current moment, in cmH2O, indicating the pressure of the ventilator output gas; V t (t) is the tidal volume at the current moment, in ml, indicating the amount of gas in and out during each breath; C l (t) is the lung elasticity of the newborn, in ml / cmH2O, which indicates the lung's ability to expand the volume of gas; R aw (t) is the airway resistance of the newborn, the unit is cmH2O / L / s, which describes the airway resistance to airflow; γ(t) is the adjustment coefficient reflecting the lung development and physiological status of the newborn.

[0038] Furthermore, the nonlinear modeling and dynamic optimization method of the relationship between the pressure and volume of the ventilator:

[0039] By real-time monitoring of the lung elasticity and airway resistance of newborns, combined with physiological characteristics, dynamic calculation of lung resilience and airway ventilation can be performed to optimize the ventilator treatment strategy; the following function model is used to describe the characteristics of lung elasticity and airway resistance changing over time:

[0040] C l (t) = C l0 ·exp(-αt)

[0041] R aw (t) = R aw0 (1+βt)

[0042] in:

[0043] C l0 is the lung elastic constant of the newborn at birth, indicating the initial lung volume expansion capacity; α is the coefficient describing the attenuation of lung elasticity over time, reflecting the speed of gradual lung development and perfection; R aw0 is the airway resistance constant of the newborn at birth, indicating the initial airway resistance; β is the coefficient describing the gradual increase of airway resistance as the newborn grows; t is the time variable, indicating the developmental stage of the newborn, in hours or days.

[0044] Furthermore, the nonlinear modeling and dynamic optimization method of the relationship between the pressure and volume of the ventilator:

[0045] Through the nonlinear regression algorithm, based on the real-time collection of neonatal physiological data including airway pressure and tidal volume, the relationship between pressure and volume is optimized, and the following nonlinear regression model is used to describe the relationship between pressure and volume:

[0046] P aw (t) = A·(V t (t)) α +B·exp(-γV t(t))

[0047] in:

[0048] A is the amplification coefficient in the regression model, which reflects the strength of the relationship between pressure and tidal volume; B is the constant term in the regression model, which represents the basic relationship between pressure and tidal volume; α is the exponential coefficient, which describes the nonlinear relationship between pressure and tidal volume; γ is the attenuation constant, which describes the dynamic effect of lung elasticity and airway resistance; V t (t) is the tidal volume, in ml, which indicates the amount of gas in and out during each breath.

[0049] The monitoring system of the neonatal ventilator of the present invention has multiple beneficial effects, which significantly improves the accuracy and safety of neonatal respiratory support, which are specifically embodied in the following aspects:

[0050] By real-time monitoring of the newborn's physiological parameters (such as airway pressure, tidal volume, blood oxygen saturation, heart rate, etc.), and combining individual differences (such as birth weight, gestational age, lung development status, etc.), this system can customize a personalized respiratory support plan for each newborn. Unlike traditional standardized treatment methods, this personalized treatment can accurately adapt to the special physiological needs of each newborn, thereby improving the effectiveness of treatment.

[0051] The system uses advanced algorithms such as nonlinear regression models, convolutional neural networks (CNNs), and multi-parameter data fusion technology to dynamically adjust ventilator parameters based on real-time collected physiological data. In particular, when the lungs of newborns are at different stages of development and airway resistance and lung elasticity change, the system can quickly respond and optimize treatment parameters such as airway pressure and tidal volume to ensure that respiratory support is always in the optimal state. This real-time adjustment mechanism effectively avoids the adverse consequences of traditional ventilators when the airway pressure is too high or too low, such as pneumothorax and lung damage.

[0052] Through convolutional neural networks (CNN) and machine learning algorithms, the future respiratory needs of newborns can be predicted in advance based on historical data and real-time monitoring results. By predicting future needs, the ventilator can automatically adjust the output parameters to provide more accurate support for newborns and avoid emergencies caused by untimely adjustment of respiratory parameters. This predictive mechanism makes the ventilator not only a reactive treatment tool, but also an intelligent preventive treatment system. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The present invention is a flow chart of a monitoring system of a neonatal ventilator.

[0054] Figure 2 The present invention is a flowchart of a method for constructing a real-time physiological dynamic adaptation mechanism of a neonatal ventilator.

[0055] Figure 3 The present invention is a flow chart of the nonlinear modeling and dynamic optimization method of the relationship between pressure and volume of the ventilator. DETAILED DESCRIPTION

[0056] The specific implementation modes of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0057] The core goal of the monitoring system of the neonatal ventilator is to provide accurate respiratory support for the newborn, ensure the maximization of gas exchange efficiency, and avoid excessive or insufficient respiratory support. Traditional ventilators often rely on fixed standard models for parameter setting, but the physiological characteristics of newborns (such as airway pressure, tidal volume, respiratory rate, etc.) and lung development status vary greatly, and standard models often cannot effectively adapt to these changes. Therefore, the monitoring system based on real-time monitoring data and individualized physiological characteristics can dynamically adjust the working parameters of the ventilator according to the different physiological needs and development stages of newborns, and provide more personalized and accurate respiratory support. The key to the implementation of this system is to monitor the respiratory data of the newborn such as airway pressure, tidal volume, respiratory rate, etc. in real time, and combine the physiological characteristics of the newborn such as birth weight, gestational age, lung elasticity, airway resistance, etc. to build a respiratory pattern model based on physiological characteristics. By collecting these multi-dimensional physiological data, the current lung condition of the newborn can be accurately described and its future respiratory needs can be predicted.

[0058] In order to ensure that the parameters of the ventilator can flexibly adapt to these changes, the system uses an adaptive control algorithm to dynamically adjust the output of the ventilator based on real-time feedback data. Specifically, when the system detects abnormal changes in airway pressure, tidal volume or respiratory rate, it immediately adjusts key parameters such as pressure and frequency by comparing them with the physiological characteristic model. For example, if the airway pressure of the newborn is detected to be too high, the system will automatically adjust the tidal volume or adjust the airway pressure to ensure that gas exchange during breathing is not disturbed and to avoid damage to the newborn's lungs. Through this mechanism of real-time data feedback and dynamic adjustment, the monitoring system of the neonatal ventilator can maintain an accurate response to the breathing needs of the newborn throughout the treatment process, thereby providing personalized and efficient treatment plans.

[0059] During the breathing process of newborns, due to the incomplete development of their lungs, airway resistance and lung elasticity have great individual differences, so it is necessary to dynamically adjust the parameters of the ventilator according to the specific physiological characteristics of the newborn. The system constructs a relationship model between pressure and volume by combining the airway resistance and lung elasticity of the newborn. Airway resistance and lung elasticity are key physiological parameters that affect the support effect of the ventilator for newborns. Airway resistance describes the resistance of the airway to airflow, while lung elasticity describes the volume expansion capacity of the lungs to gas. These two parameters fluctuate with the changes in the lung development and other physiological characteristics of the newborn. Therefore, the traditional fixed parameter model cannot effectively cope with these changes. In this system, by real-time monitoring of respiratory parameters such as airway pressure and tidal volume of newborns, combined with the physiological characteristics of newborns, a more dynamic and personalized pressure-volume relationship model is established.

[0060] This model is not fixed, but dynamically adjusts the balance between airway pressure and tidal volume according to the different development stages and physiological states of the newborn's lungs. In order to further improve the accuracy and reliability of this model, this system uses a nonlinear regression method to obtain the data value of the relationship between pressure and volume. The traditional linear regression method is difficult to adapt to the complex nonlinear changes in the development of the newborn's lungs, while the nonlinear regression method can better capture the nonlinear relationship between airway pressure and tidal volume, thereby improving the accuracy and response speed of the model. Through this nonlinear regression model, the system can evaluate the lung development process of the newborn in real time, and timely calculate the current airway pressure, tidal volume, lung resilience and airway ventilation characteristics of the newborn. These data help the system optimize the balance between pressure and volume in real time during each respiratory cycle, ensuring that the ventilator always provides the most suitable respiratory support for the newborn. Specifically, when the system detects that the airway pressure is too high or the tidal volume is too low, it will automatically adjust the airway pressure according to dynamic characteristics such as lung resilience and airway ventilation to avoid damage to the newborn's lungs.

[0061] At different physiological stages, especially just after being weaned from mechanical ventilation, newborns may gradually transition to a state of spontaneous breathing. At this time, the ventilator needs to dynamically judge their respiratory needs based on their respiratory drive and rhythm changes, and make corresponding adjustments. To achieve this goal, the system relies on real-time monitoring of the newborn's respiratory rate, airway pressure fluctuations, and tidal volume changes. Through the precise collection of these parameters, the system can evaluate the newborn's respiratory drive and rhythm in real time, and then analyze whether it has entered a state of spontaneous breathing. First of all, respiratory rate and tidal volume are important indicators for judging respiratory drive. Changes in respiratory rate can reflect the rhythmicity of the newborn's breathing. When the frequency exceeds a certain threshold, it may mean that the newborn has begun to rely on spontaneous breathing, and the ventilator should reduce its ventilation support. Changes in tidal volume are directly related to the amount of gas exchange per breath of the newborn. The fluctuation of this parameter can reveal the newborn's breathing depth. If the tidal volume gradually returns to the normal physiological range, it indicates the possibility of spontaneous breathing.

[0062] At the same time, airway pressure fluctuations are an important factor in assessing the respiratory status of newborns. The airway pressure waveform can reflect the matching of the ventilator and the newborn's breathing. Especially when the newborn gradually recovers spontaneous breathing, the airway pressure waveform often shows periodic fluctuations, and the amplitude and shape of the waveform may be different from the mechanical ventilation mode. Therefore, the system not only identifies the timing of spontaneous breathing by monitoring the changes in these physiological parameters, but also determines whether the newborn has sufficient spontaneous breathing ability by analyzing the characteristics of the airway pressure waveform and the respiratory cycle. The characteristic analysis of the airway pressure waveform can help the system distinguish the different manifestations between mechanical ventilation and spontaneous breathing. For example, mechanical ventilation may present a stable pressure waveform, while spontaneous breathing will show a certain volatility. Therefore, by real-time analysis and comprehensive judgment of these data, the system can adjust the working mode of the ventilator in time when the newborn begins to enter the state of spontaneous breathing, for example, reducing the support of mechanical ventilation, switching to assisted ventilation mode or completely offline ventilation, so as to provide the most suitable respiratory support for the newborn.

[0063] The physiological characteristics, lung development status, and clinical background of newborns will affect their needs for ventilators. Therefore, traditional ventilators cannot meet the unique needs of each newborn. The treatment model based on individual differences can accurately reflect the real-time physiological state of the newborn and tailor respiratory support for it. By comprehensively analyzing these physiological data and individual characteristics, the system can establish a dynamic and personalized treatment model to ensure that each adjustment of the ventilator can better match the needs of the newborn. Specifically, heart rate and blood oxygen saturation are important indicators for assessing the systemic oxygenation status and circulatory function of the newborn. These data can reflect the current vital signs of the newborn and its need for ventilation support; airway pressure is directly related to the ventilation mode and respiratory support effect of the ventilator. It can help determine whether the newborn is currently receiving adequate mechanical ventilation support or is close to spontaneous breathing.

[0064] Combined with these data, the system can dynamically optimize the working parameters of the ventilator, such as airway pressure, tidal volume, respiratory rate, etc., according to the individual physiological characteristics of each newborn, to ensure the personalization and accuracy of the treatment plan. In addition, the system also uses machine learning algorithms to achieve dynamic optimization based on historical data and real-time monitoring results. Machine learning can establish a predictive model for neonatal respiratory support through training and pattern recognition of a large amount of historical data, and continuously adjust and optimize the ventilator parameters based on real-time monitoring results. When the physiological state of the newborn changes, the system can automatically adjust each output value of the ventilator based on the feedback of the machine learning algorithm on the historical data to maintain the most suitable respiratory needs of the newborn. For example, if the blood oxygen saturation of the newborn is monitored to decrease, the machine learning algorithm can determine whether it is necessary to increase the airway pressure or adjust the parameters such as tidal volume through the analysis of historical data to ensure that the oxygenation state of the newborn is effectively improved. At the same time, through continuous optimization, the system can dynamically adjust the treatment strategy according to the long-term physiological changes and historical treatment responses of the newborn, so as to achieve a more refined and personalized respiratory support effect, thereby improving the safety and effectiveness of the treatment process of the newborn.

[0065] Embodiment 1:

[0066] In the real-time physiological dynamic adaptation mechanism of the neonatal ventilator, the system uses a dynamic equation to describe the relationship between airway pressure and time to achieve personalized ventilator parameter adjustment. Specifically, the form of the equation is:

[0067]

[0068] Among them, P(t) represents the airway pressure at time t, and A, B, C, D and E are constant coefficients calculated according to the individual physiological characteristics of the newborn (including birth weight, gestational age, lung function, etc.). The specific value range of these coefficients varies according to the individual differences and physiological status of the newborn, and usually needs to be optimized through clinical data and machine learning algorithms. The exponential function exp(D·t) is used to simulate the nonlinear fluctuations of airway pressure over time, reflecting the rapid changes in pressure during breathing. The constant term E is used as a bias term to adjust the final value of the airway pressure. Next, a specific example is used to illustrate how to apply this dynamic equation, and each step is explained in detail.

[0069] A premature baby with a birth weight of 1500 grams and a gestational age of 32 weeks needs respiratory support. These physiological characteristics have been analyzed through clinical data and machine learning algorithms, and the corresponding parameter ranges have been determined. According to the physiological state of the newborn, the system will dynamically calculate the corresponding parameters, and these parameters are set through clinical experiments to obtain the following values:

[0070] A=30

[0071] B=10

[0072] C=15

[0073] D=0.05

[0074] E=1.5

[0075] Now we need to calculate the airway pressure at different time points using these parameters. We choose to calculate the airway pressure at time t=1 second, t=2 seconds, and t=3 seconds respectively.

[0076] Step 1: Calculate the airway pressure at t = 1 second

[0077] Substituting into the formula, we get:

[0078]

[0079] First calculate the exponential function part:

[0080] exp(0.05·1)=exp(0.05)≈1.05127

[0081] Then substitute into the equation:

[0082]

[0083] P(1)≈1.164+1.5=2.664

[0084] Therefore, at t = 1 second, the airway pressure is approximately 2.664 cmH2O.

[0085] Step 2: Calculate the airway pressure at t = 2 seconds

[0086] Substituting into the formula, we get:

[0087]

[0088] First calculate the exponential function part:

[0089] exp(0.05·2)=exp(0.1)≈1.10517

[0090] Then substitute into the equation:

[0091]

[0092] P(2)≈1.13+1.5=2.63

[0093] Therefore, at t = 2 seconds, the airway pressure is approximately 2.63 cmH2O.

[0094] Step 3: Calculate the airway pressure at t = 3 seconds

[0095] Substituting into the formula, we get:

[0096]

[0097] First calculate the exponential function part:

[0098] exp(0.05·3)=exp(0.15)≈1.16183

[0099] Then substitute into the equation:

[0100]

[0101] P(3)≈1.095+1.5=2.595

[0102] Therefore, at t = 3 seconds, the airway pressure is approximately 2.595 cmH2O.

[0103] Through the above steps, combined with the physiological data of the newborn with a birth weight of 1500 grams and a gestational age of 32 weeks, the airway pressure at different time points was successfully calculated. As time goes by, the airway pressure shows a nonlinear trend of change, which is exactly the airway pressure fluctuation during the breathing process of the newborn simulated by the exponential function.

[0104] This embodiment uses a convolutional neural network (CNN) to predict future respiratory needs and adjust the real-time physiological dynamic adaptation mechanism of ventilator parameters in advance. In order to explain this mechanism, actual data will be introduced and examples will be used to demonstrate the feasibility of the entire step. First, briefly review the formula and its meaning:

[0105] The given formula is:

[0106]

[0107] in,

[0108] is the predicted respiratory demand at time t, that is, the expected neonatal ventilator parameter requirements (such as airway pressure, tidal volume, etc.). i W (t) represents the value of the i-th physiological parameter at time t. Common physiological parameters include airway pressure, tidal volume, blood oxygen saturation, heart rate, etc. i is the weight coefficient obtained by the convolutional neural network (CNN) model during the training process, which indicates the influence of each physiological parameter on the prediction result. b is the bias term, which is used to adjust the baseline value of the prediction result. f is the activation function (such as ReLU or Sigmoid), which determines the nonlinear characteristics of the model output. N is the dimension of the input physiological data, which indicates the number of physiological parameters used by the model.

[0109] Assume that there is a premature infant with a birth weight of 1500 grams and a gestational age of 32 weeks, who is currently receiving assisted ventilation from a neonatal ventilator. The infant's airway pressure, tidal volume, blood oxygen saturation, and heart rate data are collected in real time as input physiological parameters. Based on the physiological data of the newborn, the model is trained through a convolutional neural network (CNN) to obtain the corresponding weight coefficient W i , to predict their future respiratory needs.

[0110] Example: Real-time prediction of newborns

[0111] The following physiological parameters are set at the current moment (set to t = 0 seconds):

[0112] Airway pressure: X1(0) = 12.5 cmH2O

[0113] Tidal volume: X2(0) = 6 ml

[0114] Blood oxygen saturation: X3(0) = 92%

[0115] Heart rate: X4(0)=150bpm

[0116] The training process of the CNN model is set to obtain the weight of each parameter in the following process:

[0117] W1 = 0.3 (the influence of airway pressure on respiratory demand)

[0118] W2 = 0.25 (the influence of tidal volume on respiratory demand)

[0119] W3 = -0.4 (negative correlation between blood oxygen saturation and respiratory demand)

[0120] W4 = 0.15 (the influence of heart rate on respiratory demand)

[0121] In addition, there is a bias term b = 1.5 and a ReLU activation function is selected (f(x) = max(0, x)). This means that the future respiratory needs of the newborn will be predicted in this model based on the current physiological data.

[0122] According to the formula, the weighted sum of the input parameters and the corresponding weights is first performed:

[0123]

[0124] Substituting the values:

[0125] 0.3·12.5+0.25·6+(-0.4)·92+0.15·150

[0126] =3.75+1.536.8+22.5=-9.05

[0127] Then add the bias term b = 1.5, and we get:

[0128] -9.05+1.5=-7.55

[0129] Then apply the ReLU activation function (f(x)=max(0,x)) to process this value:

[0130] f(-7.55)=0

[0131] Through this calculation, the predicted value is 0. This indicates that at the current moment, based on the input physiological data (airway pressure, tidal volume, blood oxygen saturation, heart rate), the newborn's respiratory demand is low. In fact, this may mean that the newborn's ventilator settings should be at a lower ventilation level because its physiological state does not show signs of needing more support for the time being. Over time, the system will continue to monitor the newborn's physiological data and calculate the predicted value in real time. If the input physiological data changes at this time (for example, the airway pressure increases and the blood oxygen saturation decreases), the CNN model will recalculate and give an updated prediction of respiratory demand. This will help the ventilator dynamically adjust parameters such as airway pressure and tidal volume according to the actual physiological state of the newborn to ensure that respiratory support always matches its needs.

[0132] Through this real-time prediction and feedback mechanism, the ventilator can not only respond to current physiological needs, but also predict and adapt to upcoming physiological fluctuations in advance. For example, if the blood oxygen saturation of a newborn drops to 88%, the system will predict the increase in respiratory demand through CNN and adjust the airway pressure or tidal volume in advance to avoid severe hypoxia in the newborn.

[0133] This embodiment constructs a comprehensive physiological state map by jointly analyzing multiple physiological parameters to accurately reflect the actual respiratory needs of newborns. The implementation process and feasibility of this technology will be deeply understood through specific examples and calculation steps.

[0134] The formula mentioned is:

[0135]

[0136] in,

[0137] Ψ(t) represents the comprehensive physiological state map at time t, reflecting the overall physiological needs of the newborn. k (t) represents the value of the kth physiological parameter at time t, including airway pressure, tidal volume, respiratory rate, blood oxygen saturation, etc. k is the weighting coefficient, which indicates the contribution of each physiological parameter to the comprehensive physiological state map. β is the bias term, which is used to adjust the output value of the map. M is the total number of physiological parameters involved in the fusion.

[0138] Assume that a newborn is receiving ventilator-assisted therapy, with a birth weight of 2000 grams, a gestational age of 34 weeks, and is experiencing mild respiratory distress syndrome. Multiple physiological data are collected in real time, including airway pressure, tidal volume, respiratory rate, and blood oxygen saturation. Through multi-parameter data fusion technology, these physiological data are comprehensively processed to construct an accurate comprehensive physiological state map to determine the real respiratory needs of the newborn and adjust the ventilator parameters.

[0139] Assume that the following data are collected at a certain time (t = 0 seconds):

[0140] Airway pressure X1(0) = 15 cmH2O

[0141] Tidal volume X2(0) = 8 ml

[0142] Respiratory rate X3(0) = 50 breaths / min

[0143] Blood oxygen saturation X4(0) = 88%

[0144] According to the previous data analysis and physiological research, the model determines the weighting coefficients of each physiological parameter (α k) and the bias term (β):

[0145] α1=0.4 (airway pressure contributes more to the overall physiological state)

[0146] α2 = 0.3 (tidal volume has a moderate effect on the overall physiological state)

[0147] α3=0.2 (The respiratory rate has little influence on the comprehensive physiological state)

[0148] α4=0.1(Blood oxygen saturation has little influence on the overall physiological state)

[0149] Bias term β = 0.5

[0150] These weighting coefficients and bias terms reflect the contribution of each physiological parameter to the comprehensive physiological state map. The value range of the weighting coefficient is generally adjusted according to clinical experience, experimental data and the complexity of the physiological state of the newborn.

[0151] According to the formula, weighted summation is performed first:

[0152]

[0153] Substituting the values:

[0154] =0.4·15+0.3·8+0.2·50+0.1·88

[0155] =6+2.4+10+8.8=27.2

[0156] Next, add a bias term of β = 0.5:

[0157] 27.2+0.5=27.7

[0158] Finally, the value of the comprehensive physiological state map at time t=0 is:

[0159] Ψ(0)=27.7

[0160] Through this calculation, the value of the comprehensive physiological state map is 27.7. This means that at the current moment, the overall physiological needs of the newborn are high and the respiratory support may be in a state of urgency. Based on this comprehensive value, the ventilator can automatically adjust relevant parameters, such as increasing airway pressure or tidal volume, to ensure that the support provided by the ventilator can effectively meet the needs of the newborn.

[0161] Once the system calculates the comprehensive physiological state map, it can immediately feed back to the ventilator to adjust the output of the device. The preset airway pressure range for the baby is set to 10 to 15 cmH2O, the tidal volume range is 6 to 8 ml, and the blood oxygen saturation should be maintained above 90%. If the system's comprehensive physiological state map indicates a higher demand, the ventilator will automatically adjust the airway pressure to 15 cmH2O and increase the tidal volume to 8 ml to improve the oxygen supply to the newborn and ensure airway patency.

[0162] Through this multi-parameter data fusion technology, the system can process data from multiple sensors in real time, and make weighted adjustments based on the relative importance of each physiological parameter, ultimately deriving an accurate comprehensive physiological state map. This allows the ventilator to not only respond to current physiological needs, but also predict future respiratory needs in advance based on comprehensive assessments, thereby providing personalized and precise respiratory support for newborns.

[0163] Embodiment 2:

[0164] This embodiment constructs a nonlinear dynamic pressure-volume relationship model to achieve dynamic optimization and personalized adjustment of the ventilator. By real-time monitoring of the respiratory data and physiological state of the newborn, the ventilator output can be optimized to ensure that the respiratory support of each newborn achieves the best effect. Specific data will be used to illustrate the feasibility of the steps and the practical application of the technology.

[0165] The dynamic pressure and volume relationship model is as follows:

[0166]

[0167] in:

[0168] P aw (t) is the airway pressure of the newborn at the current moment (unit: cmH2O), indicating the pressure of the ventilator output gas.

[0169] V t (t) is the tidal volume at the current moment (unit: ml), which indicates the amount of gas entering and leaving during each breath.

[0170] C l (t) is the lung elasticity of the newborn (unit: ml / cmH2O), which indicates the lung's ability to expand the volume of gas.

[0171] R aw (t) is the airway resistance of the neonate (unit: cmH2O / L / s), which describes the airway resistance to airflow.

[0172] γ(t) is an adjustment coefficient that reflects the neonatal lung development and physiological status, indicating the state of dynamic adjustment of lung resilience and airway ventilation. It changes with the neonatal lung development stage and its physiological needs.

[0173] Assume that there is a newborn baby with a current birth weight of 1800 grams and a gestational age of 32 weeks, who is receiving ventilator-assisted ventilation. The following physiological data are collected in real time:

[0174] Tidal volume V t (t) = 6 ml (the amount of gas in and out each time you breathe).

[0175] Lung elasticity (Indicates the lungs' ability to expand with gas).

[0176] Airway resistance (Reflects the airway resistance to airflow).

[0177] The adjustment coefficient γ(t) = 0.1, which indicates that the lungs of newborns are relatively poorly developed and have low lung resilience.

[0178] First, the ratio between tidal volume and lung elasticity needs to be calculated:

[0179]

[0180] This ratio represents the pressure effect of tidal volume on lung elasticity under the current physiological state.

[0181] Next, calculate the product between airway resistance and tidal volume:

[0182]

[0183] This value reflects the effect of airway resistance on airflow.

[0184] Adding the above two parts together, we get a preliminary calculation of the airway pressure:

[0185]

[0186] This value represents the airway pressure resulting from the combined effects of tidal volume, lung elastance, and airway resistance.

[0187] In order to further adjust the airway pressure to adapt to the lung development and physiological state of the newborn, the adjustment coefficient γ(t) is introduced to obtain the final airway pressure:

[0188]

[0189] At this moment, the model calculates that the airway pressure is 103.4 cmH2O. This means that in order to meet the physiological needs of the newborn, the ventilator needs to output an airway pressure of 103.4 cmH2O to ensure that the gas effectively enters the lungs and maintains respiratory support. As time goes by, the lung elasticity of the newborn is improved and the lung development gradually improves, and the physiological data monitored in real time will change. For example, after a few hours, the lung elasticity of the newborn C l (t) Increase to 2ml / cmH2O, airway resistance R aw (t) decreased to 10 cmH2O / L / s, while the tidal volume V t (t) remained unchanged at 6 ml, and the adjustment factor γ(t) changed to 0.05 (reflecting improved lung development).

[0190] Calculate the airway pressure again:

[0191] 1. Tidal volume to lung elasticity ratio:

[0192]

[0193] 2. The influence of airway resistance on tidal volume:

[0194]

[0195] 3. Comprehensive pressure calculation:

[0196]

[0197] 4. Introduction of adjustment coefficient:

[0198]

[0199] After several hours of treatment and improvement in physiological status, the calculated airway pressure of the newborn dropped to 66.15 cmH2O. Through this dynamic optimization method, the ventilator automatically adjusts the airway pressure based on real-time data to ensure that respiratory support suitable for the current lung status of the newborn is always provided during treatment.

[0200] This embodiment uses the neonatal lung elasticity (C l (t)) and airway resistance (R aw (t)) to achieve precise treatment strategies for ventilators. Through nonlinear modeling of these physiological characteristics and dynamic adjustment over time, the ventilator can intelligently adjust treatment parameters according to the physiological state of the newborn and the progress of lung development. This solution aims to achieve personalized respiratory support by monitoring the lung state of the newborn in real time and optimizing the balance between airway pressure and tidal volume.

[0201] The dynamic modeling method includes two functions:

[0202] 1. Model of lung elasticity changing over time:

[0203] C l (t) = C l0 ·exp(-αt)

[0204] in:

[0205] C l (t) represents the lung elasticity of the newborn at time t, in ml / cmH2O, reflecting the ability of the lungs to expand with gas.

[0206] C l0 It is the lung elastic constant of a newborn at birth, indicating the initial lung volume expansion capacity.

[0207] α is a coefficient that describes the decay of lung elasticity over time, reflecting the speed at which the lungs gradually develop and improve. Usually, the value of α ranges from 0.02 to 0.05, indicating the rate at which lung elasticity gradually improves.

[0208] t is a time variable, which can be expressed in hours or days, indicating the developmental stage of the newborn.

[0209] 2. Model of airway resistance changing over time:

[0210] R aw (t) = R aw0 (1+βt)

[0211] in:

[0212] R aw (t) represents the airway resistance of the newborn at time t, in cmH2O / L / s, describing the airway resistance to airflow. aw0 is the airway resistance constant of the newborn at birth, indicating the initial airway resistance. β is a coefficient that describes the gradual increase in airway resistance as the newborn develops. Usually, the value of β ranges from 0.001 to 0.005, reflecting the speed at which the airway gradually becomes more resistant during development. t is a time variable, indicating the developmental stage of the newborn.

[0213] Assume that there is a newborn with a birth weight of 1500 g, a gestational age of 30 weeks, and a current age of 48 hours (2 days). The following initial data are obtained using real-time monitoring:

[0214] Initial lung elasticity (Indicates the volume expansion capacity of the lungs at birth).

[0215] Initial airway resistance (Indicates the amount of airway resistance at birth).

[0216] The attenuation coefficient α = 0.03 represents the rate at which lung elasticity gradually improves over time.

[0217] The growth coefficient β = 0.002, indicating that airway resistance gradually increases with development.

[0218] 1. Changes in lung elasticity over time:

[0219] Use the formula to calculate the lung elasticity of the newborn after 48 hours (2 days):

[0220] C l (48) = 1.2 exp(-0.03 48)

[0221]

[0222] According to the calculation results, the value of the newborn's lung elasticity after 48 hours is 0.284ml / cmH2O, which means that the lungs have begun to develop significantly during this period and the lung's expansion capacity has improved.

[0223] 2. Changes in airway resistance over time:

[0224] Calculate the airway resistance of the newborn after 48 hours using the formula:

[0225] R aw (48) = 12·(1+0.002·48)

[0226]

[0227] According to the calculation results, the value of the newborn's airway resistance after 48 hours is 13.152 cmH2O / L / s, indicating that the airway resistance increases slightly as the newborn grows.

[0228] Real-time monitoring of neonatal lung elasticity and airway resistance The ventilator can adjust the therapy by optimizing the balance between pressure and volume. For example, setting the tidal volume V t (48) = 8 ml (the amount of gas in and out each time you breathe in). The ventilator will dynamically calculate the relationship between airway pressure and tidal volume to ensure that the ventilator provides appropriate support for the newborn.

[0229] Using the dynamic pressure and volume relationship model described above:

[0230]

[0231] Setting γ(48) = 0.05 indicates that lung resilience and airway ventilation are still in a certain degree of development.

[0232] It can calculate:

[0233] 1. Tidal volume to lung elasticity ratio:

[0234]

[0235] 2. The influence of airway resistance on tidal volume:

[0236]

[0237] 3. Comprehensive pressure calculation:

[0238]

[0239] 4. Introduction of adjustment coefficient:

[0240]

[0241] According to the calculation results, the ventilator should provide an airway pressure of 139.055 cmH2O to support the breathing of the newborn. In this way, the ventilator can accurately adjust the airway pressure to ensure that the respiratory support matches the lung development status and airway ventilation of the newborn.

[0242] This embodiment optimizes the relationship between pressure and volume by processing the real-time collected physiological data of the newborn (including airway pressure and tidal volume). This method uses a nonlinear regression model to describe the relationship between pressure and tidal volume, aiming to adjust the output of the ventilator in real time to make it more accurately adapt to the physiological needs of the newborn. The implementation steps and feasibility of the scheme will be explained through specific examples, and actual data will be substituted to show how to optimize the treatment strategy of the ventilator through the model.

[0243] The nonlinear regression model described uses the following formula to describe the relationship between pressure and tidal volume:

[0244] P aw (t) = A·(V t (t)) α +B·exp(-γV t (t))

[0245] in:

[0246] P aw (t) represents the airway pressure at time t (unit: cmH2O), reflecting the pressure of the ventilator output gas;

[0247] V t (t) represents the tidal volume at time t (unit: ml), the amount of gas in and out of each breath;

[0248] A is the amplification coefficient in the regression model, reflecting the strength of the relationship between pressure and tidal volume;

[0249] B is the constant term in the regression model, representing the basic relationship between pressure and tidal volume;

[0250] α is the exponential coefficient, describing the nonlinear relationship between pressure and tidal volume;

[0251] γ is the decay constant, describing the dynamic effect of lung elasticity and airway resistance.

[0252] In order to adapt the model to the physiological characteristics of newborns, the values ​​of the coefficients need to be adjusted according to the individual differences of newborns. Under different treatment stages and physiological conditions of newborns, the following value ranges can be set: A is the amplification coefficient, with a value range of 0.8 to 1.5, reflecting the strength of the relationship between pressure and tidal volume; B is a constant term, with a value range of 0.1 to 0.5, indicating the basic pressure; α is an exponential coefficient, with a value range of 0.5 to 1.5, indicating the strength of the nonlinear relationship between pressure and tidal volume; γ is an attenuation constant, with a value range of 0.02 to 0.1, describing the influence of lung elasticity and airway resistance on pressure changes.

[0253] Suppose there is a newborn baby with a birth weight of 1600 grams and a gestational age of 32 weeks, and is now in the 72nd hour of treatment. Through real-time monitoring, the following data are obtained:

[0254] Tidal volume V t (72) = 6.5 ml (volume of gas in and out per breath);

[0255] The coefficient A = 1.2, indicating the strength of the relationship between pressure and tidal volume;

[0256] The coefficient B = 0.3, which represents the basic relationship between pressure and tidal volume;

[0257] The exponential coefficient α = 1.0 indicates a nonlinear relationship between pressure and tidal volume;

[0258] The decay constant γ = 0.05 describes the dynamic effects of lung elasticity and airway resistance.

[0259] Calculation steps:

[0260] 1. Calculate the relationship between pressure and tidal volume at 72 hours:

[0261] A nonlinear regression model was used to calculate neonatal airway pressure at 72 hours:

[0262] P aw (72) = 1.2 (6.5) 1.0 +0.3 exp(-0.05 6.5)

[0263] First calculate the exponential part of the tidal volume:

[0264] (6.5) 1.0 =6.5

[0265] Then calculate the attenuation term:

[0266] exp(-0.05·6.5)=exp(-0.325)≈0.722

[0267] Next, substitute and calculate the airway pressure:

[0268]

[0269] According to calculations, the current airway pressure of the newborn is 8.017 cmH2O. Through this nonlinear regression model, the ventilator can dynamically adjust the airway pressure at the 72nd hour to adapt it to the actual needs of the newborn. This result shows that by combining the physiological data of the newborn (such as tidal volume) and the nonlinear characteristics of the regression model, the ventilator can optimize treatment and ensure that the airway pressure is always maintained within the appropriate range of the newborn's physiological needs.

[0270] Based on real-time data and this nonlinear regression model, the ventilator is able to dynamically adjust airway pressure to optimize it as tidal volume and the physiological state of the newborn change. For example, when tidal volume changes, the ventilator automatically adjusts airway pressure to avoid excessive or insufficient respiratory support. This approach ensures that each newborn can get the most appropriate ventilator settings based on their specific physiological needs.

[0271] By optimizing the relationship between pressure and tidal volume through a nonlinear regression model and combining it with the newborn's real-time physiological data, the ventilator can accurately adjust the airway pressure to provide personalized respiratory support for the newborn. This method can effectively avoid the shortcomings of traditional single parameter settings, making the treatment more precise and dynamic, and meeting the physiological needs of newborns at different treatment stages and developmental states. This dynamic optimization method based on real-time data feedback not only improves the accuracy of treatment, but also reduces the risk of complications, and improves the treatment effect and survival rate of newborns.

Claims

1. The monitoring system of the neonatal ventilator is characterized by The following steps are involved: S1. Real-time physiological dynamic adaptation mechanism of neonatal ventilator: S1.

1. By real-time monitoring of the airway pressure, tidal volume, and respiratory rate parameters of the newborn, combined with the individual physiological characteristics of the newborn, a breathing pattern model based on physiological characteristics is constructed; S1.

2. The ventilator is regulated by an adaptive control algorithm based on real-time data feedback to dynamically adjust the ventilator parameters; S2. Nonlinear modeling and dynamic optimization of the relationship between ventilator pressure and volume: S2.

1. Establish a pressure-volume relationship model by combining the airway resistance and lung elasticity of the newborn; calculate the data of different development stages of the newborn's lungs, and use nonlinear regression method to obtain the relationship data value between pressure and volume; S2.2, the ventilator shall calculate the current airway pressure and tidal volume, and shall also calculate the balance between pressure and volume based on the dynamic characteristics of lung resilience and airway ventilation; S3, Intelligent switching mechanism between ventilator-assisted and spontaneous breathing: S3.

1. Based on the changes in respiratory rate, airway pressure fluctuations and tidal volume, monitor the respiratory drive and rhythm of the newborn in real time to determine whether the newborn has entered the state of spontaneous breathing; by analyzing the airway pressure waveform and respiratory cycle characteristics, identify the timing of the newborn's spontaneous breathing; S4. Adaptive optimization of ventilator parameters and individualized treatment pathways: S4.

1. Establish a treatment model based on individual differences by continuously collecting the heart rate, blood oxygen saturation, and airway pressure data of the newborn, combined with birth weight, gestational age, and disease background factors; the treatment model guides the individualized setting of the ventilator; S4.

2. Use machine learning algorithms to dynamically optimize individual ventilator parameters based on historical data and real-time monitoring results; Each output value of the ventilator is adjusted by an algorithm.

2. The monitoring system of the neonatal ventilator according to claim 1, characterized in that The method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator comprises: A dynamic equation is used to describe the relationship between airway pressure and time; the dynamic equation calculates the current physiological state of the newborn and also makes adaptive adjustments based on birth weight and gestational age factors; the equation is as follows: in: P(t) represents the airway pressure at time t; A, B, C, D, and E are constant coefficients, representing parameters calculated based on individual physiological characteristics of newborns including weight, gestational age, and lung function; exp(D·t) is an exponential function, representing the nonlinear change of airway pressure over time; the exponential function is used to simulate the rapid change of airway pressure during breathing; t is a time variable, representing the change of airway pressure monitored at a specific time point; and E is a constant bias term used to adjust the final airway pressure.

3. The monitoring system of the neonatal ventilator according to claim 2 is characterized in that The method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator comprises: Use convolutional neural network CNN to predict future respiratory needs based on the real-time physiological data of the newborn and adjust the ventilator parameters in advance; introduce the prediction model, and the output is the predicted value of future respiratory needs. The formula is as follows: in: represents the predicted respiratory demand at time t; X i (t) represents the value of the ith physiological parameter at time t; physiological parameters include airway pressure, tidal volume, blood oxygen saturation, heart rate, etc.; each parameter X i (t) reflects the immediate physiological condition of the newborn; W i is the weight coefficient obtained during the deep learning model training process; each weight W i It indicates the influence of the physiological parameter on the final prediction result; b is the bias term; f is the activation function that determines the nonlinear characteristics of the model output; N is the dimension of the input physiological data, indicating how many orders of magnitude of physiological parameters the model uses for prediction.

4. The monitoring system of the neonatal ventilator according to claim 3 is characterized in that The method for constructing the real-time physiological dynamic adaptation mechanism of the neonatal ventilator includes: introducing multi-parameter data fusion technology, by jointly analyzing multiple physiological data, reflecting the real needs of the newborn; fusing multi-dimensional physiological data, using weighted average to integrate the data, and obtaining a comprehensive physiological state map.

5. The monitoring system of the neonatal ventilator according to claim 1 is characterized in that The nonlinear modeling and dynamic optimization method of the relationship between ventilator pressure and volume: through individualized physiological data, combined with the airway resistance and lung elasticity of the newborn, a nonlinear dynamic pressure and volume relationship model is constructed, and the balance of pressure and volume is adjusted in real time; according to the different stages of lung development and physiological status of the newborn, the output of the ventilator is optimized.

6. The monitoring system of the neonatal ventilator according to claim 5, characterized in that Nonlinear modeling and dynamic optimization method of the relationship between pressure and volume of the ventilator: By real-time monitoring of the lung elasticity and airway resistance of newborns, combined with physiological characteristics, dynamic calculation of lung resilience and airway ventilation can be performed to optimize the ventilator treatment strategy; The following function model is used to describe the characteristics of lung elasticity and airway resistance changing with time: C l (t)=C l0 exp(-αt) R aw (t)=R aw0 ·(1+βt) in: C l0 is the lung elastic constant of the newborn at birth, indicating the initial lung volume expansion capacity; α is the coefficient describing the attenuation of lung elasticity over time, reflecting the speed of gradual lung development and perfection; R aw0 is the airway resistance constant of the newborn at birth, indicating the initial airway resistance; β is the coefficient describing the gradual increase of airway resistance as the newborn develops; t is a time variable, indicating the developmental stage of the newborn, in hours or days.

7. The monitoring system of the neonatal ventilator according to claim 6, characterized in that Nonlinear modeling and dynamic optimization method of the relationship between pressure and volume of the ventilator: Through the nonlinear regression algorithm, based on the real-time collection of neonatal physiological data including airway pressure and tidal volume, the relationship between pressure and volume is optimized, and the following nonlinear regression model is used to describe the relationship between pressure and volume: P aw (t)=A·(V t (t)) α +B·exp(-γV t (t)) in: A is the amplification coefficient in the regression model, which reflects the strength of the relationship between pressure and tidal volume; B is the constant term in the regression model, which represents the basic relationship between pressure and tidal volume; α is the exponential coefficient, which describes the nonlinear relationship between pressure and tidal volume; γ is the attenuation constant, which describes the dynamic effect of lung elasticity and airway resistance; V t (t) is the tidal volume, in ml, which indicates the amount of gas in and out during each breath.

Images