A method and system for continuous estimation of dynamic respiratory mechanics parameters in a ventilator

Abstract

The present application belongs to the technical field of breathing machine, and particularly relates to a method and system for continuously estimating dynamic respiratory mechanics parameters of a breathing machine, which comprises the following steps: defining an airway system based on lungs, airways and breathing machine ventilation pipeline in a tester's body, and establishing a basic respiratory mechanics formula; obtaining sampling values including pressure, flow rate and tidal volume within a set inspiration cycle, and establishing a matrix equation based on the basic respiratory mechanics formula; and using a recursive least square method to estimate compliance and airway resistance in real time according to the sampling values, so as to obtain the variation trend of the two parameters over time within the entire inspiration stage, and realize continuous estimation. The method of the present application is not limited by ventilation modes, and can accurately calculate airway resistance and compliance under any ventilation mode; the parameters can be calculated in real time and continuously, the real dynamic characteristics of respiratory mechanics can be reflected, and the control strategy of some ventilation modes of the breathing machine can be optimized.

Description

Technical Field

[0001] This invention belongs to the field of ventilator technology, and particularly relates to a method and system for continuous estimation of dynamic respiratory mechanics parameters of a ventilator. Background Technology

[0002] The air resistance and compliance of the test subject are important monitoring parameters during surgery. These parameters not only reflect the real-time physical condition of the test subject, allowing doctors to implement appropriate ventilation methods and parameters based on their size, but also significantly affect the ventilation performance of the anesthesia machine because the ventilation process of the anesthesia machine is designed with different airflow rates and ventilation times based on these two parameters.

[0003] Currently, a commonly used method is to calculate patient compliance and air resistance under VCV (volume-controlled ventilation) mode, and monitor the peak inspiratory pressure P during inspiration. peak Ventilation platform pressure P plat Peak airway flow rate f peak Positive end-expiratory pressure (P) peep Inspiratory tidal volume V insp Then, the subject's compliance and airway resistance are calculated using the following formula:

[0004]

[0005]

[0006] This calculation method can only be performed in VCV mode and cannot be used in pressure-based control modes. Alternatively, by acquiring and monitoring pressure, flow rate, and tidal volume signals within a specific respiratory cycle, the parameters R and C can be estimated using the least squares method according to the basic respiratory mechanics formulas.

[0007] Breathing mechanics parameters, especially compliance (C), are actually parameters that change dynamically during the respiratory cycle, reflected as the slope of the PV curve. The air resistance and compliance calculated by the two methods mentioned above are values ​​at a single instant, i.e., they assume that compliance (C) remains constant during breathing, and therefore cannot accurately reflect the user's true breathing status. Therefore, a method is needed that can monitor changes in mechanics parameters in real time and continuously during breathing. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and to propose a method and system for continuous estimation of dynamic respiratory mechanical parameters of a ventilator.

[0009] To achieve the above objectives, this invention proposes a continuous estimation method for dynamic respiratory mechanics parameters of a ventilator, the method comprising:

[0010] Step 1) Define the airway system based on the lungs, airway and ventilator tubing in the test subject's body, and establish basic respiratory mechanics formulas;

[0011] Step 2) Acquire sampled values ​​including pressure, flow rate and tidal volume within the set inspiratory cycle, and establish a matrix equation based on the basic respiratory mechanics formula;

[0012] Step 3) Using the recursive least squares method, compliance and air resistance are estimated in real time based on the sampled values, thereby obtaining the changing trends of these two parameters over time during the entire inhalation phase, and achieving continuous estimation.

[0013] As an improvement to the above method, step 1) specifically includes:

[0014] Define the lungs, airway, and ventilator tubing of the test subject as a single airway system, and establish a basic respiratory mechanics formula where pressure P, tidal volume V, and flow rate F satisfy the following equation:

[0015] P = EV + RF + P0

[0016] Where P0 is positive end-expiratory pressure; E is the system's elasticity, which is the reciprocal of compliance; and R represents air resistance.

[0017] As an improvement to the above method, step 2) includes:

[0018] Within a set inspiratory cycle, n sets of sampled values ​​are obtained, including pressure values ​​[P1, P2, ..., P]. n ,]、Tidal volume [V 1, V2,…,V n ,] and flow velocities [F1,F2,…,F n Based on the basic respiratory mechanics formula, the following equation is established:

[0019]

[0020] And abbreviated as:

[0021] P = Wθ

[0022] Where P is a matrix composed of pressure values, W is a matrix composed of tidal volume and flow velocity, and θ is a matrix of parameters to be estimated.

[0023] As an improvement to the above method, step 3) includes:

[0024] Step 3-1) Iterate through each group of sampled values ​​and perform the processing in step 3-2) to obtain the optimal estimated parameters for each group of sampled values. Continue until the traversal is complete, then proceed to step 3-3);

[0025] Step 3-2) Use the recursive least squares method to solve for the optimal estimated parameters of the k-th sample value.

[0026]

[0027]

[0028] s k =[V k ,F k ,1]

[0029] in, These are the optimal estimation parameters for the k-th and (k-1)-th sampled values, respectively, where ρ is the forgetting factor, ranging from 0 to 1, T represents the matrix transpose, and C... k C k-1 Let s represent the k-th and (k-1)-th rows of the covariance matrix of W, respectively. k V represents the tidal volume of the kth group. k and flow velocity F k The combination of P k Indicates the pressure of the k-th group;

[0030] Step 3-3) According to each The elasticity and air resistance are estimated in real time. Since elasticity and compliance are reciprocals of each other, the changing trends of compliance and air resistance over time during the entire inhalation phase are obtained.

[0031] On the other hand, the present invention proposes a continuous estimation system for dynamic respiratory mechanics parameters of a ventilator, the system comprising:

[0032] The Basic Respiratory Mechanics Formula Construction Module is used to define the airway system based on the lungs, airway and ventilator ventilation circuit in the test subject's body and establish basic respiratory mechanics formulas.

[0033] The matrix equation construction module is used to acquire sampled values ​​including pressure, flow rate, and tidal volume within a set inspiratory cycle, and to establish a matrix equation based on basic respiratory mechanics formulas; and

[0034] The respiratory mechanics parameter estimation module is used to estimate compliance and air resistance in real time based on the sampled values ​​using the recursive least squares method, thereby obtaining the changing trends of these two parameters over time during the entire inhalation phase and achieving continuous estimation.

[0035] As an improvement to the above system, the processing procedure of the basic respiratory mechanics formula construction module specifically includes:

[0036] Define the lungs, airway, and ventilator tubing of the test subject as a single airway system, and establish a basic respiratory mechanics formula where pressure P, tidal volume V, and flow rate F satisfy the following equation:

[0037] P = EV + RF + P0

[0038] Where P0 is positive end-expiratory pressure, E is the system's elasticity, which is the reciprocal of compliance, and R represents air resistance.

[0039] As an improvement to the above system, the processing procedure of the matrix equation construction module includes:

[0040] Within a set inspiratory cycle, n sets of sampled values ​​are obtained, including pressure values ​​[P1, P2, ..., P]. n ,]、Tidal volume [V1,V2,…,V n ,] and flow velocities [F1,F2,…,F n Based on the basic respiratory mechanics formula, the following equation is established:

[0041]

[0042] And simplified to:

[0043] P = Wθ

[0044] Where P is a matrix composed of pressure values, W is a matrix composed of tidal volume and flow velocity, and θ is a matrix of parameters to be estimated.

[0045] As an improvement to the above system, the processing procedure of the respiratory mechanics parameter estimation module includes:

[0046] Step 3-1) Iterate through each group of sampled values ​​and perform the processing in step 3-2) to obtain the optimal estimated parameters for each group of sampled values. Continue until the traversal is complete, then proceed to step 3-3);

[0047] Step 3-2) Use the recursive least squares method to solve for the optimal estimated parameters of the k-th sample value.

[0048]

[0049]

[0050] s k =[V k ,F k ,1]

[0051] in, These are the optimal estimation parameters for the k-th and (k-1)-th sampled values, respectively, where ρ is the forgetting factor, ranging from 0 to 1, T represents the matrix transpose, and C... k C k-1 Let s represent the k-th and (k-1)-th rows of the covariance matrix of W, respectively. k V represents the tidal volume of the kth group.k and flow velocity F k The combination of P k Indicates the pressure of the k-th group;

[0052] Step 3-3) According to each The elasticity and air resistance are estimated in real time. Since elasticity and compliance are reciprocals of each other, the changing trends of compliance and air resistance over time during the entire inhalation phase are obtained.

[0053] Compared with the prior art, the advantages of the present invention are:

[0054] 1. The method of the present invention is not limited by the ventilation mode and can accurately calculate air resistance and compliance in any ventilation mode;

[0055] 2. The data used in the method of the present invention is not ventilation data at a single moment, but data of the entire process from the beginning to the end of inhalation, and the calculated results are closer to the actual situation of the test subject.

[0056] 3. The method of the present invention is based on the least squares error criterion, which minimizes the error of the calculation result and makes it closest to the true value;

[0057] 4. The method of this invention uses recursive calculation, which provides accurate results while saving system memory space;

[0058] 5. The method of the present invention can calculate parameters in real time and continuously, and can reflect the true dynamic characteristics of respiratory mechanics. Attached Figure Description

[0059] Figure 1 This is a flowchart of the continuous estimation method for dynamic respiratory mechanics parameters of a ventilator according to the present invention;

[0060] Figure 2 is a schematic diagram of the input signals of this method, where Figure 2(a) is pressure P, Figure 2(b) is flow velocity F, and Figure 2(c) is capacity V;

[0061] Figure 3 is a schematic diagram of the changes in air resistance and compliance obtained by the method of the present invention, wherein Figure 3(a) is air resistance R and Figure 3(b) is compliance C. Detailed Implementation

[0062] To overcome the shortcomings of the two calculation methods mentioned above, a more accurate calculation method that is applicable to more modes and can continuously monitor parameter changes was designed. This method is based on the basic respiratory mechanics formula and uses the recursive least squares method to estimate compliance C and air resistance R in real time from the pressure, flow rate and tidal volume signals during the inspiratory phase. Thus, the changing trend of these two parameters over time during the entire inspiratory phase can be observed.

[0063] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0064] Example 1

[0065] like Figure 1 As shown, Embodiment 1 of the present invention proposes a continuous estimation method for dynamic respiratory mechanics parameters of a ventilator, the method comprising:

[0066] Step 1) Define the airway system based on the lungs, airway and ventilator tubing in the test subject's body, and establish basic respiratory mechanics formulas;

[0067] Step 2) Acquire sampled values ​​including pressure, flow rate and tidal volume within the set inspiratory cycle, and establish a matrix equation based on the basic respiratory mechanics formula;

[0068] Step 3) Using the recursive least squares method, compliance and air resistance are estimated in real time based on the sampled values, thereby obtaining the changing trends of these two parameters over time during the entire inhalation phase, and achieving continuous estimation.

[0069] The specific analysis is as follows:

[0070] If we define the lungs, airway, and ventilator tubing in the test subject's body as a single airway system, then the pressure P, volume V, and flow rate F in this system have the following relationships:

[0071] P = EV + RF + P0

[0072] Where P0 represents positive end-expiratory pressure, E represents the system's elasticity (the reciprocal of compliance), and R represents air resistance. Assuming that within a given inspiratory cycle, each signal contains n sampling points, and each data set is represented by a column vector, with elements representing the first to nth observations of pressure, flow rate, or tidal volume, then the above formula also has a matrix multiplication form:

[0073]

[0074] Abbreviated as:

[0075] P = Wθ

[0076] Where P and W represent matrices composed of sampling points of the observed signal, and θ represents the parameter matrix to be estimated. The ultimate goal is to find the optimal estimate of θ. Make P and The sum of squared errors is minimized, that is:

[0077]

[0078] The loss function can be set as:

[0079] L(θ) = ||Wθ - P|| 2

[0080] =(Wθ-P) T (Wθ-P)

[0081] =θ T W T Wθ-2θ T W T P+PP T

[0082] Set the partial derivative of the loss function with respect to the parameters to zero:

[0083]

[0084] And when When the loss function is minimized, that is, when the partial derivative of the loss function is zero:

[0085]

[0086] The expression for the optimal parameters can be obtained by solving:

[0087]

[0088] The ultimate goal of this method is to transform the process of finding the optimal parameters into a recursive form, treating the parameter matrix as time-varying. Each observation at any given time corresponds to a unique parameter matrix, and the optimal solution for the k-th parameter set is estimated using the first to (k-1) sets of observation data. The recursive form is as follows:

[0089]

[0090] Where the subscript k represents the sampling time, and is the row index for matrices W and P, and η represents a bias, this formula shows that for time-varying systems like the respiratory mechanics model, the parameters at the current time can be obtained by adding a bias to the estimated value at the previous time. The detailed derivation of the optimal parameter estimate at each time is as follows (note the distinction between bold and non-bold, which represent matrices or vectors). First, let:

[0091]

[0092] C can be viewed as the covariance matrix of W, where:

[0093]

[0094] then:

[0095]

[0096]

[0097] The optimal parameter estimation expression under the least squares error criterion can be further derived as follows:

[0098]

[0099] Then perform an equivalent substitution:

[0100]

[0101] Here, ε is equivalent to the bias in predicting the pressure at time k using the (k-1)th estimated parameter matrix, while H is equivalent to a correction matrix for this bias. Finally, the estimate of the parameter matrix at time k can be obtained by adding a corrected bias to the parameter estimate at time k-1. This method can be summarized as the following four iterative steps:

[0102]

[0103]

[0104]

[0105]

[0106] Note that during the iteration process, to avoid a large number of inverse matrix operations, the fourth formula can be written as:

[0107]

[0108] The initial value of matrix C can be a large number multiplied by the identity matrix. Furthermore, considering that the information provided by the observation signal closer to time k is more valuable for the next estimation step during the recursive estimation process, a weight needs to be introduced to change the influence of past observation data on future parameter predictions; this weight is called the forgetting factor. After introducing the forgetting factor, the method can ultimately be summarized into the following two steps:

[0109]

[0110]

[0111] Where ρ is the forgetting factor, which takes a value between 0 and 1, usually closer to 1.

[0112] Example 2

[0113] Embodiment 2 of the present invention proposes a continuous estimation system for dynamic respiratory mechanics parameters of a ventilator, implemented based on the method of Embodiment 1. The system includes:

[0114] The Basic Respiratory Mechanics Formula Construction Module is used to define the airway system based on the lungs, airway and ventilator ventilation circuit in the test subject's body and establish basic respiratory mechanics formulas.

[0115] The matrix equation construction module is used to acquire sampled values ​​including pressure, flow rate and tidal volume within a set inspiratory cycle, and to establish matrix equations based on basic respiratory mechanics formulas.

[0116] The respiratory mechanics parameter estimation module is used to estimate compliance and air resistance in real time based on the sampled values ​​using the recursive least squares method, thereby obtaining the changing trends of these two parameters over time during the entire inhalation phase and achieving continuous estimation.

[0117] Effect verification:

[0118] When applied to real-world data, this method will produce results in the form shown in Figures 2 and 3:

[0119] Figure 2 shows the raw signals sampled during a certain period of mechanical ventilation. Figure 2(a) shows the airway pressure monitoring value P, Figure 2(b) shows the flow rate monitoring value F, and Figure 2(c) shows the tidal volume V, which are the three inputs of this method. The horizontal axis represents the sampling points. During the inspiration phase, the sampling points range from the 1st to the 60th, so our k value ranges from 1 to 60.

[0120] Figure 3 is a schematic diagram showing the changing trends of air resistance and compliance obtained using the method of the present invention, where Figure 3(a) represents air resistance R and Figure 3(b) represents compliance C. It can be seen that compliance and air resistance gradually increase at the end of inspiration, which is consistent with the mechanical properties of the lungs during inspiration.

[0121] The invention features the following advantages: it is applicable to all ventilation modes; based on the least squares error criterion, it outputs the parameters at each observation time, ultimately presenting the trend of parameter changes; and it uses recursive calculation, saving memory space.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for continuous estimation of dynamic respiratory mechanics parameters of a ventilator, the method comprising: Step 1) Define the airway system based on the lungs, airway, and ventilator tubing in the test subject's body, and establish basic respiratory mechanics formulas; Step 2) Acquire sampled values ​​including pressure, flow rate and tidal volume within the set inspiratory cycle, and establish a matrix equation based on the basic respiratory mechanics formula; Step 3) Using the recursive least squares method, compliance and air resistance are estimated in real time based on the sampled values, thereby obtaining the changing trends of these two parameters over time throughout the entire inhalation phase, achieving continuous estimation; specifically including: Step 3-1) traversing each set of sampling values, performing step 3-2) processing to obtain the optimal estimation parameter of each set of sampling values , until the traversal is completed and go to step 3-3); Step 3-2) Use the recursive least squares method to solve for the _th ... k Optimal estimation parameters of group sample values : in, ρ The forgetting factor has a value between 0 and 1. T Indicates matrix transpose. Let W represent the covariance matrix of the first and second elements respectively. k row and number k- 1 line, Indicates the first k Group tidal volume and flow rate The combination Indicates the first k Group pressure; Step 3-3) According to each The elasticity and air resistance are estimated in real time. Since elasticity and compliance are reciprocals of each other, the changing trends of compliance and air resistance over time during the entire inhalation phase are obtained.

2. The method for continuous estimation of dynamic respiratory mechanics parameters of a ventilator according to claim 1, characterized in that, Step 1) specifically includes: Define the lungs, airway, and ventilator tubing of the test subject as a single airway system, and establish a basic respiratory mechanics formula where pressure P, tidal volume V, and flow rate F satisfy the following equation: in, P 0 Positive end-expiratory pressure; E For the system's elasticity, it is the reciprocal of its compliance; R This indicates air resistance.

3. The method for continuous estimation of dynamic respiratory mechanics parameters of a ventilator according to claim 2, characterized in that, Step 2) includes: Achieve within the set inhalation cycle n Group of sampled values, the sampled values ​​including pressure values ​​[ ], tidal volume and flow rate Based on the basic respiratory mechanics formula, the following equation is established: And abbreviated as: Where P is a matrix composed of pressure values, W is a matrix composed of tidal volume and flow velocity, and θ is a matrix of parameters to be estimated.

4. A continuous estimation system for dynamic respiratory mechanics parameters of a ventilator, characterized in that, The system includes: The Basic Respiratory Mechanics Formula Construction Module is used to define the airway system based on the lungs, airway and ventilator ventilation circuit in the test subject's body and establish basic respiratory mechanics formulas. The matrix equation construction module is used to acquire sampled values ​​including pressure, flow rate, and tidal volume within a set inspiratory cycle, and to establish a matrix equation based on basic respiratory mechanics formulas; and The respiratory mechanics parameter estimation module uses a recursive least squares method to estimate compliance and air resistance in real time based on sampled values, thereby obtaining the changing trends of these two parameters over time throughout the entire inspiratory phase, achieving continuous estimation. The processing steps of the respiratory mechanics parameter estimation module include: Step 3-1) Iterate through each group of sampled values ​​and perform the processing in Step 3-2) to obtain the optimal estimated parameters for each group of sampled values. Continue until the traversal is complete, then proceed to step 3-3). Step 3-2) Use the recursive least squares method to solve for the _th ... k Optimal estimation parameters of group sample values : in, ρ The forgetting factor has a value between 0 and 1. T Indicates matrix transpose. Let W represent the covariance matrix of the first and second elements respectively. k row and number k- 1 line, Indicates the first k Group tidal volume and flow rate The combination Indicates the first k Group pressure; Step 3-3) According to each The elasticity and air resistance are estimated in real time. Since elasticity and compliance are reciprocals of each other, the changing trends of compliance and air resistance over time during the entire inhalation phase are obtained.

5. The continuous estimation system for dynamic respiratory mechanics parameters of a ventilator according to claim 4, characterized in that, The processing steps of the basic respiratory mechanics formula construction module specifically include: Define the lungs, airway, and ventilator tubing of the test subject as a single airway system, and establish a basic respiratory mechanics formula where pressure P, tidal volume V, and flow rate F satisfy the following equation: in, P 0 Positive end-expiratory pressure (PEEP) E The system's elasticity is the reciprocal of its compliance. R This indicates air resistance.

6. The continuous estimation system for dynamic respiratory mechanics parameters of a ventilator according to claim 5, characterized in that, The processing steps of the matrix equation construction module include: Within a set inspiratory cycle, n sets of sampled values ​​are obtained, including pressure values. ], tidal volume and flow rate Based on the basic respiratory mechanics formula, the following equation is established: And simplified to: Where P is a matrix composed of pressure values, W is a matrix composed of tidal volume and flow velocity, and θ is a matrix of parameters to be estimated.

Images