Method and device for predicting lung function curve, storage medium and program product

Abstract

The invention provides a lung function curve prediction method, a lung function curve prediction device, a storage medium and a program product. The prediction method comprises the following steps: acquiring flow rates of collected M sampling points; inputting the flow rates of the M sampling points into a pre-constructed and trained lung function curve prediction model, and predicting a lung function curve by using the trained lung function curve prediction model; the construction of the prediction model comprises the following steps: constructing a Taylor series expansion of an N-order K-element equation, and using the Taylor series expansion to represent the relationship between the flow velocity of the t-th sampling point of the lung function curve and the flow velocity of N sampling points before the t-th sampling point; wherein the training of the prediction model comprises the following steps: constructing a training data set; and training the lung function curve prediction model by using the training data set and a predetermined neural network algorithm until a predetermined training target is met. By utilizing the technical scheme, the real lung function curve data can be predicted by using a limited number of effective data points in the lung function examination process.

Description

Technical Field

[0001] The present invention relates to the field of pulmonary function curve prediction, and in particular to a pulmonary function curve prediction method, device, storage medium and program product. Background Art

[0002] Pulmonary function tests are crucial in the diagnosis of respiratory diseases and are crucial for treatment evaluation, serving as the primary and most critical evaluation indicator. However, these tests require high patient cooperation, otherwise valid data is difficult to obtain. This can easily lead to misdiagnosis in elderly or less educated patients due to poor quality control of test data. Currently, there is a need to predict a complete pulmonary function curve using a small number of sample points, but no existing solution exists that can achieve this predictive capability. Summary of the Invention

[0003] The embodiments of the present invention provide a method, device, storage medium and program product for predicting a pulmonary function curve, so as to predict the actual pulmonary function curve data of the subject under test with only a limited number of valid data points in the first part of the pulmonary function test process.

[0004] In order to achieve the above object, a method for predicting a pulmonary function curve is provided, comprising:

[0005] Obtaining the flow rate of M sampling points collected during the exhalation process of the subject under test during the pulmonary function test;

[0006] Inputting the flow rates of the M sampling points into a pre-built and trained pulmonary function curve prediction model, and using the trained pulmonary function curve prediction model to predict the pulmonary function curve of the subject to be tested;

[0007] The construction of the lung function curve prediction model includes the following steps:

[0008] Constructing a Taylor series expansion of an N-order K-variable equation, and using the Taylor series expansion to represent the relationship between the flow rate at the t-th sampling point on the pulmonary function curve and the flow rates at N sampling points before the t-th sampling point during exhalation of a pulmonary function test, where N, K, and t are natural numbers greater than or equal to 1, K≤N; and M≥N;

[0009] The training of the pulmonary function curve prediction model includes the following steps:

[0010] Constructing a training data set for the pulmonary function curve prediction model using data of multiple pulmonary function curves that meet predetermined quality control qualification conditions;

[0011] The pulmonary function curve prediction model is trained using a training data set and a predetermined neural network algorithm until a predetermined training goal is met.

[0012] Preferably, the prediction method uses the following Taylor series expansion of an N-order K-ary equation to express the relationship between the flow rate at the t-th sampling point on the pulmonary function curve and the flow rates at N sampling points before the t-th sampling point during the exhalation process of the pulmonary function test:

[0013] f(t)=a (t-1)K [f(t-1)] K +a (t-1)(K-1) [f(t-1)] (K-1) +...+a (t-1) f(t-1)

[0014] +b (t-1) +...+a (t-N)K [f(tN)] K

[0015] +a (t-N)(K-1) [f(tN)] (K-1) +...+a (t-N) f(tN)+b (t-N)

[0016] f(t) represents the flow velocity at the t-th sampling point; f(t-1) represents the flow velocity at the t-1-th sampling point; f(tN) represents the flow velocity at the tN-th sampling point; a (t-1)K to a (t-1) They represent the coefficients of the corresponding order terms of f(t-1); a (t-N)K to a (t-N) They represent the coefficients of the corresponding order terms of f(tN); b (t-1) to b (t-N) They represent the corresponding constant terms respectively.

[0017] Preferably, in the prediction method, the data of the plurality of lung function curves that meet the predetermined quality control qualification conditions include:

[0018] Lung function curves of different genders, ages and / or BMI ranges that meet the predetermined quality control qualification criteria.

[0019] Preferably, in the prediction method, the predetermined neural network algorithm is a recurrent neural network algorithm.

[0020] Preferably, in the prediction method, the predetermined training goal is that the verification accuracy of the prediction data of the same pulmonary function curve is greater than a predetermined accuracy threshold.

[0021] Preferably, the prediction method, wherein the step of training the pulmonary function curve prediction model using a training data set and a predetermined neural network algorithm comprises:

[0022] Define a training starting equation and a training ending equation; the ending equation is an N-element K-order equation; the starting equation is an N1-element K1-order equation; N1≤N, K1≤K;

[0023] During the training process, when the number of elements and the order of the equation do not reach the number of elements and the order of the termination equation, it is judged whether the predicted value meets a predetermined verification standard; if the verification standard is not met, the number of elements and / or the order of the equation are increased, and the training is continued; if the predetermined verification standard is met, it is continued to be judged whether the predicted value meets a predetermined training target, and if the predetermined training target is met, the training is terminated, and if the predetermined training target is not met, the number of elements and / or the order of the equation are increased, and the training is continued; wherein the verification standard is: the difference between the predicted value and the true value is within a predetermined error threshold range;

[0024] During the training process, when the number of elements and order of the equations reach the number of elements and order of the termination equation and the predicted value does not meet the verification criteria and / or training objectives, the parameters of the neural network algorithm are adjusted, the starting equation and / or the termination equation are adjusted, and training is continued.

[0025] Preferably, in the prediction method, the accuracy threshold is ≥95%; and the error threshold range is [-5%, +5%].

[0026] On the other hand, a device for predicting a pulmonary function curve is provided, comprising a memory and a processor, wherein the memory stores at least one program, and the at least one program is executed by the processor to implement any of the prediction methods described above.

[0027] On the other hand, a computer-readable storage medium is provided, wherein the storage medium stores at least one program, and the at least one program is executed by a processor to implement any of the prediction methods described above.

[0028] On the other hand, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the prediction method as described above is implemented.

[0029] The above technical solution has the following technical effects:

[0030] The technical solution of the embodiment of the present invention collects a certain amount of quality-controlled pulmonary function curve data and trains and tests a pulmonary function curve prediction model constructed in advance using a Taylor series expansion based on a predetermined neural network algorithm. That is, a relationship model between points in the pulmonary function curve can be used to predict data points at later sampling times using multiple sampling data points at previous moments, thereby achieving effective prediction of the pulmonary function curve of the subject to be tested.

[0031] Furthermore, the validity of invalid lung function curves can be predicted;

[0032] It solves the problem of misdiagnosis when the subjects, such as patients, are unable to master effective lung function test methods, resulting in inaccurate collected data. At the same time, it simplifies the process of lung function monitoring, which is conducive to the implementation of lung function screening for all people. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of a lung function test in the prior art;

[0034] Figure 2 A schematic flow chart of a method for predicting a lung function curve according to an embodiment of the present invention;

[0035] Figure 3 A schematic diagram of a process for training a pulmonary function curve prediction model according to a method for predicting a pulmonary function curve according to an embodiment of the present invention;

[0036] Figure 4 FIG. 4 is a schematic structural diagram of a device for predicting a lung function curve according to an embodiment of the present invention. DETAILED DESCRIPTION

[0037] To further illustrate various embodiments, the present invention is provided with accompanying drawings. These drawings form part of the present disclosure and are primarily used to illustrate the embodiments and, in conjunction with the relevant description in the specification, to explain the operating principles of the embodiments. By referring to these drawings, one of ordinary skill in the art will understand other possible embodiments and the advantages of the present invention. The components in the figures are not drawn to scale, and similar reference numerals are generally used to represent similar components.

[0038] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.

[0039] The basic measurement method of pulmonary function test is to let the subject take a full breath and then exhale quickly and forcefully until all the air is exhaled. Figure 1 Schematic diagram of lung function test in the prior art. Figure 1In this example, PEF stands for peak expiratory flow; FEF25%, FEF50%, and FEF75% represent the instantaneous flow rates at forced expiration of 25%, 50%, and 75% of vital capacity, respectively. TLC stands for total lung capacity; RV stands for residual volume. The present invention primarily focuses on predictive research using expiratory data from pulmonary function tests; pulmonary function conclusions are determined solely based on expiratory data.

[0040] During the exhalation process of a pulmonary function test, the working principle of the lungs is that the diaphragm relaxes and rises, the abdominal muscles contract, and the internal intercostal muscles contract to move the ribs downward, thereby compressing the lung space to expel the lung gas. When the lung gas is expelled, the movement is only affected by the initial kinetic energy obtained during exhalation, the shape of the respiratory tract, and the subsequent exhaled gas molecules.

[0041] The inventors discovered that the exhalation process in pulmonary function tests can be understood through the principles of showerheads. Showerheads can have different shapes, similar to the shape of the respiratory tract. Under conditions of constant water pressure and showerhead, the trajectory of water outflow is fixed. When the showerhead is turned off at the same speed, the final trajectory of the water flow is also fixed. This means that the size of the subsequent water flow affects the flow trajectory. Therefore, the flow trajectory can be predicted based on the initial flow rate.

[0042] At the same time, the human respiratory tract is a relatively stable structure. Based on the above principle, the inventors hypothesized that during pulmonary function testing, the flow rate before exhalation can predict the flow rate after exhalation. An N-dimensional, K-order equation exists that can express the relationship between the flow rates before and after exhalation. Since the water outflow trajectory is fixed and the flow rate changes relatively continuously at different showerhead closing speeds, this equation is presumed to be a linear equation. Taylor series are a tool for approximating complex equations.

[0043] The inventors of the present invention used an N-ary K-order equation and a neural network algorithm, such as a recurrent neural network (RNN), to construct an automated model for predicting pulmonary function curve data under limited sample size conditions. The inventors of the present invention hypothesized that the N-ary K-order equation is a linear equation. In one embodiment, the N-ary K-order equation is represented using its Taylor series expansion.

[0044] Among them, the Taylor series expansion of the function f(x) is shown as follows:

[0045]

[0046] An embodiment of the present invention uses a Taylor series expansion to represent the relationship between the flow rates of the preceding and following sampling points during the exhalation process of a pulmonary function test. In a specific implementation, a Taylor series expansion is used to represent the relationship between the flow rate of the t-th sampling point on the pulmonary function curve and the flow rates of the N sampling points before the t-th sampling point during the exhalation process of a pulmonary function test, where N, K, and t are natural numbers greater than or equal to 1, and K≤N. In a specific implementation, the relationship between the flow rate at sampling time t, i.e., the exhaled flow rate f(t), and the flow rates of the N sampling points before t is represented by the following Taylor series expansion:

[0047] (t) = a (t-1)K [f(t-1)] K +a (t-1)(K-1) [f(t-1)] (K-1) +...+a (t-1) f(t-1)

[0048] +b (t-1) +...+a (t-N)K [f(tN)] K

[0049] +a (t-N)(K-1) [f(tN)] (K-1) +...+a (t-N) f(tN)+b (t-N)

[0050] f(t) represents the flow velocity at the t-th sampling point, i.e., the sampling time t; f(t-1) represents the flow velocity at the t-1-th sampling point; f(tN) represents the flow velocity at the tN-th sampling point; a (t-1)K to a (t-1) They represent the coefficients of the corresponding order terms of f(t-1); a (t-N)K to a (t-N) They represent the coefficients of the corresponding order terms of f(tN); b (t-1) to b (t-N) They represent the corresponding constant terms respectively.

[0051] Take N=2, K=2 as an example to illustrate the construction of Taylor series expression for a quadratic equation of two variables or a two-stage equation of two variables:

[0052] F(x,y)=ax*x+bx+c+d*y*y+e*y+f;

[0053] Among them, a, b, and c in the formula are the coefficients of the quadratic term, the linear term, and the zero-order term of the variable x, respectively; d, e, and f in the formula are the coefficients of the quadratic term, the linear term, and the zero-order term of the variable y, respectively.

[0054] The data for the pulmonary function training sampling points are: [f1,f2,f3,f4...fn], where fn represents the nth flow velocity sampling value. Replace x and y with f(t-1) and f(t-2) to construct the equation for predicting the flow velocity f(t) at the tth sampling point:

[0055] f(t)=k 11 (t-1)*f(t-1)*f(t-1)+k 12 (t-1)*f(t-1)+b(t-1)+k 21 (t-

[0056] 2)*f(t-2)*f(t-2)+k 22 (t-2)*f(t-2)+b(t-2)

[0057] f(t): represents the flow velocity value at the t-th sampling point; k 11 : represents the coefficient of the quadratic term of f(t-1); k 12 : represents the coefficient of the first-order term of f(t-1); k 21 : represents the coefficient of the second-order term of f(t-2); k 22 : represents the coefficient of the first-order term of f(t-2); b(t-1): represents the coefficient of the zero-order term of f(t-1); b(t-2): represents the coefficient of the zero-order term of f(t-2).

[0058] Example 1:

[0059] Figure 2 FIG. 1 is a flow chart of a method for predicting a lung function curve according to an embodiment of the present invention. Figure 2 The prediction method of this embodiment includes the following steps:

[0060] S1, obtaining the flow rate of M sampling points collected during the exhalation process of the subject during the pulmonary function test, where M is a natural number greater than or equal to 1;

[0061] S2, inputting the flow rates of the M sampling points into a pre-built and trained lung function curve prediction model, and using the trained lung function curve prediction model to predict the lung function curve of the subject to be tested;

[0062] The construction of the lung function curve prediction model includes the following steps:

[0063] Construct a Taylor series expansion of an N-order K-variable equation, and use the Taylor series expansion to express the relationship between the flow rate at the t-th sampling point on the pulmonary function curve and the flow rates at the N sampling points before the t-th sampling point during exhalation of a pulmonary function test, where N, K, and t are natural numbers greater than or equal to 1, K≤N, and M≥N.

[0064] The training of the lung function curve prediction model includes the following steps:

[0065] A training data set for a pulmonary function curve prediction model is constructed using data of multiple pulmonary function curves that meet predetermined quality control qualification conditions; preferably, the data of multiple pulmonary function curves that meet predetermined quality control qualification conditions include: pulmonary function curves of different genders, different ages, and / or different BMI ranges that meet predetermined quality control qualification conditions;

[0066] The pulmonary function curve prediction model is trained using a training data set and a predetermined neural network algorithm until a predetermined training target is met; preferably, the predetermined neural network algorithm is a recurrent neural network algorithm (RNN).

[0067] In one specific implementation, the predetermined training goal is: the verification accuracy of the predicted data of the same pulmonary function curve is greater than a predetermined accuracy threshold. Preferably, the accuracy threshold is ≥95%. In a preferred specific implementation, the accuracy threshold = 95%.

[0068] Example 2:

[0069] Figure 3 This is a flow chart of a method for predicting a pulmonary function curve in this embodiment of the present invention, in which a training data set and an RNN algorithm are used to train a pulmonary function curve prediction model, i.e., an N-variable K-order equation expressed by a Taylor series expansion for expressing the relationship between previous and subsequent data points in the pulmonary function curve.

[0070] like Figure 3 In this embodiment, the training of the lung function curve prediction model includes the following steps:

[0071] Prepare a training data set; in a specific implementation, the training data set includes lung function curve data of different genders, different ages and / or different BMI (body mass index) ranges that meet predetermined quality control standards;

[0072] Prepare a test dataset; the test dataset is the same as the training dataset, except that it is not used for model training but for model validation.

[0073] Setting a training goal; preferably, the training goal is that the prediction data verification accuracy of the same curve is greater than 95%;

[0074] Setting a validation standard; in one specific implementation, since the allowable error range for flow rate in the pulmonary function guidelines is ±5%, the validation standard is defined as the difference between the predicted value and the true value cannot exceed ±5%. Less than ±5% indicates that the prediction is accurate;

[0075] Define a starting equation for training and a terminating equation for training; the starting equation is an N1-element and K1-order equation, i.e., a K1-order equation; the terminating equation is an N-element K-order equation, i.e., a K-order equation; wherein N1≤N, K1≤K;

[0076] In one specific implementation, the starting equation is defined as a 5-variable first-order equation, that is, the first 5 sampling points are used to predict the flow velocity of the next sampling point, that is, the sixth sampling point. This is a preferred embodiment, but in other implementations, starting equations with other numbers and orders can be used; for example, the starting equation is defined as a 4-variable first-order equation; or a 2-variable second-order equation; or a 5-variable quadratic equation, etc.; these are not described in detail here.

[0077] In a specific implementation, taking into account computing power and prediction timeliness, for example, to ensure that the data sampling frequency does not exceed 1 second, the order of the termination equation can be set to no more than 5, that is, the highest degree of each element in the equation in the Taylor display does not exceed 5; this is a preferred embodiment, but in other implementations, termination equations with other numbers of elements and orders can be used; for example, a termination equation of 5 elements and 4th order, etc., which will not be repeated here;

[0078] Conduct training and verify the training results, i.e., the predicted values, including:

[0079] Determine whether the predicted value meets the verification criteria; if so, then: further determine whether it meets the training goal; if it meets the training goal, end the training; if it does not meet the training goal, determine whether the current number of equation elements is greater than or equal to N; if not, adjust the model element number by 1 and continue training; if the current number of equation elements is greater than or equal to N; then continue to determine whether the current equation order is greater than or equal to K; if not, adjust the model order by 1 and continue training; if the current equation order is greater than or equal to K, further adjust the RNN network parameters and redefine the starting equation and / or ending equation for training; the training process continues until the number and order of the equation are no greater than the number and order of the ending equation, and the predicted value obtained by training meets both the predetermined verification criteria and the training goal, then end the training; obtain a trained pulmonary function curve prediction model.

[0080] The trained pulmonary function curve prediction model can then be used to predict the subject's pulmonary function curve. Specifically, only a few valid data points, i.e., valid flow rate sampling points, are needed for the subject at the beginning of the pulmonary function test to predict subsequent pulmonary function curve data. The required number of valid data points corresponds to the number of elements in the termination equation of the trained pulmonary function curve prediction model. Specifically, the number of valid data points used for prediction must be greater than or equal to the number of elements in the termination equation of the prediction model.

[0081] The lung function curve prediction method according to the embodiment of the present invention can use a plurality of previous valid flow rate data points to predict the flow rate of subsequent lung function data points with a high accuracy.

[0082] Utilizing embodiments of the present invention, the validity of an invalid pulmonary function curve can be further predicted. Specifically, several data points in the front portion of the pulmonary function curve to be judged for validity can be selected and used as input into a trained prediction model to obtain a predicted pulmonary function curve. The predicted pulmonary function curve is then compared with the pulmonary function curve to be judged for validity. If the deviation is within a predetermined deviation range, the pulmonary function curve to be judged for validity can be determined to be a valid pulmonary function curve; otherwise, the pulmonary function curve to be judged for validity is determined to be an invalid pulmonary function curve.

[0083] Example 3:

[0084] The present invention also provides a device for predicting lung function curves, such as Figure 4 As shown, the device includes a processor 401, a memory 402, a bus 403, and a computer program stored in the memory 402 and executable on the processor 401. The processor 401 includes one or more processing cores. The memory 402 is connected to the processor 401 via the bus 403. The memory 402 is used to store program instructions. When the processor executes the computer program, the steps in the above-mentioned method embodiment of the present invention are implemented.

[0085] Furthermore, as an executable solution, the device for predicting the pulmonary function curve can be a computer unit, which can be a computing device such as a desktop computer, laptop, PDA, or cloud server. The computer unit may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the above-mentioned computer unit structure is merely an example of a computer unit and does not constitute a limitation of the computer unit. The computer unit may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the computer unit may also include input and output devices, network access devices, buses, etc., which are not limited in the embodiments of the present invention.

[0086] Furthermore, as an executable solution, the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the computer unit and connects various parts of the entire computer unit using various interfaces and lines.

[0087] The memory can be used to store the computer programs and / or modules, and the processor implements the various functions of the computer unit by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required for a function; the data storage area can store data created based on the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0088] Example 4:

[0089] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method in the embodiment of the present invention are implemented.

[0090] If the module / unit integrated in the computer unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above-mentioned various method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM) and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0091] Embodiment 5:

[0092] The present invention also provides a computer program product, comprising a computer program, which implements the steps of the method described above when the computer program is executed by a processor.

[0093] Although the present invention has been particularly shown and described in conjunction with preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made to the present invention without departing from the spirit and scope of the invention as defined in the appended claims, and all such changes are within the scope of protection of the present invention.

Claims

1. A method for predicting a lung function curve, characterized in that: include: Obtaining the flow rate of M sampling points collected during the exhalation process of the subject under test during the pulmonary function test; Inputting the flow rates of the M sampling points into a pre-built and trained pulmonary function curve prediction model, and using the trained pulmonary function curve prediction model to predict the pulmonary function curve of the subject to be tested; The construction of the lung function curve prediction model includes the following steps: Constructing a Taylor series expansion of an N-order K-variable equation, and using the Taylor series expansion to represent the relationship between the flow rate at the t-th sampling point on the pulmonary function curve and the flow rates at N sampling points before the t-th sampling point during exhalation of a pulmonary function test, where N, K, and t are natural numbers greater than or equal to 1, K≤N; and M≥N; The training of the pulmonary function curve prediction model includes the following steps: Constructing a training data set for the pulmonary function curve prediction model using data of multiple pulmonary function curves that meet predetermined quality control qualification conditions; The pulmonary function curve prediction model is trained using a training data set and a predetermined neural network algorithm until a predetermined training goal is met.

2. The prediction method according to claim 1, characterized in that The following Taylor series expansion of the Nth-order K-ary equation is used to express the relationship between the flow rate at the tth sampling point on the pulmonary function curve and the flow rates at the N sampling points before the tth sampling point during the exhalation process of the pulmonary function test: f(t) represents the flow velocity at the t-th sampling point; f(t-1) represents the flow velocity at the t-1-th sampling point; f(tN) represents the flow velocity at the tN-th sampling point; to a (t-1) They represent the coefficients of the corresponding degree terms of f(t-1); to a (t-N) They represent the coefficients of the corresponding order terms of f(tN); b (t-1) to b (t-N) They represent the corresponding constant terms respectively.

3. The prediction method according to claim 1, wherein: The data of the multiple pulmonary function curves that meet the predetermined quality control qualification conditions include: Pulmonary function curves of different genders, ages and / or BMI ranges that meet the predetermined quality control qualification criteria.

4. The prediction method according to claim 1, wherein: The predetermined neural network algorithm is a recurrent neural network algorithm.

5. The prediction method according to claim 1, wherein: The predetermined training goal is that the verification accuracy of the prediction data of the same pulmonary function curve is greater than a predetermined accuracy threshold.

6. The prediction method according to claim 5, characterized in that Training the pulmonary function curve prediction model using a training data set and a predetermined neural network algorithm includes: Define a training starting equation and a training ending equation; the ending equation is an N-element K-order equation; the starting equation is an N1-element K1-order equation; N1≤N, K1≤K; During the training process, when the number of elements and the order of the equation do not reach the number of elements and the order of the termination equation, it is judged whether the predicted value meets a predetermined verification standard; if the verification standard is not met, the number of elements and / or the order of the equation are increased, and the training is continued; if the predetermined verification standard is met, it is continued to be judged whether the predicted value meets a predetermined training target, and if the predetermined training target is met, the training is terminated, and if the predetermined training target is not met, the number of elements and / or the order of the equation are increased, and the training is continued; wherein the verification standard is: the difference between the predicted value and the true value is within a predetermined error threshold range; During the training process, when the number of elements and order of the equations reach the number of elements and order of the termination equation and the predicted value does not meet the verification criteria and / or training objectives, the parameters of the neural network algorithm are adjusted, the starting equation and / or the termination equation are adjusted, and training is continued.

7. The prediction method according to claim 6, characterized in that The accuracy threshold is ≥95%; the error threshold range is [-5%, +5%].

8. A device for predicting lung function curve, characterized in that: The method comprises a memory and a processor, wherein the memory stores at least one program, and the at least one program is executed by the processor to implement the prediction method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that The storage medium stores at least one program, and the at least one program is executed by a processor to implement the prediction method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the prediction method according to any one of claims 1 to 7 is implemented.

Images