CN115444367B

CN115444367B - Monitoring method and system of respiratory function rehabilitation training instrument based on artificial intelligence

Info

Publication number: CN115444367B
Application number: CN202211062173.7A
Authority: CN
Inventors: 高天; 王倩; 林仙枝
Original assignee: Guangxi Rio Tinto Medical Technology Co ltd
Current assignee: Guangxi Rio Tinto Medical Technology Co ltd
Priority date: 2022-09-01
Filing date: 2022-09-01
Publication date: 2023-09-05
Anticipated expiration: 2042-09-01
Also published as: CN115444367A

Abstract

The invention discloses a monitoring method and a system of a respiratory function rehabilitation training device based on artificial intelligence, wherein the monitoring method comprises the following steps: the method comprises the steps that a server is used for responding to a respiratory function rehabilitation monitoring request, and a first respiratory state identification model is called for processing, so that respiratory state parameters after respiratory physiological parameter conversion are obtained; and determining a plurality of target users with continuous abnormal respiratory state characteristics based on the respiratory state parameters, and pushing corresponding medical instrument product pushing information and rehabilitation training information pushing information for each target user.

Description

Monitoring method and system of respiratory function rehabilitation training instrument based on artificial intelligence

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a monitoring method and a monitoring system of a respiratory function rehabilitation training instrument based on artificial intelligence.

Background

At present, a corresponding rehabilitation instrument is generally configured for respiratory function rehabilitation training of a patient, besides auxiliary training, physiological parameters of the corresponding patient are recorded, and as the condition of each patient is different, the adopted rehabilitation training strategy and the corresponding acquired physiological parameters are also different, the judgment of how the pathological condition of the patient is determined by the respiratory function rehabilitation training instrument is inconvenient, and the difficulty is caused for the follow-up recommendation of targeted medical instrument products and the pushing of rehabilitation training consultation.

Disclosure of Invention

The invention aims to provide a monitoring method and a system of a respiratory function rehabilitation training device based on artificial intelligence.

In a first aspect, an embodiment of the present invention provides a monitoring system for an artificial intelligence-based respiratory function rehabilitation training device, where the monitoring system for an artificial intelligence-based respiratory function rehabilitation training device includes a server and a respiratory function rehabilitation training device communicatively connected with the server;

the respiratory function rehabilitation training device is used for calling the configured sensor to collect physiological parameters of a user;

the server is used for responding to the respiratory function rehabilitation monitoring request and establishing communication connection with the respiratory function rehabilitation training instrument; acquiring physiological parameters of a user from a respiratory function rehabilitation training instrument; acquiring a respiratory physiological parameter to be evaluated from the physiological parameter of a user, and calling a first respiratory state identification model to process to obtain a respiratory state parameter after the respiratory physiological parameter is converted; according to the respiratory state parameters, a preset respiratory function rehabilitation training table is combined for evaluation, and a monitoring result of the respiratory function rehabilitation training instrument is obtained; determining a plurality of target users with continuous abnormal breathing state characteristics based on monitoring results of the breathing function training instrument; and determining on-target line pushing information matched with each target user according to the continuous abnormal breathing state characteristics of each target user, wherein the on-target line pushing information comprises medical instrument product pushing information and rehabilitation training information pushing information.

In a second aspect, an embodiment of the present invention provides a method for monitoring a respiratory function rehabilitation training device based on artificial intelligence, including:

responding to a respiratory function rehabilitation monitoring request, and establishing communication connection with a respiratory function rehabilitation training instrument;

acquiring physiological parameters of a user from a respiratory function rehabilitation training instrument, wherein the physiological parameters of the user are acquired by a sensor configured by the respiratory function rehabilitation training instrument;

acquiring a respiratory physiological parameter to be evaluated from the physiological parameter of a user, and calling a first respiratory state identification model to process to obtain a respiratory state parameter after the respiratory physiological parameter is converted;

according to the respiratory state parameters, a preset respiratory function rehabilitation training table is combined for evaluation, and a monitoring result of the respiratory function rehabilitation training instrument is obtained;

obtaining a monitoring result of the respiratory function training instrument;

determining a plurality of target users with continuous abnormal breathing state characteristics based on monitoring results of the breathing function training instrument;

and determining on-target line pushing information matched with each target user according to the continuous abnormal breathing state characteristics of each target user, wherein the on-target line pushing information comprises medical instrument product pushing information and rehabilitation training information pushing information.

Compared with the prior art, the invention has the beneficial effects that: the invention discloses a monitoring system of a respiratory function rehabilitation training device based on artificial intelligence, which comprises: the method comprises the steps that a server is used for responding to a respiratory function rehabilitation monitoring request, and a first respiratory state identification model is called for processing, so that respiratory state parameters after respiratory physiological parameter conversion are obtained; and determining a plurality of target users with continuous abnormal respiratory state characteristics based on the respiratory state parameters, and pushing corresponding medical instrument product pushing information and rehabilitation training information pushing information for each target user.

Drawings

In order to more clearly illustrate the technical solution of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described. It is appreciated that the following drawings depict only certain embodiments of the invention and are therefore not to be considered limiting of its scope. Other relevant drawings may be made by those of ordinary skill in the art without undue burden from these drawings.

FIG. 1 is a schematic flow chart of steps of a method for monitoring an artificial intelligence based respiratory function rehabilitation training device according to an embodiment of the present invention;

FIG. 2 is a schematic block diagram of a monitoring system of an artificial intelligence based respiratory function rehabilitation training device according to an embodiment of the present invention;

fig. 3 is a schematic block diagram of a computer device according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

Referring to fig. 1 in combination, fig. 1 is a schematic step flow diagram of a monitoring method of an artificial intelligence-based respiratory function rehabilitation training device according to an embodiment of the present invention, where the monitoring method of the artificial intelligence-based respiratory function rehabilitation training device may be implemented by a server in a monitoring system of the artificial intelligence-based respiratory function rehabilitation training device as an execution subject, and the server is in communication connection with the respiratory function rehabilitation training device, and the specific flow of the monitoring method of the artificial intelligence-based respiratory function rehabilitation training device may be as follows:

210. and responding to the respiratory function rehabilitation monitoring request, and establishing communication connection with the respiratory function rehabilitation training device.

220. The physiological parameters of the user are acquired from the respiratory function rehabilitation training device, and the physiological parameters of the user are acquired by sensors configured by the respiratory function rehabilitation training device.

230. And acquiring the respiratory physiological parameters to be evaluated from the physiological parameters of the user, and calling the first respiratory state identification model to process so as to obtain respiratory state parameters after the respiratory physiological parameters are converted.

240. And (3) carrying out evaluation according to the respiratory state parameters and combining with a preset respiratory function rehabilitation training table to obtain a monitoring result of the respiratory function rehabilitation training instrument.

250. And obtaining a monitoring result of the respiratory function training instrument.

260. And determining a plurality of target users with continuous abnormal breathing state characteristics based on the monitoring result of the breathing function training instrument.

270. And determining target on-line pushing information matched with each target user according to the continuous abnormal breathing state characteristics of each target user, wherein the target on-line pushing information comprises medical instrument product pushing information and rehabilitation training information pushing information.

In the embodiment of the invention, the respiratory function rehabilitation training device can comprise a plurality of sensing devices for collecting a plurality of sensing data of a human body in real time, for example, according to respiratory rehabilitation scenes, the corresponding sensors can be called to monitor the indexes such as electrocardio, blood oxygen saturation, blood pressure, respiratory frequency and the like, so that when the respiratory function rehabilitation monitoring is needed, the respiratory function rehabilitation training device can be firstly connected with the respiratory function rehabilitation training device in a communication way so as to achieve the purpose of acquiring accurate parameters in real time. After the physiological parameters of the user are acquired, the first respiratory state identification model can be called for processing, so as to obtain respiratory state parameters after the respiratory physiological parameters are converted, and it is understood that the physiological parameters of the user are parameters which intuitively reflect the physical state of the patient, the index number is large, and the physiological parameters are objective data, so that the respiratory state of the user can be quickly and accurately identified, the physiological data are required to be converted into respiratory state parameters (such as a corresponding identifier or numerical value, which is not limited here) capable of intuitively reflecting the current specific situation, then the monitoring result of the respiratory function rehabilitation training instrument can be obtained, a plurality of target users with continuous abnormal respiratory state characteristics can be determined according to the monitoring result of the respiratory function rehabilitation training instrument, and the pushing information of each user comprises pushing information of medical equipment products and pushing information of rehabilitation training information according to the monitoring result corresponding to each user, so that the user can be accurately provided with the required consultation of the user instead of the unified consultation pushing information in the prior art.

In order to more clearly describe the solution provided by the embodiments of the present application, the above step 230 may be implemented by the following detailed steps.

101. A respiratory physiological parameter to be evaluated is acquired, the respiratory physiological parameter including at least one respiratory physiological index.

The respiratory physiological parameters are analyzed to obtain various respiratory physiological indexes in the respiratory physiological parameters, and in the embodiment of the application, the respiratory physiological indexes include but are not limited to indexes such as respiratory frequency, blood sugar, heartbeat, blood oxygen concentration and the like. Optionally, in some embodiments, the respiratory physiological parameter may be converted from corresponding sensor data.

102. And obtaining a respiratory feature vector corresponding to at least one respiratory physiological index in respiratory physiological parameters through a first respiratory state recognition model, wherein the first respiratory state recognition model is trained according to a plurality of sample respiratory data pairs, the sample respiratory data pairs comprise sample respiratory state data, the sample respiratory physiological data obtained by converting the sample respiratory state data through a second respiratory state recognition model, the second respiratory state recognition model is trained according to an optimized respiratory data pair, one optimized respiratory data pair comprises one optimized sample respiratory state data and a plurality of optimized sample respiratory physiological data, and the optimized sample respiratory physiological data is mapping data of the optimized sample respiratory state data.

The first and second respiratory state recognition models may be convolutional Neural networks (ConvolutionNeural Network, CNN), recurrent Neural networks (RNN, recurrent Neural Network), time recurrent Neural networks (LSTM, long Short Term Memory), bi-directional recurrent Neural networks (BiRNN, bidirectionalRecurrent Neural networks), and the like. It should be noted that the above examples should not be construed as limiting the first and second respiratory state recognition models.

The second respiratory state recognition model is trained according to the optimized respiratory data pair, one optimized respiratory data pair comprises optimized sample respiratory state data and a plurality of corresponding optimized sample respiratory physiological data with the same scene, and the second respiratory state recognition model is trained according to a plurality of mapping data. The trained second respiratory state recognition model converts sample respiratory state data, and can output corresponding user states of respiratory physiological sides with more diversity.

103. And predicting at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector through a first respiratory state identification model, and converting each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index.

Optionally, in some embodiments, the step of predicting, by the first respiratory state recognition model, at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector, and converting each respiratory physiological index into the first confidence coefficient of the corresponding pre-stored respiratory state index may include:

the breathing characteristic vector is corresponding to the second breathing characteristic matrix through a matching relation between a first breathing characteristic matrix at a breathing physiological side and a second breathing characteristic matrix at a breathing state side in the first breathing state identification model, so that a reference breathing characteristic vector is obtained;

according to the respiration feature vector of the respiration physiological index set at the respiration state side in the second respiration feature matrix and the matching degree of the respiration feature vector, at least two pre-stored respiration state indexes corresponding to each respiration physiological index in the respiration physiological index set are determined, and each respiration physiological index is converted into a first confidence coefficient of the corresponding pre-stored respiration state index.

The respiratory physiological index set belongs to the respiratory state side, and can be composed of a plurality of respiratory physiological indexes.

The respiratory feature vector corresponding to the respiratory physiological index is corresponding to the second respiratory feature matrix to obtain a reference respiratory feature vector, and convolution operation and pooling operation can be performed on the respiratory feature vector of the respiratory physiological index to obtain the reference respiratory feature vector corresponding to the second respiratory feature matrix.

The step of determining at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index in the respiratory state index set according to the respiratory feature vector of the respiratory physiological index set at the respiratory state side in the second respiratory feature matrix and the matching degree of the respiratory feature vector with the reference respiratory feature vector, and converting each respiratory physiological index into a first confidence degree of the corresponding pre-stored respiratory state index may include:

calculating a respiration characteristic vector of the respiration physiological index set at the respiration state side in the second respiration characteristic matrix, and matching the respiration characteristic vector with a reference respiration characteristic vector;

determining at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index in the respiratory physiological index set according to the matching degree; and predicting the first confidence that each respiratory physiological index is converted into the corresponding pre-stored respiratory state index.

The matching degree may specifically be a euclidean distance, a cosine distance, a manhattan distance, or the like, which is not limited in this embodiment. The smaller the matching degree is, the larger the difference between the breathing physiological index set and the scene of the corresponding breathing physiological index is, and the larger the matching degree is, the closer the breathing physiological index set and the scene of the corresponding breathing physiological index are.

In this embodiment, the respiratory physiological index set with the matching degree smaller than the preset distance may be determined as a pre-stored respiratory status index, where the preset distance may be set according to the actual situation, which is not limited in this embodiment. For example, the number of pre-stored respiratory status indicators that may be acquired as needed may be set.

The first confidence level of each respiratory physiological index converted into the corresponding pre-stored respiratory state index can be predicted through the full connection layer in the first respiratory state identification model.

104. And determining the predicted respiratory state index corresponding to each respiratory physiological index from the prestored respiratory state indexes according to the first confidence coefficient through the first respiratory state identification model.

For each respiratory physiological index, the pre-stored respiratory state index with the largest first confidence coefficient can be used as a predicted respiratory state index corresponding to the respiratory physiological index.

105. And performing weighted average operation on each predicted respiratory state index through the first respiratory state identification model to obtain respiratory state parameters after respiratory physiological parameter conversion.

The first respiratory state recognition model and the second respiratory state recognition model are trained, and the second respiratory state recognition model is trained according to a plurality of mapping data, so that sample respiratory physiological data in a sample respiratory data pair constructed by the second respiratory state recognition model are relatively more diversified; the first breathing state recognition model is obtained by training a sample breathing data pair constructed according to the second breathing state recognition model, and for the first breathing state recognition model, the diversity of preset user breathing state parameters in the sample breathing data pair can be increased, so that the conversion quality of the first breathing state recognition model is enhanced through richer information.

The fusion mode of the predicted respiratory state indexes can be specifically that all the predicted respiratory state indexes are spliced according to a certain mode to obtain respiratory state parameters.

It should be noted that the first respiratory state recognition model may be trained by a plurality of sample training sets, where the sample training sets may include sample respiratory data pairs and initial respiratory data pairs, where each initial respiratory data pair includes paired initial sample respiratory physiological data and initial sample respiratory state data, and the initial sample respiratory physiological data and the initial sample respiratory state data have the same scene. The first respiratory state recognition model may be provided to the artificial intelligence based switching device after being trained by other devices, or may be trained by the artificial intelligence based switching device itself.

If the artificial intelligence based conversion device performs training by itself, the following examples are further provided in the embodiments of the present invention before the step of "obtaining the respiratory physiological parameter to be evaluated:

acquiring sample respiratory state data and a plurality of initial respiratory data pairs, wherein each initial respiratory data pair comprises paired initial sample respiratory physiological data and initial sample respiratory state data;

Converting the sample respiratory state data through the second respiratory state recognition model to obtain converted sample respiratory physiological data, and combining the converted sample respiratory physiological data with the corresponding sample respiratory state data to form a sample respiratory data pair;

determining a sample breath data pair and an initial breath data pair as a sample training set of a first breath state recognition model, wherein initial sample breath physiological data and sample breath physiological data are original sample breath physiological data, and initial sample breath state data and sample breath state data are original sample breath state data;

and converting the original sample respiratory physiological data through the first respiratory state identification model to obtain converted predicted sample respiratory state data, and performing conversion training from a respiratory physiological side to a respiratory state side on the first respiratory state identification model according to a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data.

Wherein, a group of sample training sets comprises paired original sample respiratory physiological data and original sample respiratory state data, and the sample training sets can be regarded as a combination of the sample respiratory data pair and the initial respiratory data pair.

Wherein the source data of the first respiratory state recognition model is the target data of the second respiratory state recognition model, the target data of the first respiratory state recognition model is the source data of the second respiratory state recognition model, that is, the source data of the first respiratory state recognition model is the respiratory physiological side, and the target data of the first respiratory state recognition model is the respiratory state side. It will be appreciated that the first respiratory state recognition model may be regarded as a forward respiratory state recognition model and the second respiratory state recognition model as a backward respiratory state recognition model. Wherein a forward direction may be considered a source data to target data direction and a reverse direction may be considered a target data to source data direction.

According to the application, the sample respiratory state data can be converted into the sample respiratory physiological data (namely the sample respiratory physiological data mentioned in the previous embodiment) through the backward respiratory state recognition model, so that a sample respiratory data pair is constructed, the combination of the sample respiratory state data set and the optimized sample respiratory state data set is enhanced, the quality of the forward respiratory state recognition model is further improved, and the flow can be seen in the following examples:

1001. training a reverse breathing state recognition model by using a training target driven by diversity, wherein the diversity specifically refers to the diversity of mapping data; specifically, optimized sample breath state data set pairs can be obtained, each optimized sample breath state data set pair can comprise optimized sample breath state data and a plurality of optimized sample breath physiological data corresponding to the optimized sample breath state data, the optimized sample breath physiological data are mapping data of the optimized sample breath state data, the optimized sample breath state data are converted through a reverse breath state identification model to obtain converted predicted sample breath physiological data, and the reverse breath state identification model is trained according to a cost function between each mapping data corresponding to the optimized sample breath state data and the predicted sample breath physiological data;

1002. Converting sample respiratory state data by using a trained reverse respiratory state recognition model, and outputting recognition data with more diversity, wherein the recognition data is sample respiratory physiological data, and the sample respiratory physiological data and the corresponding sample respiratory state data form sample respiratory data pairs, so that the information of the constructed sample respiratory data pair source data is enriched;

1003. merging the sample breath data pair with the optimized sample breath state data set to obtain merged sample training data, wherein the merged sample training data comprises merged optimized sample breath physiological data and merged optimized sample breath state data; and taking the combined optimized sample respiratory physiological data as input of a forward respiratory state recognition model, obtaining converted predicted sample respiratory state data, and training the forward respiratory state recognition model according to cross entropy of maximum likelihood estimation between the predicted sample respiratory state data and the corresponding optimized sample respiratory state data.

In one embodiment, the respiratory physiological side is chinese, the respiratory state side is english, the first respiratory state recognition model is a respiratory state recognition model of a translated english, and the second respiratory state recognition model is a respiratory state recognition model of a translated english.

It is emphasized, among other things, that this second respiratory state recognition model is already trained. Since the second respiratory state recognition model is trained from the plurality of mapping data, the sample respiratory physiological data in the sample respiratory data pair constructed by the second respiratory state recognition model is relatively more diverse. For the first breathing state identification model, the diversity of the preset user breathing state parameters in the sample breathing data pair can be increased, and the preset breathing state data are enriched.

Optionally, in some embodiments, the step of converting, by the first respiratory state recognition model, the original sample respiratory physiological data to obtain converted predicted sample respiratory state data, and performing, according to a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data, a conversion training from a respiratory physiological side to a respiratory state side on the first respiratory state recognition model may include:

obtaining a respiration characteristic vector corresponding to each sample respiration physiological index in the original sample respiration physiological data through a first respiration state identification model;

Predicting at least two undetermined sample respiratory physiological indexes corresponding to each sample respiratory physiological index according to the respiratory feature vector, and converting each sample respiratory physiological index into a fourth confidence coefficient of the corresponding undetermined sample respiratory physiological index;

determining target indexes corresponding to the breath physiological indexes of the samples from the breath physiological indexes of the samples to be determined according to the fourth confidence coefficient, and performing weighted average operation on the target indexes to obtain breath state data of the predicted samples;

calculating a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data;

and optimizing the model framework of the first respiratory state recognition model according to the first price function to obtain a trained first respiratory state recognition model.

In the optimization process of the first breath state recognition model, the fourth confidence coefficient of the breath physiological index of the undetermined sample, which is close to the application scene matching degree in the mapping data (namely the corresponding original sample breath state data), is increased.

The specific process of predicting the undetermined sample respiratory physiological index corresponding to each sample respiratory physiological index according to the respiratory feature vector of each sample respiratory physiological index can refer to the above embodiment. The step of predicting at least two pending sample respiratory physiological indexes corresponding to each sample respiratory physiological index according to the respiratory feature vector, and converting each sample respiratory physiological index into a fourth confidence coefficient of the corresponding pending sample respiratory physiological index may include:

according to the respiratory feature vector of the respiratory physiological index set at the respiratory state side in the second respiratory feature matrix and the matching degree of the respiratory feature vector with the reference respiratory feature vector, determining at least two undetermined sample respiratory physiological indexes corresponding to each sample respiratory physiological index in the respiratory physiological index set, and converting each sample respiratory physiological index into a fourth confidence coefficient of the corresponding undetermined sample respiratory physiological index.

For each sample respiratory physiological index, the undetermined sample respiratory physiological index with the highest fourth confidence coefficient can be used as a target index corresponding to the sample respiratory physiological index.

Wherein the first respiratory state recognition model outputs each of the sample respiratory state data not only guided by the mapping data in the training set (i.e., the corresponding raw sample respiratory state data), but also considers those state data that are not covered by the reference data in the mapping data, but may be more accurate.

Wherein, the step of calculating a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data may include:

calculating index difference cost of each target index in the predicted sample respiratory state data and corresponding original data in the original sample respiratory state data;

and performing weighted average operation on index difference cost corresponding to each target index in the predicted sample respiratory state data to obtain a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data.

The index difference cost can be calculated according to the breathing characteristic vector corresponding to the target index and the vector distance between the breathing characteristic vectors corresponding to the original data. The vector distance may include a euclidean distance or a cosine distance, etc.

The fusing of the index difference costs corresponding to the target indexes may specifically be that the index difference costs corresponding to the target indexes are weighted and summed to obtain the first cost function.

Optionally, in some embodiments, the step of calculating a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data may include:

Obtaining confidence coefficient of original data of each sample respiration physiological index converted into corresponding original sample respiration state data;

according to the confidence, a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data is calculated.

Optionally, in some embodiments, the step of optimizing the model architecture of the first respiratory state recognition model according to the first cost function to obtain a trained first respiratory state recognition model may specifically include: and optimizing the model framework of the first respiratory state recognition model by adopting a back propagation algorithm (BP, backPropagation) or a random gradient Descent algorithm (SGD, stochasticGradient Descent), and optimizing the model framework of the first respiratory state recognition model according to the first price function to ensure that the first price function is smaller than a preset cost function, so as to obtain a trained first respiratory state recognition model. The preset cost function can be set according to actual conditions.

The second breath state recognition model is trained according to an optimized breath data pair, and one optimized breath data pair comprises optimized sample breath state data and a plurality of corresponding optimized sample breath physiological data. The second respiratory state recognition model may be provided to the artificial intelligence based switching device after being trained by other devices, or may be trained by the artificial intelligence based switching device itself.

Alternatively, in this embodiment, the first respiratory state recognition model may be pre-trained according to the pair of auxiliary respiratory data. A pair of auxiliary respiratory data includes a target sample respiratory physiological data, and a corresponding plurality of optimized sample respiratory state data, the optimized sample respiratory state data being a reference transformation of the target sample respiratory physiological data. The first respiratory state recognition model may be provided to the artificial intelligence based switching device after being trained by other devices, or may be trained by the artificial intelligence based switching device itself.

(1) Training of a second respiratory state recognition model.

If the second respiratory state recognition model is trained by the conversion device according to the artificial intelligence, before the step of converting the sample respiratory state data through the second respiratory state recognition model to obtain converted sample respiratory physiological data, the following example is further provided:

obtaining an optimized respiratory data pair, wherein one optimized respiratory data pair comprises one optimized sample respiratory state data and a plurality of corresponding optimized sample respiratory physiological data, and the optimized sample respiratory physiological data is a reference conversion of the optimized sample respiratory state data;

Converting the optimized sample respiratory state data through a preset second respiratory state identification model to obtain converted predicted sample respiratory physiological data;

calculating a second cost function between the predicted sample respiratory physiological data of the same optimized sample respiratory state data and the corresponding optimized sample respiratory physiological data;

and optimizing a model framework of a preset second respiratory state recognition model according to the second cost function to obtain the second respiratory state recognition model.

The model architecture of the preset second respiratory state recognition model can be optimized by adopting a back propagation algorithm or a random gradient descent algorithm, and the model architecture of the preset second respiratory state recognition model is optimized according to a second cost function, so that the second cost function is smaller than the preset cost function, and the second respiratory state recognition model is obtained. The preset cost function can be set according to actual conditions.

Optionally, in some embodiments, the step of converting the optimized sample breath state data by presetting a second breath state recognition model to obtain converted predicted sample breath physiological data may include:

obtaining breath characteristic vectors corresponding to all target indexes in the breath state data of the optimized sample through presetting a second breath state identification model;

Predicting at least two pre-stored respiratory physiological index variables corresponding to each target index according to the respiratory feature vector, and converting each target index into a second confidence coefficient of the corresponding pre-stored respiratory physiological index variable;

determining target respiratory physiological index variables corresponding to all target indexes from prestored respiratory physiological index variables according to the second confidence coefficient;

and performing weighted average operation on each target respiratory physiological index variable to obtain predicted sample respiratory physiological data after the conversion of the optimized sample respiratory state data.

For each target index, the pre-stored respiratory physiological index variable with the maximum second confidence coefficient can be used as a target respiratory physiological index variable corresponding to the target index, and the target respiratory physiological index variable can also be regarded as a conversion result of the target index. The fusion mode of the target respiratory physiological index variables can be specifically that the target respiratory physiological index variables are spliced according to a certain mode to obtain the respiratory physiological data of the prediction sample.

Optionally, in some embodiments, the step of predicting at least two pre-stored respiratory physiological index variables corresponding to each target index according to the respiratory feature vector, and converting each target index into the second confidence coefficient of the corresponding pre-stored respiratory physiological index variable may include:

Acquiring a respiratory physiological index variable set, wherein the respiratory physiological index variable set comprises a plurality of respiratory physiological index variables;

the method comprises the steps that through presetting a matching relation between a second respiration characteristic matrix at a respiration state side and a first respiration characteristic matrix at a respiration physiological side in a second respiration state identification model, a respiration characteristic vector of a target index is corresponding to the first respiration characteristic matrix, and a first reference respiration characteristic vector is obtained;

according to the respiratory feature vector of the respiratory physiological index variable in the first respiratory feature matrix and the matching degree of the respiratory feature vector with the first reference respiratory feature vector, determining at least two pre-stored respiratory physiological index variables corresponding to each target index in the respiratory physiological index variable, and converting each target index into a second confidence coefficient of the corresponding pre-stored respiratory physiological index variable.

The respiratory physiological index variable set may be a respiratory physiological side full physiological database, or may be a subset of the respiratory physiological side full physiological database.

The respiratory feature vector of the target index is corresponding to the first respiratory feature matrix to obtain a first reference respiratory feature vector, and convolution operation and pooling operation can be performed on the respiratory feature vector of the target index to obtain the first reference respiratory feature vector corresponding to the first respiratory feature matrix.

The matching degree may specifically be a euclidean distance, a cosine distance, a manhattan distance, or the like, which is not limited in this embodiment. The respiratory physiological index variable with the matching degree smaller than the preset distance can be determined as a pre-stored respiratory physiological index variable, the preset distance can be set according to actual conditions, and the embodiment is not limited to this.

In the optimization process of presetting the second respiratory state recognition model, the second confidence coefficient of the pre-stored respiratory physiological index variable, which is similar to the application scene matching degree in the mapping data (namely the corresponding optimized sample respiratory physiological data), is increased.

Optionally, in some embodiments, the step of "calculating the second cost function between the predicted sample respiratory physiological data and the corresponding optimized sample respiratory physiological data of the same optimized sample respiratory state data" may include:

for the same optimized sample respiratory state data, calculating a pre-stored respiratory physiological index variable corresponding to a target index of the same optimized sample respiratory state data, and a first homogeneity coefficient between contrast indexes corresponding to the target index in the optimized sample respiratory physiological data corresponding to the optimized sample respiratory state data;

calculating index difference cost between prestored respiratory physiological index variables corresponding to all target indexes in the optimized sample respiratory state data and the contrast indexes according to the first homogeneous coefficient and the second confidence coefficient;

And performing weighted average operation on index difference cost corresponding to each target index to obtain a second cost function between the predicted sample respiratory physiological data of the optimized sample respiratory state data and the corresponding optimized sample respiratory physiological data.

The target index corresponds to a plurality of prestored respiratory physiological index variables, and the first homogeneity coefficient comprises homogeneity coefficients between each prestored respiratory physiological index variable corresponding to the target index and the corresponding contrast index.

The first homogeneity coefficient can be measured by a vector distance, and the vector distance specifically refers to a vector distance between a respiratory feature vector of a pre-stored respiratory physiological index variable and a respiratory feature vector of a contrast index. The larger the vector distance, the smaller the first homogeneity coefficient; the smaller the vector distance, the larger the first homogeneity coefficient. The vector distance may be a euclidean distance or a cosine distance.

The weighted average operation is performed on the index difference costs corresponding to the target indexes, which may be specifically performed by weighted summation of the index difference costs corresponding to the target indexes.

(2) Pre-training of the first respiratory state recognition model.

If the first respiratory state recognition model is pre-trained by the artificial intelligence conversion device, before the step of converting the original sample respiratory physiological data through the first respiratory state recognition model to obtain converted predicted sample respiratory state data, the following example is further provided:

Acquiring a pair of auxiliary respiratory data, wherein one pair of auxiliary respiratory data comprises one target sample respiratory physiological data and a corresponding plurality of optimized sample respiratory state data, and the optimized sample respiratory state data is a reference conversion of the target sample respiratory physiological data;

converting the target sample respiratory physiological data by presetting a first respiratory state identification model to obtain converted predicted and optimized sample respiratory state data;

calculating a third price function between predicted optimized sample breath state data of the same target sample breath physiological data and each corresponding optimized sample breath state data;

and optimizing a model framework of a preset first respiratory state recognition model according to the third price function to obtain the first respiratory state recognition model.

The model architecture of the preset first respiratory state recognition model can be optimized by adopting a back propagation algorithm or a random gradient descent algorithm, and the model architecture of the preset first respiratory state recognition model is optimized according to a third cost function, so that the third cost function is smaller than the preset cost function, and the first respiratory state recognition model is obtained. The preset cost function can be set according to actual conditions.

Optionally, in some embodiments, the step of converting the target sample respiratory physiological data by presetting a first respiratory state identification model to obtain converted predicted and optimized sample respiratory state data may include:

acquiring a respiration characteristic vector corresponding to each target index in the respiration physiological data of the target sample by presetting a first respiration state identification model;

predicting at least two pre-stored respiratory state index variables corresponding to each target index according to the respiratory feature vector, and converting each target index into a third confidence coefficient of the corresponding pre-stored respiratory state index variable;

determining target respiratory state index variables corresponding to all target indexes from prestored respiratory state index variables according to the third confidence coefficient;

and performing weighted average operation on each target respiratory state index variable to obtain the predicted and optimized sample respiratory state data after the target sample respiratory physiological data conversion.

For each target index, the pre-stored respiratory state index variable with the highest third confidence coefficient can be used as a target respiratory state index variable corresponding to the target index, and the target respiratory state index variable can also be regarded as a conversion result of the target index. The fusion mode of the target respiratory state index variables can be specifically that the target respiratory state index variables are spliced according to a certain mode to obtain the respiratory state data of the prediction optimization sample.

Optionally, in some embodiments, the step of predicting at least two pre-stored respiratory state index variables corresponding to each target index according to the respiratory feature vector, and converting each target index into a third confidence coefficient of the corresponding pre-stored respiratory state index variable may include:

acquiring a respiratory state index variable set, wherein the respiratory state index variable set comprises a plurality of respiratory state index variables;

the method comprises the steps that through presetting a matching relation between a first respiration characteristic matrix on a respiration physiological side and a second respiration characteristic matrix on a respiration state side in a first respiration state identification model, a respiration characteristic vector of a target index is corresponding to the second respiration characteristic matrix, and a second reference respiration characteristic vector is obtained;

and determining at least two pre-stored respiratory state index variables corresponding to each target index in the respiratory state index variables and a third confidence coefficient for converting each target index into the corresponding pre-stored respiratory state index variable according to the respiratory feature vector of the respiratory state index variable in the second respiratory feature matrix and the matching degree of the respiratory feature vector with the second reference respiratory feature vector.

The respiratory state index variable set may be a respiratory state-side full-physiological database, or may be a subset of a respiratory state-side full-physiological database.

The respiratory feature vector of the target index is corresponding to the second respiratory feature matrix to obtain a second reference respiratory feature vector, and convolution operation and pooling operation can be performed on the respiratory feature vector of the target index to obtain the second reference respiratory feature vector corresponding to the second respiratory feature matrix.

The matching degree may specifically be a euclidean distance, a cosine distance, a manhattan distance, or the like, which is not limited in this embodiment. The respiratory state index variable with the matching degree smaller than the preset distance can be determined as the pre-stored respiratory state index variable, the preset distance can be set according to actual conditions, and the embodiment is not limited to this.

In the optimization process of the preset first respiratory state recognition model, the third confidence coefficient of the pre-stored respiratory state index variable, which is similar to the application scene matching degree in the mapping data (namely the corresponding optimized sample respiratory state data), is increased.

Optionally, in some embodiments, the step of "calculating a third cost function between predicted optimized sample breath state data and corresponding optimized sample breath state data for the same target sample breath physiological data" may include:

For the same target sample respiratory physiological data, calculating a pre-stored respiratory state index variable corresponding to a target index of the respiratory physiological data, and a second homogeneity coefficient between contrast indexes corresponding to the target index in optimized sample respiratory state data corresponding to the target sample respiratory physiological data;

calculating index difference cost between prestored respiratory state index variables corresponding to all target indexes in the target sample respiratory physiological data and the contrast indexes according to the second homogeneity coefficient and the third confidence coefficient;

and performing weighted average operation on index difference cost corresponding to each target index to obtain a third price function between the predicted and optimized sample respiratory state data of the target sample respiratory physiological data and the corresponding optimized sample respiratory state data.

The target index corresponds to a plurality of prestored respiratory state index variables, and the second homogeneity coefficient comprises homogeneity coefficients between each prestored respiratory state index variable corresponding to the target index and the corresponding contrast index.

The second homogeneity coefficient can be measured by a vector distance, which refers to a vector distance between a breathing characteristic vector of a pre-stored breathing state index variable and a breathing characteristic vector of a contrast index. The larger the vector distance, the smaller the second homogeneity coefficient; the smaller the vector distance, the larger the second homogeneity coefficient. The vector distance may be a euclidean distance or a cosine distance.

The specific calculation method of the third cost function may refer to description of the related embodiment (calculation of the second cost function) in the training of the second respiratory state recognition model, which is not described herein.

As can be seen from the above, the embodiment can obtain the respiratory physiological parameters to be evaluated, where the respiratory physiological parameters include at least one respiratory physiological index; obtaining a respiratory feature vector corresponding to at least one respiratory physiological index in respiratory physiological parameters through a first respiratory state recognition model, wherein the first respiratory state recognition model is trained according to a plurality of sample respiratory data pairs, the sample respiratory data pairs comprise sample respiratory state data, the sample respiratory physiological data obtained by converting the sample respiratory state data through a second respiratory state recognition model, the second respiratory state recognition model is trained according to an optimized respiratory data pair, one optimized respiratory data pair comprises one optimized sample respiratory state data and a plurality of optimized sample respiratory physiological data, and the optimized sample respiratory physiological data is mapping data of the optimized sample respiratory state data; predicting at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector through a first respiratory state identification model, and converting each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index; determining predicted respiratory state indexes corresponding to each respiratory physiological index from prestored respiratory state indexes according to a first confidence coefficient through a first respiratory state identification model; and performing weighted average operation on each predicted respiratory state index through the first respiratory state identification model to obtain respiratory state parameters after respiratory physiological parameter conversion. According to the embodiment of the application, the second respiratory state recognition model can be trained according to the plurality of optimized sample respiratory physiological data, so that the diversity of constructed sample respiratory data pairs is increased, richer preset respiratory state data are provided for training the first respiratory state recognition model, and the conversion quality is further improved.

The method according to the previous embodiment will be described in further detail below with the specific integration of the conversion device according to artificial intelligence in a server, which may be a cloud server or the like.

In order to more clearly describe the solution provided by the embodiments of the present application, the embodiments of the present application further provide the following examples:

201. the server receives the to-be-evaluated respiratory physiological parameters sent by the terminal, wherein the respiratory physiological parameters comprise at least one respiratory physiological index.

202. The method comprises the steps that a server obtains a respiration characteristic vector corresponding to at least one respiration physiological index in respiration physiological parameters through a first respiration state identification model, wherein the first respiration state identification model is obtained through training according to a plurality of sample respiration data pairs, the sample respiration data pairs comprise sample respiration state data, the sample respiration physiological data obtained through conversion of the sample respiration state data by a second respiration state identification model are obtained through training according to an optimized respiration data pair, one optimized respiration data pair comprises one optimized sample respiration state data and a plurality of optimized sample respiration physiological data, and the optimized sample respiration physiological data is mapping data of the optimized sample respiration state data.

In this embodiment, the first respiratory state recognition model and the second respiratory state recognition model may be NMT models according to a recurrent neural network, a convolutional neural network, and self-attention, NMT models using RNN, CNN, and self-attention in a mixed manner, or the like.

The second respiratory state recognition model is trained according to the optimized respiratory data pair, wherein one optimized respiratory data pair comprises one optimized sample respiratory state data and a plurality of corresponding optimized sample respiratory physiological data, namely the second respiratory state recognition model is trained according to a plurality of mapping data. The trained second respiratory state recognition model converts sample respiratory state data, and can output corresponding user states of respiratory physiological sides with more diversity.

203. The server predicts at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector through a first respiratory state identification model, and converts each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index.

204. The server determines predicted respiratory state indexes corresponding to the respiratory physiological indexes from the prestored respiratory state indexes according to the first confidence coefficient through the first respiratory state identification model.

205. And the server executes weighted average operation on each predicted respiratory state index through the first respiratory state identification model to obtain respiratory state parameters after respiratory physiological parameter conversion.

206. The server sends the breathing state parameters to the terminal.

The first respiratory state recognition model and the second respiratory state recognition model are trained, and the second respiratory state recognition model is trained according to a plurality of mapping data, so that sample respiratory physiological data in a sample respiratory data pair constructed by the second respiratory state recognition model are relatively more diversified; the first respiratory state recognition model is obtained by training a sample respiratory data pair constructed according to the second respiratory state recognition model, and for the first respiratory state recognition model, the diversity of the respiratory state parameters of a preset user in the sample respiratory data pair can be increased, so that the conversion quality of the first respiratory state recognition model is enhanced through richer respiratory state parameters.

After receiving the respiratory state parameter, the terminal may display the converted respiratory state parameter on a display of the electronic device.

In a specific embodiment, the first respiratory state recognition model is used as a forward respiratory state recognition model, the second respiratory state recognition model is used as a backward respiratory state recognition model, the forward respiratory state recognition model converts the source data content to the target data content, the backward respiratory state recognition model converts the target data content to the source data content, and the specific training process of the forward respiratory state recognition model and the backward respiratory state recognition model is as follows:

2001. Acquiring an initial respiratory data pair, which specifically comprises a source data content K0 and a target data content H0;

2002. the training direction is selected, so that a scheme for increasing the diversity of the identification data of the forward respiratory state identification model can be selected, and a scheme for improving the conversion quality of the forward respiratory state identification model by improving the reverse conversion technology can be selected;

2003. if step 2002 selects a scheme for increasing the diversity of the identification data of the forward respiratory state identification model, the forward respiratory state identification model may be trained according to the training targets driven by the diversity, specifically, multiple target end candidate identification data may be obtained, both the target end candidate identification data and the target data content H0 may be used as target data mapping data of the source data content K0 (may be considered as the optimized sample respiratory state data mentioned in the foregoing embodiment), and the forward respiratory state identification model is trained according to the target data mapping data and the source data content K0;

2004. according to training in step 2003, optimizing a model architecture of the forward breathing state recognition model, so that a cost function between the converted predicted and optimized sample breathing state data and target data mapping data meets a preset condition, and obtaining a trained forward breathing state recognition model M, wherein the forward breathing state recognition model M can increase diversity of recognition data;

2005. If step 2002 selects a scheme of improving the conversion quality of the forward respiratory state recognition model by improving the reverse conversion technology, the reverse respiratory state recognition model may be trained according to the training targets driven by diversity, specifically, multiple source candidate recognition data may be obtained, both the source candidate recognition data and the source data content K0 may be used as source data mapping data (which may be considered as the optimized sample respiratory physiological data mentioned in the previous embodiment) of the target data content H0, and the reverse respiratory state recognition model may be trained according to the source data mapping data and the target data content H0;

2006. according to training in step 2005, optimizing a model architecture of a backward breathing state recognition model, so that a cost function between the converted predicted sample breathing physiological data and the source data mapping data meets a preset condition, and obtaining a trained backward breathing state recognition model;

2007. acquiring single target respiratory state data H1;

2008. according to the reverse breathing state recognition model trained in the step 2006, converting the single target breathing state data H1 to obtain a pseudo source data content K1, wherein the pseudo source data content K1 and the single target breathing state data H1 form a sample breathing data pair, and the reverse breathing state recognition model is trained according to a plurality of source data mapping data, so that the diversity of the constructed pseudo source data content K1 can be increased, and richer preset breathing state data can be provided;

2009. Combining the sample respiratory data pair with the initial respiratory data pair to obtain a sample training set of a forward respiratory state recognition model, wherein the sample training set comprises source data content (K0+K1) and target data content (H0+H2);

2010. the source data content (K0+K1) can be used as input of a forward respiratory state recognition model, converted predicted sample respiratory state data is obtained, and the forward respiratory state recognition model is subjected to conversion training from the source data to the target data according to cross entropy of maximum likelihood estimation between the predicted sample respiratory state data and the corresponding target data content (H0+H2);

2011. through training in step 2010, optimizing a model framework of a forward respiration state recognition model so that cross entropy of maximum likelihood estimation meets preset conditions, and obtaining a trained forward respiration state recognition model M1; because the backward breathing state recognition model has stronger capability of outputting various recognition data, the source end of the combined sample training set contains richer information, and the conversion quality of the final forward breathing state recognition model M1 can be improved.

Optionally, in some embodiments, training to increase the diversity of the identification data may be performed on the forward respiratory state identification model (see step 2001-step 2004), or the conversion quality of the forward respiratory state identification model may be enhanced by merely improving the reverse conversion technique (see step 2001, step 2002, and step 2005-step 2011).

Alternatively, in other embodiments, the forward respiratory state recognition model M1 may be trained based on the forward respiratory state recognition model M obtained in step 2004, that is, the forward respiratory state recognition model is trained twice; specifically, the forward respiration state recognition model and the backward respiration state recognition model can be trained to increase the diversity of recognition data, a trained forward respiration state recognition model M and a trained backward respiration state recognition model are obtained, a sample respiration data pair is obtained through the trained backward respiration state recognition model, and the forward respiration state recognition model M is trained to convert source data into target data, so that a forward respiration state recognition model M1 is obtained.

In the current related art, training of respiratory state recognition models typically uses initial respiratory data with a single reference to model architectural optimization of the library, which limits the use of resources and blindly and unreasonably approximates the reference transformations.

According to the embodiment of the application, the reverse breathing state recognition model can be trained according to the training targets driven by diversity, so that the diversity of preset user breathing state parameters of the constructed training set is increased, the forward breathing state recognition model is enhanced by enriching source data information in the training set, the forward breathing state recognition model is combined with cross entropy loss of the neural network machine breathing state recognition model NMT, and the conversion quality of the forward breathing state recognition model is finally improved.

The application can be used for any neural network machine conversion system (such as rnns search, transformer, etc.) implemented according to any deep learning framework.

As can be seen from the above, in this embodiment, the server may receive the respiratory physiological parameter to be evaluated sent by the terminal, where the respiratory physiological parameter includes at least one respiratory physiological index; obtaining a respiratory feature vector corresponding to at least one respiratory physiological index in respiratory physiological parameters through a first respiratory state recognition model, wherein the first respiratory state recognition model is trained according to a plurality of sample respiratory data pairs, the sample respiratory data pairs comprise sample respiratory state data, the sample respiratory physiological data obtained by converting the sample respiratory state data through a second respiratory state recognition model, the second respiratory state recognition model is trained according to an optimized respiratory data pair, one optimized respiratory data pair comprises one optimized sample respiratory state data and a plurality of optimized sample respiratory physiological data, and the optimized sample respiratory physiological data is mapping data of the optimized sample respiratory state data; the server predicts at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector through a first respiratory state identification model, and converts each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index; determining predicted respiratory state indexes corresponding to each respiratory physiological index from prestored respiratory state indexes according to a first confidence coefficient through a first respiratory state identification model; performing weighted average operation on each predicted respiratory state index through a first respiratory state identification model to obtain respiratory state parameters after respiratory physiological parameter conversion; the server sends the breathing state parameters to the terminal. According to the embodiment of the application, the second respiratory state recognition model can be trained according to the plurality of optimized sample respiratory physiological data, so that the diversity of constructed sample respiratory data pairs is increased, richer preset respiratory state data are provided for training the first respiratory state recognition model, and the conversion quality is further improved.

Referring to fig. 2 in combination, fig. 2 is a schematic block diagram of a monitoring system 110 of an artificial intelligence based respiratory function rehabilitation training device according to an embodiment of the invention, including:

the response module 1101 is configured to establish a communication connection with the respiratory function rehabilitation training device in response to the respiratory function rehabilitation monitoring request.

The obtaining module 1102 is configured to obtain a physiological parameter of a user from the respiratory function rehabilitation training device, where the physiological parameter of the user is collected by a sensor configured by the respiratory function rehabilitation training device.

The monitoring module 1103 is configured to obtain a respiratory physiological parameter to be evaluated from the physiological parameters of the user, and invoke the first respiratory state recognition model to perform processing, so as to obtain a respiratory state parameter after the respiratory physiological parameter is converted; and (3) carrying out evaluation according to the respiratory state parameters and combining with a preset respiratory function rehabilitation training table to obtain a monitoring result of the respiratory function rehabilitation training instrument.

It should be noted that, the implementation principle of the monitoring system 110 of the respiratory function rehabilitation training device based on artificial intelligence may refer to the implementation principle of the monitoring method of the respiratory function rehabilitation training device based on artificial intelligence, which is not described herein. It should be understood that the division of the modules of the above apparatus is merely a division of a logic function, and may be fully or partially integrated into a physical entity or may be physically separated when actually implemented. And these modules may all be implemented in software in the form of calls by the processing element; or can be realized in hardware; the method can also be realized in a form of calling software by a processing element, and the method can be realized in a form of hardware by a part of modules.

For example, the modules above may be one or more integrated circuits configured to implement the methods above, such as: one or more specific integrated circuits (application specific integrated circuit, ASIC), or one or more microprocessors (digital signal processor, DSP), or one or more field programmable gate arrays (field programmable gate array, FPGA), or the like. For another example, when a module above is implemented in the form of a processing element scheduler code, the processing element may be a general purpose processor, such as a central processing unit (centralprocessing unit, CPU) or other processor that may invoke the program code. For another example, the modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

The embodiment of the invention provides a computer device 100, wherein the computer device 100 comprises a processor and a nonvolatile memory storing computer instructions, and when the computer instructions are executed by the processor, the computer device 100 executes the monitoring system 110 of the respiratory function rehabilitation training device based on artificial intelligence. As shown in fig. 3, fig. 3 is a block diagram of a computer device 100 according to an embodiment of the present invention. Computer device 100 includes an artificial intelligence based respiratory function rehabilitation trainer monitoring system 110, a memory 111, a processor 112, and a communication unit 113.

For data transmission or interaction, the memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly. For example, the elements may be electrically connected to each other via one or more communication buses or signal lines. The monitoring system 110 of the artificial intelligence based respiratory function rehabilitation trainer comprises at least one software function module which may be stored in the memory 111 in the form of software or firmware (firmware) or cured in the Operating System (OS) of the computer device 100. The processor 112 is configured to execute the monitoring system 110 of the artificial intelligence based respiratory function rehabilitation training device stored in the memory 111, for example, a software function module and a computer program included in the monitoring system 110 of the artificial intelligence based respiratory function rehabilitation training device.

The embodiment of the invention provides a readable storage medium, which comprises a computer program, wherein the computer program controls computer equipment where the readable storage medium is located to execute the monitoring method of the respiratory function rehabilitation training instrument based on artificial intelligence when running.

The foregoing description, for purpose of explanation, has been presented with reference to particular embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the disclosure and its practical application, to thereby enable others skilled in the art to best utilize the disclosure and various embodiments with various modifications as are suited to the particular use contemplated. The foregoing description, for purpose of explanation, has been presented with reference to particular embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the disclosure and its practical application, to thereby enable others skilled in the art to best utilize the disclosure and various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. The monitoring system of the respiratory function rehabilitation training device based on the artificial intelligence is characterized by comprising a server and the respiratory function rehabilitation training device in communication connection with the server;

the server is used for responding to the respiratory function rehabilitation monitoring request and establishing communication connection with the respiratory function rehabilitation training instrument; acquiring the physiological parameters of the user from the respiratory function rehabilitation training device; acquiring a respiratory physiological parameter to be evaluated from the physiological parameter of the user, and calling a first respiratory state identification model to process to obtain a respiratory state parameter after the respiratory physiological parameter is converted; according to the respiratory state parameters, a preset respiratory function rehabilitation training table is combined for evaluation, and a monitoring result of the respiratory function rehabilitation training instrument is obtained; determining a plurality of target users with continuous abnormal breathing state characteristics based on the monitoring result of the breathing function rehabilitation training device; determining target on-line pushing information matched with each target user according to the continuous abnormal breathing state characteristics of each target user, wherein the target on-line pushing information comprises medical instrument product pushing information and rehabilitation training information pushing information;

The server is further configured to:

acquiring the respiratory physiological parameter to be evaluated, wherein the respiratory physiological parameter comprises at least one respiratory physiological index;

obtaining a respiratory feature vector corresponding to the at least one respiratory physiological index in the respiratory physiological parameter through a first respiratory state recognition model, wherein the first respiratory state recognition model is trained according to a plurality of sample respiratory data pairs, the sample respiratory data pairs comprise sample respiratory state data and sample respiratory physiological data obtained by converting the sample respiratory state data through a second respiratory state recognition model, the second respiratory state recognition model is trained according to optimizing respiratory data on a model framework of a preset second respiratory state recognition model, one optimizing respiratory data pair comprises one optimizing sample respiratory state data and a plurality of optimizing sample respiratory physiological data, the optimizing sample respiratory physiological data is mapping data of the optimizing sample respiratory state data, the model framework of the preset second respiratory state recognition model is optimized according to a conversion result of the optimizing sample respiratory state data by the preset second respiratory state recognition model, and a second cost function between optimizing sample respiratory physiological data corresponding to the optimizing sample respiratory state data is optimized;

Predicting at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index according to the respiratory feature vector through the first respiratory state identification model, and converting each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index;

determining a predicted respiratory state index corresponding to each respiratory physiological index from the pre-stored respiratory state indexes according to the first confidence coefficient through the first respiratory state identification model;

and carrying out weighted average operation on each predicted respiratory state index through the first respiratory state identification model to obtain respiratory state parameters after the respiratory physiological parameters are converted.

2. The system of claim 1, wherein the server is further configured to:

the breathing characteristic vector is corresponding to a second breathing characteristic matrix through a matching relation between a first breathing characteristic matrix at a breathing physiological side and a second breathing characteristic matrix at a breathing state side in the first breathing state identification model, so that a reference breathing characteristic vector is obtained;

according to the respiratory feature vector of the respiratory physiological index set at the respiratory state side in the second respiratory feature matrix and the matching degree of the respiratory feature vector, determining at least two pre-stored respiratory state indexes corresponding to each respiratory physiological index in the respiratory physiological index set, and converting each respiratory physiological index into a first confidence coefficient of the corresponding pre-stored respiratory state index.

3. The system of claim 1, wherein the server is further configured to:

acquiring an optimized respiratory data pair;

acquiring a respiratory feature vector corresponding to each target index in the optimized sample respiratory state data through a preset second respiratory state identification model;

the breathing characteristic vector of the target index is corresponding to the first breathing characteristic matrix through a matching relation between a second breathing characteristic matrix at a breathing state side and a first breathing characteristic matrix at a breathing physiological side in the preset second breathing state identification model, so that a first reference breathing characteristic vector is obtained;

determining at least two pre-stored respiratory physiological index variables corresponding to each target index in the respiratory physiological index variables according to the respiratory feature vectors of the respiratory physiological index variables in the first respiratory feature matrix and the matching degree of the respiratory feature vectors with the first reference respiratory feature vectors, and converting each target index into a second confidence degree of the corresponding pre-stored respiratory physiological index variable;

Determining target respiratory physiological index variables corresponding to the target indexes from the prestored respiratory physiological index variables according to the second confidence coefficient;

performing weighted average operation on each target respiratory physiological index variable to obtain predicted sample respiratory physiological data after the conversion of the optimized sample respiratory state data;

calculating index difference cost between a prestored respiratory physiological index variable corresponding to each target index and a contrast index in the optimized sample respiratory state data according to the first homogeneity coefficient and the second confidence coefficient;

performing weighted average operation on index difference cost corresponding to each target index to obtain a second cost function between predicted sample respiratory physiological data of the optimized sample respiratory state data and corresponding optimized sample respiratory physiological data;

optimizing a model framework of a preset second respiratory state identification model according to the second cost function to obtain a second respiratory state identification model;

Converting the sample respiratory state data through a second respiratory state identification model to obtain converted sample respiratory physiological data, and forming a sample respiratory data pair by the converted sample respiratory physiological data and the corresponding sample respiratory state data;

determining the sample breath data pair and the initial breath data pair as a sample training set of the first breath state recognition model, wherein the initial sample breath physiological data and the converted sample breath physiological data are original sample breath physiological data, and the initial sample breath state data and the sample breath state data are original sample breath state data;

and converting the original sample respiratory physiological data through a first respiratory state identification model to obtain converted predicted sample respiratory state data, and performing conversion training from a respiratory physiological side to a respiratory state side on the first respiratory state identification model according to a first cost function between the predicted sample respiratory state data and the original sample respiratory state data corresponding to the original sample respiratory physiological data.

4. A system according to claim 3, wherein the server is further configured to:

and optimizing a model framework of a preset first respiratory state identification model according to the third price function to obtain the first respiratory state identification model.

5. The system of claim 4, wherein the server is further configured to:

acquiring a respiratory feature vector corresponding to each target index in the target sample respiratory physiological data through a preset first respiratory state identification model;

predicting at least two pre-stored respiratory state index variables corresponding to the target indexes according to the respiratory feature vectors corresponding to the target indexes, and converting the target indexes into third confidence coefficients of the corresponding pre-stored respiratory state index variables;

Determining target respiratory state index variables corresponding to the target indexes from the prestored respiratory state index variables according to the third confidence coefficient;

6. The system of claim 5, wherein the server is further configured to:

the breathing characteristic vector of the target index is corresponding to the second breathing characteristic matrix through a matching relation between a first breathing characteristic matrix of a breathing physiological side and a second breathing characteristic matrix of a breathing state side in the preset first breathing state identification model, so that a second reference breathing characteristic vector is obtained;

according to the respiratory feature vector of the respiratory state index variable in the second respiratory feature matrix and the matching degree of the respiratory feature vector with the second reference respiratory feature vector, determining at least two pre-stored respiratory state index variables corresponding to each target index in the respiratory state index variables, and converting each target index into a third confidence coefficient of the corresponding pre-stored respiratory state index variable.

7. The system of claim 6, wherein the server is further configured to:

for the same target sample respiratory physiological data, calculating a pre-stored respiratory state index variable corresponding to a target index of the same target sample respiratory physiological data, and a second homogeneity coefficient between contrast indexes corresponding to the target index in optimized sample respiratory state data corresponding to the target sample respiratory physiological data;

calculating index difference cost between a prestored respiratory state index variable corresponding to each target index in the target sample respiratory physiological data and a contrast index according to the second homogeneity coefficient and the third confidence coefficient;

and performing weighted average operation on the index difference cost corresponding to each target index to obtain a third price function between the predicted and optimized sample respiratory state data of the target sample respiratory physiological data and the corresponding optimized sample respiratory state data.

8. A system according to claim 3, wherein the server is further configured to:

determining target indexes corresponding to the sample respiratory physiological indexes from the undetermined sample respiratory physiological indexes according to the fourth confidence coefficient, and performing weighted average operation on the target indexes to obtain predicted sample respiratory state data;

calculating a first cost function between the predicted sample respiratory state data and original sample respiratory state data corresponding to the original sample respiratory physiological data;

9. The monitoring method of the respiratory function rehabilitation training device based on artificial intelligence is characterized by being executed by a server, and comprises the following steps:

acquiring physiological parameters of a user from the respiratory function rehabilitation training device, wherein the physiological parameters of the user are acquired by a sensor configured by the respiratory function rehabilitation training device;

Acquiring a respiratory physiological parameter to be evaluated from the physiological parameter of the user, and calling a first respiratory state identification model to process to obtain a respiratory state parameter after the respiratory physiological parameter is converted;

obtaining a monitoring result of the respiratory function rehabilitation training device;

determining a plurality of target users with continuous abnormal breathing state characteristics based on the monitoring result of the breathing function rehabilitation training device;

determining target on-line pushing information matched with each target user according to the continuous abnormal breathing state characteristics of each target user, wherein the target on-line pushing information comprises medical instrument product pushing information and rehabilitation training information pushing information;

the method further comprises the steps of: