Children health monitoring method and device based on smart bracelet

Abstract

The invention discloses a child health monitoring method and device based on a smart bracelet. The method comprises the following steps: S1, acquiring a training data set; s2, performing training processing on a preset children physiological monitoring model by using the training data set to obtain a trained children physiological monitoring model; s3, acquiring a child physiological parameter set by using the smart bracelet; and S4, processing the child physiological parameter set by using the trained child physiological monitoring model to obtain child health monitoring result information. The child health monitoring result information is used for representing the health degree information of the child. According to the invention, real-time and accurate monitoring of the health condition of the child is realized by using the bracelet.

Description

Technical Field

[0001] The present invention relates to the fields of communications and artificial intelligence, and in particular to a method. Background Art

[0002] A medical bracelet is a ring-shaped wearable item that carries user information and is distributed by the hospital during the user's hospitalization, so that the hospital can manage the user's treatment. However, the current medical bracelet has a relatively simple function and is only a label that stores information. It is difficult to effectively monitor and apply the user's physiological condition. Especially for child users, some of their organs are not fully developed, and their physiological indicators show variable characteristics. At the same time, children have poor expressive ability and there are inaccuracies in the expression of some physiological feelings, which leads to certain difficulties in their health monitoring.

[0003] How to use bracelets to monitor the health status of children, especially those hospitalized, in real time and accurately is an issue that needs to be addressed urgently. Summary of the invention

[0004] The present invention mainly solves the problem of how to use a wristband to conduct real-time and accurate monitoring of a child's health status. The present invention discloses a child health monitoring method and device based on a smart wristband.

[0005] In a first aspect of the embodiment of the present application, a method for monitoring children's health based on a smart bracelet is disclosed, comprising:

[0006] S1, obtain the training data set;

[0007] S2, using the training data set, training the preset child physiological monitoring model to obtain a trained child physiological monitoring model;

[0008] S3, using the smart bracelet to collect a set of children’s physiological parameters;

[0009] S4, using the trained child physiological monitoring model, processing the child physiological parameter set to obtain child health monitoring result information.

[0010] The training data set includes a set of child physiological parameters and corresponding label information; the label information is used to characterize the child health level of the training data in the set of child physiological parameters;

[0011] The child physiological parameter set includes physiological acquisition signals of four modes;

[0012] The physiological acquisition signals include electrocardiogram signals, heart rate signals, blood pressure signals and electrodermal signals.

[0013] The child physiological monitoring model includes: an input module, a feature extraction module, a feature fusion module, a feature processing module and a prediction module;

[0014] The input module is used to receive a set of physiological parameters of a child;

[0015] The feature extraction module is connected to the input module and the feature fusion module respectively, and is used to extract features from the child's physiological parameter set to obtain a feature signal set;

[0016] The feature fusion module is used to perform fusion processing on the feature signal set to obtain a fused feature signal set;

[0017] The feature processing module is connected to the feature fusion module and the prediction module respectively, and is used to perform time series correlation processing on the fused feature signal set to obtain a time series cross feature signal set;

[0018] The prediction module is used to perform prediction processing on the time series cross-feature signal set to obtain child health monitoring result information.

[0019] The feature extraction module includes four feature extractors corresponding to the signal modalities; each feature extractor is used to perform feature extraction processing on the physiological acquisition signal of the corresponding modality to obtain a corresponding feature signal;

[0020] The feature fusion module includes four multi-head attention sub-modules; each multi-head attention sub-module is used to process the physiological acquisition signal of the corresponding modality; the dimension of each multi-head attention sub-module matches the dimension of the physiological acquisition signal of the modality it processes; the input end of each multi-head attention sub-module is connected to the output end of the corresponding feature extractor;

[0021] The feature processing module includes four long short-term memory network submodules; each long short-term memory network submodule is used to process the physiological acquisition signal of the corresponding modality; the input end of each long short-term memory network submodule is connected to the output end of the corresponding multi-head attention submodule;

[0022] The prediction module includes a fully connected network and an output end; the N input ends of the fully connected network are respectively connected to the output ends of the four long short-term memory network sub-modules; the output end of the prediction module is used to output the child health monitoring result information.

[0023] The method of using the training data set to train the preset child physiological monitoring model to obtain a trained child physiological monitoring model includes:

[0024] Initialize the number of training iterations;

[0025] Inputting the training data in the training data set as input data into the child physiological monitoring model;

[0026] Using the child physiological monitoring model, the input data is processed to obtain a predicted value;

[0027] Performing difference calculation processing on the obtained predicted value and the label information corresponding to the input data to obtain a difference value;

[0028] Determine whether the difference value satisfies a convergence condition, and obtain a first determination result;

[0029] When the first judgment result is no, judging whether the training iteration number value is equal to the training number threshold, and obtaining a second judgment result;

[0030] When the second judgment result is no, determining that the model training state does not meet the training termination condition;

[0031] When the second judgment result is yes, determining that the model training state satisfies the training termination condition;

[0032] When the first judgment result is yes, determining that the model training state satisfies a training termination condition;

[0033] When the model training state does not meet the termination training condition, the parameter updating model is used to update the parameters of the feature fusion module and the feature processing module of the child physiological monitoring model, and the training iteration number value is increased by 1, triggering the execution of inputting the training data in the training data set as input data into the child physiological monitoring model;

[0034] When the model training state satisfies the training termination condition, the training process of the child physiological monitoring model is completed to obtain a trained child physiological monitoring model.

[0035] The parameter updating model is:

[0036]

[0037] θ←θ+v;

[0038] In the formula, is the difference value calculated for the i-th training data in the training data set, v is the parameter update value, θ is the parameter of the module to be updated, η is the initial parameter learning rate, α is the momentum angle parameter, 0≤α≤π / 4, It means to find the partial derivative with respect to the variable θ.

[0039] In a second aspect of an embodiment of the present invention, a child health monitoring device based on a smart bracelet is disclosed, which is implemented by using the child health monitoring method based on a smart bracelet, including:

[0040] Smart bracelet and health monitoring module;

[0041] The smart bracelet is used to collect a set of physiological parameters of the child and send the set of physiological parameters of the child to the health monitoring module;

[0042] The health monitoring module includes a child physiological monitoring model; the child physiological monitoring model is trained using a training data set and is used to process a set of child physiological parameters to obtain child health monitoring result information.

[0043] In a third aspect of an embodiment of the present invention, a child health monitoring device based on a smart bracelet is disclosed, the device comprising:

[0044] A memory storing executable program code;

[0045] a processor coupled to the memory;

[0046] The processor calls the executable program code stored in the memory to execute the child health monitoring method based on the smart bracelet.

[0047] In a fourth aspect of an embodiment of the present invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are called by a computer, they are used to execute the child health monitoring method based on a smart bracelet.

[0048] In a fifth aspect of an embodiment of the present invention, an information data processing terminal is disclosed, which is used to implement the child health monitoring method based on the smart bracelet.

[0049] The beneficial effects of the present invention are:

[0050] In view of the problem that children's physiological indicators present changeable characteristics and poor expression ability, the present invention uses a smart bracelet to realize real-time and accurate monitoring of multiple physiological indicators of children. By using a trained child physiological monitoring model, the child's physiological parameter set is processed to realize accurate monitoring of children's health.

[0051] The present invention specifically constructs a parameter updating model for the training process of the children's physiological monitoring model, realizes the accurate extraction of label difference features in training, and improves the efficiency and accuracy of model training. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart for implementing the method of the present invention. DETAILED DESCRIPTION

[0053] In order to better understand the content of the present invention, an embodiment is given here.

[0054] Figure 1 It is a flow chart for implementing the method of the present invention.

[0055] In view of the problem that children's physiological indicators present variable characteristics and poor expression ability, the present invention discloses a child health monitoring method and device based on a smart bracelet.

[0056] In a first aspect of the embodiment of the present application, a method for monitoring children's health based on a smart bracelet is disclosed, comprising:

[0057] S1, obtain the training data set;

[0058] S2, using the training data set, training the preset child physiological monitoring model to obtain a trained child physiological monitoring model;

[0059] S3, using the smart bracelet to collect a set of children’s physiological parameters;

[0060] S4, using the trained child physiological monitoring model, processing the child physiological parameter set to obtain child health monitoring result information.

[0061] The child health monitoring result information is used to represent the child's health status information.

[0062] The training data set includes a set of child physiological parameters and corresponding label information; the label information is used to characterize the child health level of the training data in the set of child physiological parameters;

[0063] The child physiological parameter set includes physiological acquisition signals of four modes;

[0064] The physiological acquisition signals include electrocardiogram signals, heart rate signals, blood pressure signals, and electrical skin signals.

[0065] The child physiological monitoring model includes: an input module, a feature extraction module, a feature fusion module, a feature processing module and a prediction module;

[0066] The input module is used to receive a set of physiological parameters of a child;

[0067] The feature extraction module is connected to the input module and the feature fusion module respectively, and is used to extract features from the child's physiological parameter set to obtain a feature signal set;

[0068] The feature fusion module is used to perform fusion processing on the feature signal set to obtain a fused feature signal set;

[0069] The feature processing module is connected to the feature fusion module and the prediction module respectively, and is used to perform time series correlation processing on the fused feature signal set to obtain a time series cross feature signal set;

[0070] The prediction module is used to perform prediction processing on the time series cross-feature signal set to obtain child health monitoring result information.

[0071] The feature extraction module includes four feature extractors corresponding to the signal modalities; each feature extractor is used to perform feature extraction processing on the physiological acquisition signal of the corresponding modality to obtain a corresponding feature signal;

[0072] The feature fusion module includes four multi-head attention sub-modules; each multi-head attention sub-module is used to process the physiological acquisition signal of the corresponding modality; the dimension of each multi-head attention sub-module matches the dimension of the physiological acquisition signal of the modality it processes; the input end of each multi-head attention sub-module is connected to the output end of the corresponding feature extractor;

[0073] The feature processing module includes four long short-term memory network submodules; each long short-term memory network submodule is used to process the physiological acquisition signal of the corresponding modality; the input end of each long short-term memory network submodule is connected to the output end of the corresponding multi-head attention submodule;

[0074] The prediction module includes a fully connected network and an output end; the N input ends of the fully connected network are respectively connected to the output ends of the N long short-term memory network sub-modules; the output end of the prediction module is used to output the child health monitoring result information.

[0075] The method of using the training data set to train the preset child physiological monitoring model to obtain a trained child physiological monitoring model includes:

[0076] Initialize the number of training iterations;

[0077] Inputting the training data in the training data set as input data into the child physiological monitoring model;

[0078] Using the child physiological monitoring model, the input data is processed to obtain a predicted value;

[0079] Performing difference calculation processing on the obtained predicted value and the label information corresponding to the input data to obtain a difference value;

[0080] Determine whether the difference value satisfies a convergence condition, and obtain a first determination result;

[0081] When the first judgment result is no, judging whether the training iteration number value is equal to the training number threshold, and obtaining a second judgment result;

[0082] When the second judgment result is no, determining that the model training state does not meet the training termination condition;

[0083] When the second judgment result is yes, determining that the model training state satisfies the training termination condition;

[0084] When the first judgment result is yes, determining that the model training state satisfies a training termination condition;

[0085] When the model training state does not meet the training termination condition, the parameter updating model is used to update the parameters of the feature fusion module and the feature processing module of the child physiological monitoring model, and the training iteration number value is increased by 1, triggering the execution of inputting the training data in the training data set as input data into the child physiological monitoring model;

[0086] When the model training state satisfies the training termination condition, the training process of the child physiological monitoring model is completed to obtain a trained child physiological monitoring model.

[0087] The parameter updating model is:

[0088]

[0089] θ←θ+v;

[0090] In the formula, is the difference value calculated for the i-th training data in the training data set, v is the parameter update value, θ is the parameter of the module to be updated, η is the initial parameter learning rate, α is the momentum angle parameter, 0≤α≤π / 4, It means to find partial derivatives with respect to variable θ;

[0091] The trained child physiological monitoring model is used to process the child physiological parameter set to obtain child health monitoring result information, including:

[0092] Using the trained child physiological monitoring model, the child physiological parameter set of the current frame is processed to obtain prediction result data of N frames subsequent to the current frame;

[0093] Using the trained child physiological monitoring model, the child physiological parameter set of the current frame and the previous N frames is processed to obtain N+1 frame enhancement information;

[0094] A circulant matrix is ​​constructed using the enhancement information of the current frame and its previous N frames and the next N frames; the i-th row of the circulant matrix is ​​the N+1 frame enhancement information obtained by the set of child physiological parameters of the next i-1 frame after the current frame and the next N frames before the next i-1 frame;

[0095] Perform data enhancement processing on the circulatory matrix to obtain the child health monitoring result information

[0096] The data enhancement processing includes:

[0097] Extract the sub-diagonal vector of the circulant matrix;

[0098] The sampling sequence of the normalized Gaussian distribution function is used to perform inner product operation with the sub-diagonal vector to obtain the child health monitoring result information of the current frame data.

[0099] Taking the current frame as the 10th frame as an example, the expression of the constructed circulant matrix is:

[0100]

[0101] The numbers in the matrix represent the prediction result data of the sequence frame.

[0102] The feature extraction module includes 4 feature extraction submodules for extracting physiological acquisition signals of corresponding modalities; for heart rate signals, the corresponding feature extraction submodule is implemented by using a Transformer network; for blood pressure signals, the corresponding feature extraction submodule is implemented by using a pre-trained VGG16net model; for electrocardiogram signals, a heart rate variability extraction model is used to implement it; the heart rate variability extraction model is used to calculate the heart rate variability (HRV) value based on the collected user heart rate, and the heart rate variability (HRV) value refers to the change in the difference between successive heartbeat cycles. For skin electrical signals, an autoencoder neural network or an autoencoder or a DAE network is used to implement it;

[0103] The fully connected network is used to implement the Softmax function.

[0104] The multi-head attention submodule is used to perform modal cross-feature calculation processing on the corresponding feature signal to obtain the modal cross-feature; the output of a multi-head attention submodule is 4 modal cross-feature vectors; each multi-head attention submodule includes n attention heads;

[0105] The modal cross-feature calculation process is calculated as follows:

[0106]

[0107]

[0108] Among them, F l-u is the modal cross feature vector of the feature signal of the current multi-head attention submodule and the u-th feature signal, l is the serial number of the current multi-head attention submodule, l = 1, 2, ..., N, u = 1, 2, ..., N, h i is the output of the i-th attention head of the current multi-head attention submodule, are the key, value, and weighted value of the query vector of the i-th attention head of the current multi-head attention submodule, i = 1, 2, ..., n, q, k, and v are the key, value, and query vector input by the current multi-head attention submodule, respectively; the values ​​of v and k are the l-th feature signal, and the value of q is the u-th feature signal; f is the first calculation function, and its expression is:

[0109]

[0110] Among them, d is a preset calculation constant, and s(q2, q1) is a calculation intermediate quantity.

[0111] The values ​​of the label information and the child health monitoring result information are 1, 2 or 3, and the corresponding health levels of the children are healthy, sub-healthy and abnormal.

[0112] The difference value satisfies the convergence condition, which means that the difference value is less than a preset convergence threshold; the difference value does not satisfy the convergence condition, which means that the difference value is not less than the preset convergence threshold.

[0113] The difference calculation process may be implemented using a loss function.

[0114] The loss function may adopt a cross entropy loss function.

[0115] In a second aspect of the embodiment of the present application, a child health monitoring device based on a smart bracelet is disclosed, which is implemented by using the child health monitoring method based on a smart bracelet, including:

[0116] Smart bracelet and health monitoring module;

[0117] The smart bracelet is used to collect a set of physiological parameters of the child and send the set of physiological parameters of the child to the health monitoring module;

[0118] The health monitoring module includes a child physiological monitoring model; the child physiological monitoring model is trained using a training data set and is used to process a set of child physiological parameters to obtain child health monitoring result information.

[0119] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for monitoring children's health based on a smart bracelet, characterized in that: include: S1, obtain the training data set; S2, using the training data set, training the preset child physiological monitoring model to obtain a trained child physiological monitoring model; S3, using the smart bracelet to collect a set of children’s physiological parameters; S4, using the trained child physiological monitoring model, processing the child physiological parameter set to obtain child health monitoring result information.

2. The method for monitoring children's health based on a smart bracelet as claimed in claim 1, characterized in that: The training data set includes a set of child physiological parameters and corresponding label information; the label information is used to characterize the child health level of the training data in the set of child physiological parameters; The child physiological parameter set includes physiological acquisition signals of four modes; The physiological acquisition signals include electrocardiogram signals, heart rate signals, blood pressure signals and electrodermal signals.

3. The method for monitoring children's health based on a smart bracelet as claimed in claim 2, characterized in that: The child physiological monitoring model includes: an input module, a feature extraction module, a feature fusion module, a feature processing module and a prediction module; The input module is used to receive a set of physiological parameters of a child; The feature extraction module is connected to the input module and the feature fusion module respectively, and is used to extract features from the child's physiological parameter set to obtain a feature signal set; The feature fusion module is used to perform fusion processing on the feature signal set to obtain a fused feature signal set; The feature processing module is connected to the feature fusion module and the prediction module respectively, and is used to perform time series correlation processing on the fused feature signal set to obtain a time series cross feature signal set; The prediction module is used to perform prediction processing on the time series cross-feature signal set to obtain child health monitoring result information.

4. The method for monitoring children's health based on a smart bracelet as claimed in claim 3, characterized in that: The feature extraction module includes four feature extractors corresponding to the signal modalities; each feature extractor is used to perform feature extraction processing on the physiological acquisition signal of the corresponding modality to obtain a corresponding feature signal; The feature fusion module includes 4 multi-head attention sub-modules; Each multi-head attention submodule is used to process the physiological acquisition signal of the corresponding modality; the dimension of each multi-head attention submodule matches the dimension of the physiological acquisition signal of the modality it processes; the input end of each multi-head attention submodule is connected to the output end of the corresponding feature extractor; The feature processing module includes four long short-term memory network submodules; each long short-term memory network submodule is used to process the physiological acquisition signal of the corresponding modality; the input end of each long short-term memory network submodule is connected to the output end of the corresponding multi-head attention submodule; The prediction module includes a fully connected network and an output end; the N input ends of the fully connected network are respectively connected to the output ends of the four long short-term memory network sub-modules; the output end of the prediction module is used to output the child health monitoring result information.

5. The method for monitoring children's health based on a smart bracelet as claimed in claim 3, characterized in that: The method of using the training data set to train the preset child physiological monitoring model to obtain a trained child physiological monitoring model includes: Initialize the number of training iterations; Inputting the training data in the training data set as input data into the child physiological monitoring model; Using the child physiological monitoring model, the input data is processed to obtain a predicted value; Performing difference calculation processing on the obtained predicted value and the label information corresponding to the input data to obtain a difference value; Determine whether the difference value satisfies a convergence condition, and obtain a first determination result; When the first judgment result is no, judging whether the training iteration number value is equal to the training number threshold, and obtaining a second judgment result; When the second judgment result is no, determining that the model training state does not meet the training termination condition; When the second judgment result is yes, determining that the model training state satisfies the training termination condition; When the first judgment result is yes, determining that the model training state satisfies a training termination condition; When the model training state does not meet the training termination condition, the parameter updating model is used to update the parameters of the feature fusion module and the feature processing module of the child physiological monitoring model, and the training iteration number value is increased by 1, triggering the execution of inputting the training data in the training data set as input data into the child physiological monitoring model; When the model training state satisfies the training termination condition, the training process of the child physiological monitoring model is completed to obtain a trained child physiological monitoring model.

6. The method for monitoring children's health based on a smart bracelet as claimed in claim 5, characterized in that: The parameter updating model is: θ←θ+v; In the formula, is the difference value calculated for the i-th training data in the training data set, v is the parameter update value, θ is the parameter of the module to be updated, η is the initial parameter learning rate, α is the momentum angle parameter, 0≤α≤π / 4, It means to find the partial derivative with respect to the variable θ.

7. A child health monitoring device based on a smart bracelet, implemented by the child health monitoring method based on a smart bracelet according to any one of claims 1 to 6, comprising: Smart bracelet and health monitoring module; The smart bracelet is used to collect a set of physiological parameters of the child and send the set of physiological parameters of the child to the health monitoring module; The health monitoring module includes a child physiological monitoring model; the child physiological monitoring model is trained using a training data set and is used to process a set of child physiological parameters to obtain child health monitoring result information.

8. A child health monitoring device based on a smart bracelet, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the child health monitoring method based on the smart bracelet as described in any one of claims 1 to 6.

9. A computer storable medium, characterized in that: The computer storable medium stores computer instructions, and when the computer instructions are called by a computer, they are used to execute the child health monitoring method based on a smart bracelet as described in any one of claims 1 to 6.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the child health monitoring method based on the smart bracelet as described in any one of claims 1 to 6.

Images