CN111329491A - Blood glucose prediction method and device, electronic equipment and storage medium - Google Patents
Blood glucose prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111329491A CN111329491A CN202010125312.0A CN202010125312A CN111329491A CN 111329491 A CN111329491 A CN 111329491A CN 202010125312 A CN202010125312 A CN 202010125312A CN 111329491 A CN111329491 A CN 111329491A
- Authority
- CN
- China
- Prior art keywords
- blood glucose
- data
- feature
- prediction
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
- A61B5/14532—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue for measuring glucose, e.g. by tissue impedance measurement
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Emergency Medicine (AREA)
- Optics & Photonics (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
One or more embodiments of the present specification provide a blood glucose prediction method, apparatus, electronic device, and storage medium; the method comprises the following steps: acquiring at least one characteristic data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data; generating at least one feature vector from at least one of the feature data sets; and obtaining a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model. One or more embodiments of the present disclosure predict a blood glucose value of a user by using blood glucose data, food intake data, and insulin administration data based on a sequence model, instead of predicting by using only blood glucose data, the related conditions of the user are sufficiently and accurately reflected by the blood glucose data, food intake data, and insulin administration data, thereby effectively improving the accuracy of blood glucose prediction.
Description
Technical Field
One or more embodiments of the present disclosure relate to the field of artificial intelligence technologies, and in particular, to a blood glucose prediction method, apparatus, electronic device, and storage medium.
Background
With the increase of living standard, more and more people suffer from diabetes. Research data show that the prevalence of diabetes in our country has rapidly increased from less than 1% to over 10% over the past thirty years. Can accurately predict the blood sugar level in time and effectively control the blood sugar fluctuation, and is one of the ways of treating diabetes.
With the rapid development and wide application of artificial intelligence technology, in the related art, a technical scheme for predicting blood glucose by artificial intelligence technology has appeared. However, the solutions in the related technologies generally have the problems that the related conditions of the user cannot be accurately reflected, and the accuracy is low.
Disclosure of Invention
In view of the above, an object of one or more embodiments of the present disclosure is to provide a blood glucose prediction method, apparatus, electronic device and storage medium.
In view of the above, one or more embodiments of the present specification provide a blood glucose prediction method, including:
acquiring at least one characteristic data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data;
generating at least one feature vector from at least one of the feature data sets;
and obtaining a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model.
Based on the same inventive concept, one or more embodiments of the present specification further provide a blood glucose prediction device, including:
an acquisition module configured to acquire at least one feature data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data;
a generation module configured to generate at least one feature vector from at least one of the feature data sets;
and the prediction module is configured to obtain a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model.
Based on the same inventive concept, one or more embodiments of the present specification further provide an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the method as described in any one of the above items when executing the program.
Based on the same inventive concept, one or more embodiments of the present specification also provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method as described in any one of the above.
As can be seen from the foregoing description, the blood glucose prediction method, apparatus, electronic device and storage medium provided in one or more embodiments of the present disclosure use blood glucose data, feeding data and insulin administration data to predict a blood glucose value of a user based on a sequence model, instead of using only blood glucose data to predict, the blood glucose data, feeding data and insulin administration data are used to sufficiently and accurately reflect the relevant conditions of the user, so as to effectively improve the accuracy of blood glucose prediction.
Drawings
In order to more clearly illustrate one or more embodiments or prior art solutions of the present specification, the drawings that are needed in the description of the embodiments or prior art will be briefly described below, and it is obvious that the drawings in the following description are only one or more embodiments of the present specification, and that other drawings may be obtained by those skilled in the art without inventive effort from these drawings.
FIG. 1 is a flow diagram of a method for blood glucose prediction according to one or more embodiments of the present disclosure;
FIG. 2 is a schematic diagram illustrating the operation of a blood glucose prediction sequence model according to one or more embodiments of the present disclosure;
FIG. 3 is a schematic diagram of a blood glucose prediction device according to one or more embodiments of the present disclosure;
fig. 4 is a schematic structural diagram of an electronic device according to one or more embodiments of the present disclosure.
Detailed Description
For the purpose of promoting a better understanding of the objects, aspects and advantages of the present disclosure, reference is made to the following detailed description taken in conjunction with the accompanying drawings.
It is to be noted that unless otherwise defined, technical or scientific terms used in one or more embodiments of the present specification should have the ordinary meaning as understood by those of ordinary skill in the art to which this disclosure belongs.
As described in the background section, the technical solution of predicting blood glucose by artificial intelligence in the related art generally has the problem of low accuracy. Technical schemes for predicting blood sugar through an artificial intelligence technology in the related art can be roughly divided into two categories: blood glucose prediction based on physiological models and blood glucose prediction based on data. The physiological model-based prediction predicts the blood glucose level by considering factors influencing the blood glucose level in a human body, however, because the physiological mechanism of the human body is complex, the factors influencing the blood glucose level are many and are mutually related, an accurate prediction model is difficult to establish; on the other hand, the collection of physiological signals of a human body wastes time and labor, so that the blood sugar prediction based on a physiological model is inconvenient for users to use. Therefore, data-based blood glucose prediction is more adopted in the related art.
However, the accuracy of data-based blood glucose prediction in the related art is still low. In implementing the present disclosure, the inventors found that the reason why the accuracy of data-based blood glucose prediction in the related art is low is that: the data-based blood glucose prediction is based on the historical blood glucose data of the user and ignores the physiological mechanism of the human body to predict the blood glucose level. This method is convenient for the user to use, however, since only the historical blood glucose data is utilized, there is a certain loss of accuracy. For example, assuming that the user's historical blood glucose data is identical, the blood glucose value after one hour should be different in the case where the user has used a large amount of sugar-containing food and has not consumed food. Furthermore, the use of insulin medication by the user also has a strong influence on the blood glucose level. In summary, one or more embodiments of the present disclosure provide a blood glucose prediction technical solution, which predicts a blood glucose value of a user by using blood glucose data, food intake data, and insulin administration data based on a sequence model, and does not use only the blood glucose data for prediction, but sufficiently and accurately reflect a relevant situation of the user by using the blood glucose data, the food intake data, and the insulin administration data, thereby effectively improving an accuracy rate of blood glucose prediction.
The technical solutions of one or more embodiments of the present specification are described in detail below with reference to specific embodiments.
Referring to fig. 1, a blood glucose prediction method according to an embodiment of the present specification includes the following steps:
s101, acquiring at least one characteristic data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data.
In this step, a feature data set of the user is first obtained, and the blood glucose of the user is subsequently predicted based on data included in the feature data set. A characteristic data set at least comprises three data of blood sugar data, food intake data and medicine taking data.
The blood glucose data is a blood glucose value of the user. In particular, the blood glucose data may include at least one of current blood glucose data and historical blood glucose data. The current blood glucose data refers to a blood glucose value obtained by currently performing blood glucose detection on the user. The historical blood glucose data refers to blood glucose values obtained by the user through blood glucose detection in the history of the user. According to specific implementation requirements, only the current blood glucose data or only the historical blood glucose data may be used, or both the current blood glucose data and the historical blood glucose data may be used. In this embodiment, a blood glucose value obtaining way of the user in the blood glucose data is not specifically limited, and may be obtained and uploaded by the user through self detection, or obtained from an external data source (such as a physiological index monitoring system, software, and the like).
The eating data is data describing whether or not the user has eaten food and what kind of food the user has eaten. The acquisition mode can be through uploading by the user, or other data sources (such as a user's eating plan established by a hospital).
The medication data is data describing whether a drug related to blood glucose control (generally, an insulin control drug, or directly injected insulin) is used or not, and what kind of drug is used. Similarly, the information may be obtained by uploading by the user, or may be obtained from other data sources (e.g., a user's medication plan made by a hospital).
In this embodiment, the number of feature data sets is at least one. Since the sequence model is subsequently used for prediction in this embodiment, the feature data set constitutes a sequence data for inputting the sequence model for prediction. When the number of the feature data sets is more than one, the plurality of feature data sets form sequence data arranged in time sequence according to the time sequence of the acquisition time.
Each characteristic data set corresponds to one collection time, and the blood sugar data, the food intake data and the medicine taking data which are included in the characteristic data sets correspond to one collection time, namely the blood sugar value, food intake and medicine taking of the user at the collection time. The food consumption data and the medicine consumption data do not strictly correspond to one collection time, so that the adoption of the food consumption data and the medicine consumption data can be expanded to a time period including the collection time, such as whether food is consumed and medicine is consumed in the time period from the previous collection time to the current collection time.
In addition, for the acquisition moments corresponding to the plurality of feature data sets, the acquisition moments may be equidistant or non-equidistant. In this embodiment, a case where a plurality of acquisition times are equally spaced in time will be described.
Step S102, generating at least one feature vector according to at least one feature data set.
In this step, based on the feature data set obtained in the previous step, a feature vector corresponding to the user is generated according to the data included in the feature data set, and the feature vector is used as an input of the sequence model in the subsequent prediction step.
Specifically, for blood glucose data, food intake data and medication data included in the feature data set, corresponding feature values are respectively determined, that is, a blood glucose feature value corresponding to the blood glucose data, a food intake feature value corresponding to the blood glucose data and a medication feature value corresponding to the medication data.
The blood sugar of the user often shows continuity and correlation in time, so that the blood sugar data can further comprise current blood sugar data and historical blood sugar data to further meet the actual situation. The current blood glucose data is the blood glucose value corresponding to the user at the current collection time. Historical blood glucose data, i.e., blood glucose values corresponding to the user at other historical times. The other time may be any time before the current acquisition time, or may be the acquisition time corresponding to the other feature data set acquired in the foregoing step. In this embodiment, the present blood glucose data and the historical blood glucose data are used together as an example for explanation. Obviously, only one of the current blood glucose data and the historical blood glucose data may be used in other embodiments, depending on implementation needs.
And for the blood sugar data, the corresponding blood sugar characteristic value is the blood sugar value. Further, the blood glucose data includes current blood glucose data and historical blood glucose data. Correspondingly, the current blood sugar data corresponds to a current blood sugar characteristic value, the historical blood sugar data corresponds to a historical blood sugar characteristic value, and the current blood sugar characteristic value and the historical blood sugar characteristic value are blood sugar values of the user at corresponding moments. As an example, two historical blood glucose data are taken, and the corresponding historical blood glucose characteristic value is represented as Glu according to time sequence1、Glu2(ii) a The current blood glucose feature value corresponding to the current blood glucose data represents Glu3(ii) a According to the time sequence arrangement, the blood sugar characteristic value is expressed as: glu (glutamic acid)1、Glu2、Glu3。
For meal data, its corresponding meal feature values are represented by a vector. The element position in the vector represents the kind of food, and the value of the element represents whether the food is eaten or not. The condition of the diabetic for eating the staple food is closely related to the blood sugar value, so the type of the food can be selected from the staple food; in this embodiment, a staple food is taken as an example for explanation; obviously, the food may be other foods such as vegetables and fruits according to the implementation requirement. As an example, is shown inThe vector of the food characteristic values may be expressed as (food)1,…,foodN) That is, N kinds of staple foods are available. Wherein, the value of each element is 1 or 0; a value of 1 indicates that the staple food was consumed and a value of 0 indicates that the staple food was not consumed.
For the administration data, its corresponding administration characteristic values are represented by vectors, similar in form and manner to the aforementioned eating characteristic values. As one example, the vector representing the medication characteristic value may be represented as (drug)1,…,drugN) That is, N kinds of drugs are present. Wherein, the value of each element is 1 or 0; a value of 1 indicates that the drug is used and a value of 0 indicates that the drug is not used.
When generating the feature vector, the blood glucose feature value, the eating feature value, and the medication feature value are combined to generate the feature vector, that is, the feature vector may be represented as (Glu)1,Glu2,Glu3,food1,…,foodN,drug1,…,drugN)。
A more specific example is given below, and specifically includes three feature data sets, i.e., feature data set 1, feature data set 2, and feature data set 3, which respectively correspond to acquisition time t1、t2、t3；t1、t2、t3The time intervals therebetween were all 1 hour. The blood sugar data comprises the blood sugar value at the current collection time and the blood sugar values at the previous two collection times; the eating data comprises the eating conditions of two staple foods of rice and steamed buns; the dosing data includes the use of insulin injection, a drug.
For the feature data set 1, only the acquisition time t1When the blood sugar level of (1) is 4, the blood sugar characteristic value Glu3A value of 4; blood glucose characteristic value Glu without blood glucose values of the previous two collection times1、Glu2The values of (A) are all 0. At the acquisition time t1The user does not eat, nor does the user use medications. The feature vector x is generated from the feature data set 11Is (0,0,4,0,0, 0).
For the feature data set 2, the acquisition time t2When the blood sugar level of (1) is 6, the blood sugar characteristic value Glu3The value is 6; previous acquisition time t1When the blood sugar level of (1) is 4, the blood sugar characteristic value Glu2A value of 4; based on the above data, characteristic value Glu of blood glucose1The value is 0. At the acquisition time t2The user eats the rice and the steamed bread and injects the insulin. The feature vector x is generated from the feature data set 22Is (0,4,6,1,1, 1).
For the feature data set 2, the acquisition time t3When the blood sugar level of (1) is 3, the blood sugar characteristic value Glu3The value is 3; according to the above data, Glu1、Glu2The values of (A) are respectively 4 and 6. At the acquisition time t3The user does not eat, nor does the user use medications. The feature vector x is generated from the feature data set 33Is (4,6,3,0,0, 0).
In time series, the feature vector x1、x2、x3I.e., a set of sequence data is constructed for subsequent prediction processes by the sequence model.
And S103, obtaining a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model.
In this step, x is obtained based on the above-mentioned step1、x2、x3The formed sequence data is used for predicting the blood sugar value of the user through a blood sugar prediction sequence model. The sequence model is a Recurrent Neural Network (RNN) in which sequence data is input, recursion is performed in the direction of evolution of the sequence, and all nodes are connected in a chain. Specifically, when data in the sequence data is input into the sequence model, the sequence model outputs a memory vector, which is a weight matrix of an activation function in the neural network responsible for mapping inputs of neurons to outputs, and which can reflect the effect on the result of each data preceding the current data in the sequence data. When the next data of the sequence data is input into the sequence model, the memory vector output by the previous sequence model is also input together. Wherein, the commonly used activation function can be selectedSigmoid function, Tanh function, ReLU function, etc.
In this embodiment, the blood glucose prediction sequence model is obtained by using a large number of feature data sets of different users obtained by the similar obtaining manner and construction method in the foregoing steps as a training set, and training an initial sequence model by a predetermined machine learning algorithm based on the training set. In this embodiment, the used machine learning algorithm is a Back-Propagation Through Time (BPTT) algorithm based on Time.
In this embodiment, the feature vectors obtained in the foregoing steps are sequentially input into the blood glucose prediction sequence model according to the time sequence of the acquisition time. Wherein the output of the blood glucose prediction sequence model comprises a memory vector; the input of the blood sugar prediction sequence model comprises a memory vector output by the blood sugar prediction sequence model according to a previous feature vector on the time sequence.
For the first feature vector in the time sequence, because there is no memory vector output in the preamble, an initial memory vector is set, and values of elements included in the initial memory vector may be random values or 0 values. When prediction is performed, the initial memory vector is input into the blood glucose prediction sequence model together with the time-series first feature vector.
And the blood sugar prediction sequence model obtains the blood sugar prediction result of the user by adopting a preset prediction algorithm according to the memory vector output by the last characteristic vector on the time sequence. The predetermined prediction algorithm may be any feasible prediction algorithm, such as Softmax, among others.
Based on the example of the feature vector given in the foregoing step, the prediction process in the present embodiment can be referred to as shown in fig. 2.
In particular, the feature vector x1Characteristic vector x2Characteristic vector x3Which are sequentially input into the blood glucose prediction sequence model in time series. For the feature vector x1It is associated with an initial memory vector h0Inputting a blood glucose prediction sequence togetherModel, blood glucose prediction sequence model output memory vector h1(ii) a Feature vector x2And a memory vector h1Inputting the blood sugar prediction sequence model together, outputting a memory vector h by the blood sugar prediction sequence model2(ii) a Feature vector x3And a memory vector h2Inputting the blood sugar prediction sequence model together, outputting a memory vector h by the blood sugar prediction sequence model3. Finally, according to the memory vector h3And obtaining a blood sugar prediction result of the user through a predetermined prediction algorithm.
Wherein the expression of the memory vector is: h isi＝σ(W×xi+U×hi-1+b)。
When prediction is performed, the following formula can be used for calculation: y-vTh3。
In the above formula, the value range of i is 1-3; σ () is Sigmoid function; w, U, b and v are weight parameters and are determined in the training process of the blood sugar prediction sequence model.
For the training process of the blood glucose prediction sequence model, for example, given a training set D { (x)1,1,x1,2,x1,3,y1),…,(xn,1,xn,2,xn,3,yn) And (6) minimizing a loss function through a time-based back propagation algorithm, so that weighting parameters such as W, U, b and v can be learned. The expression of the loss function is:
in the above formula, the value range of j is 1-n; y' is the prediction result in the training process.
In the present embodiment, the blood glucose prediction result of the user, that is, the predicted blood glucose level of the user can be obtained through the prediction process described above.
As can be seen from the foregoing embodiments, the blood glucose prediction method of this embodiment predicts the blood glucose value of the user by using the blood glucose data, the eating data, and the insulin administration data based on the sequence model, and does not predict only by using the blood glucose data, but sufficiently and accurately reflects the relevant conditions of the user by using the blood glucose data, the eating data, and the insulin administration data, thereby effectively improving the accuracy of blood glucose prediction.
As an optional embodiment, in the blood glucose prediction method of the foregoing embodiment, after obtaining the blood glucose prediction result of the user, the method further includes the following steps:
generating a health suggestion report of the user according to the blood sugar prediction result;
performing at least one of the following operations on the health advice report:
displaying the health advice report;
sending the health suggestion report to a preset server;
and sending the health suggestion report to a preset terminal device.
In this embodiment, the blood glucose prediction result includes a predicted blood glucose value of the user, and a health advice report of the user can be further generated based on the blood glucose value. The health advice report may be based on the blood glucose prediction result to generate a recommendation of eating, medication, or other life activities for the user, and further generate a health advice report in the form of text, voice, video, or various multimedia. The process of generating the suggestions according to the blood sugar prediction result can be obtained in a table look-up mode through the corresponding relation between the preset blood sugar value and the suggestion content, can also be obtained through algorithm calculation, and can also be obtained through a machine learning model in an artificial intelligence mode.
In this embodiment, after the health advice report is generated, the health advice report is further pushed, and a specific pushing manner for the health advice report can be flexibly selected.
For example, for an application scenario in which the method of the present embodiment is executed on a single device, the health advice report may be directly output in a display manner on a display component (display screen, projection component, etc.) of the current device, so that the user can directly see the health advice report from the display component.
For another example, for an application scenario in which the method of this embodiment is executed on a system composed of multiple devices, the health advice report may be pushed to other preset devices as recipients in the system through any data communication manner (wired connection, NFC, bluetooth, wifi, cellular mobile network, etc.). Optionally, the preset device may be a preset server, and the server is generally arranged at a cloud end and serves as a data processing and storage center, which can store and distribute the health advice report; the recipient of the distribution is a terminal device, and the holders or operators of the terminal devices may be users themselves, relatives, doctors, and the like.
For another example, for an application scenario in which the method of this embodiment is executed on a system composed of multiple devices, the health advice report may be directly sent to a preset terminal device through any data communication manner, and the terminal device may be one or more of the foregoing paragraphs.
The blood sugar prediction method further comprises the step of processing the blood sugar prediction result, generating a health advice report for the user and pushing the health advice report, so that the method is more convenient to apply.
It is to be appreciated that the method of one or more embodiments of the specification can be performed by any apparatus, device, platform, cluster of devices having computing and processing capabilities.
It should be noted that the method of one or more embodiments of the present disclosure may be performed by a single device, such as a computer or server. The method of the embodiment can also be applied to a distributed scene and completed by the mutual cooperation of a plurality of devices. In such a distributed scenario, one of the devices may perform only one or more steps of the method of one or more embodiments of the present disclosure, and the devices may interact with each other to complete the method.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Based on the same inventive concept, one or more embodiments of the present specification further provide a blood glucose prediction device. Referring to fig. 3, the blood glucose prediction apparatus includes:
an obtaining module 301 configured to obtain at least one feature data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data;
a generating module 302 configured to generate at least one feature vector from at least one of the feature data sets;
and the prediction module 303 is configured to obtain a blood glucose prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood glucose prediction sequence model.
As an optional embodiment, the generating module 302 is specifically configured to, for each feature data set, determine a blood glucose feature value corresponding to the blood glucose data, an eating feature value corresponding to the eating data, and a medication feature value corresponding to the medication data; and generating the characteristic vector according to the blood sugar characteristic value, the eating characteristic value and the medication characteristic value.
As an optional embodiment, the blood glucose data includes: at least one of current blood glucose data and historical blood glucose data; the generation module is specifically configured to determine a current blood glucose feature value corresponding to the current blood glucose data; and/or determining a historical blood glucose characteristic value corresponding to the historical blood glucose data.
As an optional embodiment, the prediction module 303 is specifically configured to sequentially input at least one of the feature vectors into the blood glucose prediction sequence model according to the time sequence of the acquisition time; wherein the output of the blood glucose prediction sequence model comprises a memory vector; the input of the blood sugar prediction sequence model comprises a memory vector output by the blood sugar prediction sequence model according to a previous feature vector on the time sequence; and obtaining a blood sugar prediction result of the user by adopting a preset prediction algorithm based on the memory vector output by the blood sugar prediction sequence model according to the last feature vector on the time sequence.
As an alternative embodiment, the prediction module 303 is specifically configured to obtain an initial memory vector for the first feature vector in the time sequence; inputting the initial memory vector and the first feature vector in the time sequence into the blood sugar prediction sequence model; and each element in the initial memory vector takes a random value or takes a zero value.
As an optional embodiment, the blood glucose prediction apparatus further includes: a push module configured to generate a health advice report for the user based on the blood glucose prediction result; performing at least one of the following operations on the health advice report: displaying the health advice report; sending the health suggestion report to a preset server; and sending the health suggestion report to a preset terminal device.
For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. Of course, the functionality of the modules may be implemented in the same one or more software and/or hardware implementations in implementing one or more embodiments of the present description.
The apparatus of the foregoing embodiment is used to implement the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which are not described herein again.
One or more embodiments of the present specification further provide an electronic device based on the same inventive concept. The electronic device comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the blood glucose prediction method according to any one of the above embodiments.
Fig. 4 is a schematic diagram illustrating a more specific hardware structure of an electronic device according to this embodiment, where the electronic device may include: a processor 1010, a memory 1020, an input/output interface 1030, a communication interface 1040, and a bus 1050. Wherein the processor 1010, memory 1020, input/output interface 1030, and communication interface 1040 are communicatively coupled to each other within the device via bus 1050.
The processor 1010 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits, and is configured to execute related programs to implement the technical solutions provided in the embodiments of the present disclosure.
The Memory 1020 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random access Memory), a static storage device, a dynamic storage device, or the like. The memory 1020 may store an operating system and other application programs, and when the technical solution provided by the embodiments of the present specification is implemented by software or firmware, the relevant program codes are stored in the memory 1020 and called to be executed by the processor 1010.
The input/output interface 1030 is used for connecting an input/output module to input and output information. The i/o module may be configured as a component in a device (not shown) or may be external to the device to provide a corresponding function. The input devices may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output devices may include a display, a speaker, a vibrator, an indicator light, etc.
The communication interface 1040 is used for connecting a communication module (not shown in the drawings) to implement communication interaction between the present apparatus and other apparatuses. The communication module can realize communication in a wired mode (such as USB, network cable and the like) and also can realize communication in a wireless mode (such as mobile network, WIFI, Bluetooth and the like).
It should be noted that although the above-mentioned device only shows the processor 1010, the memory 1020, the input/output interface 1030, the communication interface 1040 and the bus 1050, in a specific implementation, the device may also include other components necessary for normal operation. In addition, those skilled in the art will appreciate that the above-described apparatus may also include only those components necessary to implement the embodiments of the present description, and not necessarily all of the components shown in the figures.
Based on the same inventive concept, one or more embodiments of the present specification also provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the blood glucose prediction method according to any one of the embodiments described above.
Computer-readable media of the present embodiments, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, is limited to these examples; within the spirit of the present disclosure, features from the above embodiments or from different embodiments may also be combined, steps may be implemented in any order, and there are many other variations of different aspects of one or more embodiments of the present description as described above, which are not provided in detail for the sake of brevity.
It is intended that the one or more embodiments of the present specification embrace all such alternatives, modifications and variations as fall within the broad scope of the appended claims. Therefore, any omissions, modifications, substitutions, improvements, and the like that may be made without departing from the spirit and principles of one or more embodiments of the present disclosure are intended to be included within the scope of the present disclosure.
Claims (14)
1. A method of predicting blood glucose, comprising:
acquiring at least one characteristic data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data;
generating at least one feature vector from at least one of the feature data sets;
and obtaining a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model.
2. The method of claim 1, wherein generating at least one feature vector based on at least one of the feature data sets comprises:
for each feature data set, determining a blood glucose feature value corresponding to the blood glucose data, an eating feature value corresponding to the eating data and a medication feature value corresponding to the medication data;
and generating the characteristic vector according to the blood sugar characteristic value, the eating characteristic value and the medication characteristic value.
3. The method of claim 2, wherein the blood glucose data comprises: at least one of current blood glucose data and historical blood glucose data;
the determining the blood glucose characteristic value corresponding to the blood glucose data specifically includes:
determining a current blood glucose characteristic value corresponding to the current blood glucose data; and/or the presence of a gas in the gas,
and determining a historical blood glucose characteristic value corresponding to the historical blood glucose data.
4. The method of claim 1, wherein obtaining the blood glucose prediction result of the user according to the time sequence of the collection time, the feature vector and a pre-trained blood glucose prediction sequence model comprises:
sequentially inputting at least one feature vector into the blood glucose prediction sequence model according to the time sequence of the acquisition time; wherein the output of the blood glucose prediction sequence model comprises a memory vector; the input of the blood sugar prediction sequence model comprises a memory vector output by the blood sugar prediction sequence model according to a previous feature vector on the time sequence;
and obtaining a blood sugar prediction result of the user by adopting a preset prediction algorithm based on the memory vector output by the blood sugar prediction sequence model according to the last feature vector on the time sequence.
5. The method according to claim 4, wherein the sequentially inputting at least one of the feature vectors into the blood glucose prediction sequence model according to the time sequence of the collection time specifically comprises:
obtaining an initial memory vector for the first feature vector in the time sequence; inputting the initial memory vector and the first feature vector in the time sequence into the blood sugar prediction sequence model; and each element in the initial memory vector takes a random value or takes a zero value.
6. The method of predicting blood glucose as set forth in claim 1, further comprising:
generating a health suggestion report of the user according to the blood sugar prediction result;
performing at least one of the following operations on the health advice report:
displaying the health advice report;
sending the health suggestion report to a preset server;
and sending the health suggestion report to a preset terminal device.
7. A blood glucose prediction device, comprising:
an acquisition module configured to acquire at least one feature data set of a user; wherein each of the feature data sets corresponds to an acquisition time; the feature data set includes: blood glucose data, food intake data, and medication data;
a generation module configured to generate at least one feature vector from at least one of the feature data sets;
and the prediction module is configured to obtain a blood sugar prediction result of the user according to the time sequence of the acquisition time, the feature vector and a pre-trained blood sugar prediction sequence model.
8. The apparatus according to claim 7, wherein the generating module is specifically configured to determine, for each of the feature data sets, a blood glucose feature value corresponding to the blood glucose data, an eating feature value corresponding to the eating data, and a medication feature value corresponding to the medication data; and generating the characteristic vector according to the blood sugar characteristic value, the eating characteristic value and the medication characteristic value.
9. The apparatus of claim 8, wherein the blood glucose data comprises: at least one of current blood glucose data and historical blood glucose data;
the generation module is specifically configured to determine a current blood glucose feature value corresponding to the current blood glucose data; and/or determining a historical blood glucose characteristic value corresponding to the historical blood glucose data.
10. The apparatus of claim 7, wherein the prediction module is specifically configured to sequentially input at least one of the feature vectors into the blood glucose prediction sequence model according to the time sequence of the acquisition time; wherein the output of the blood glucose prediction sequence model comprises a memory vector; the input of the blood sugar prediction sequence model comprises a memory vector output by the blood sugar prediction sequence model according to a previous feature vector on the time sequence; and obtaining a blood sugar prediction result of the user by adopting a preset prediction algorithm based on the memory vector output by the blood sugar prediction sequence model according to the last feature vector on the time sequence.
11. The apparatus according to claim 10, wherein the prediction module is specifically configured to obtain an initial memory vector for the first feature vector in the time sequence; inputting the initial memory vector and the first feature vector in the time sequence into the blood sugar prediction sequence model; and each element in the initial memory vector takes a random value or takes a zero value.
12. The apparatus of claim 7, further comprising:
a push module configured to generate a health advice report for the user based on the blood glucose prediction result; performing at least one of the following operations on the health advice report: displaying the health advice report; sending the health suggestion report to a preset server; and sending the health suggestion report to a preset terminal device.
13. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1 to 6 when executing the program.
14. A non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010125312.0A CN111329491A (en) | 2020-02-27 | 2020-02-27 | Blood glucose prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010125312.0A CN111329491A (en) | 2020-02-27 | 2020-02-27 | Blood glucose prediction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111329491A true CN111329491A (en) | 2020-06-26 |
Family
ID=71173956
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010125312.0A Pending CN111329491A (en) | 2020-02-27 | 2020-02-27 | Blood glucose prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111329491A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112017778A (en) * | 2020-08-06 | 2020-12-01 | 东北大学 | Clinical-oriented multi-level blood glucose abnormity early warning method |
| CN112116967A (en) * | 2020-09-25 | 2020-12-22 | 苏州市立医院 | Information processing method and device for improving urine sample collection efficiency |
| CN113679348A (en) * | 2021-08-26 | 2021-11-23 | 平安国际智慧城市科技股份有限公司 | Blood glucose prediction method, blood glucose prediction device, blood glucose prediction apparatus, and storage medium |
| CN113951879A (en) * | 2021-12-21 | 2022-01-21 | 苏州百孝医疗科技有限公司 | Blood glucose prediction method and device and system for monitoring blood glucose level |
| WO2022083124A1 (en) * | 2020-10-22 | 2022-04-28 | 平安科技（深圳）有限公司 | Personalized diabetes health management system and device, and storage medium |
| CN115662616A (en) * | 2022-10-24 | 2023-01-31 | 重庆联芯致康生物科技有限公司 | Critical patient intelligent blood sugar management system based on CGM and management method thereof |
| CN117612737A (en) * | 2024-01-24 | 2024-02-27 | 胜利油田中心医院 | Intelligent optimization method for diabetes care data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110098548A1 (en) * | 2009-10-22 | 2011-04-28 | Abbott Diabetes Care Inc. | Methods for modeling insulin therapy requirements |
| CN103605878A (en) * | 2013-08-19 | 2014-02-26 | 浙江大学 | General blood glucose prediction method based on data modeling and model transplanting |
| US20140073892A1 (en) * | 2011-04-20 | 2014-03-13 | Novo Nordisk A/S | Glucose predictor based on regularization networks with adaptively chosen kernels and regularization parameters |
| US20150347707A1 (en) * | 2014-05-30 | 2015-12-03 | Anthony Michael Albisser | Computer-Implemented System And Method For Improving Glucose Management Through Cloud-Based Modeling Of Circadian Profiles |
| WO2019129857A1 (en) * | 2017-12-29 | 2019-07-04 | Hillo Ai | Systems and methods for prediction of glycemia and decisions support |
-
2020
- 2020-02-27 CN CN202010125312.0A patent/CN111329491A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110098548A1 (en) * | 2009-10-22 | 2011-04-28 | Abbott Diabetes Care Inc. | Methods for modeling insulin therapy requirements |
| US20140073892A1 (en) * | 2011-04-20 | 2014-03-13 | Novo Nordisk A/S | Glucose predictor based on regularization networks with adaptively chosen kernels and regularization parameters |
| CN103605878A (en) * | 2013-08-19 | 2014-02-26 | 浙江大学 | General blood glucose prediction method based on data modeling and model transplanting |
| US20150347707A1 (en) * | 2014-05-30 | 2015-12-03 | Anthony Michael Albisser | Computer-Implemented System And Method For Improving Glucose Management Through Cloud-Based Modeling Of Circadian Profiles |
| WO2019129857A1 (en) * | 2017-12-29 | 2019-07-04 | Hillo Ai | Systems and methods for prediction of glycemia and decisions support |
Non-Patent Citations (1)
| Title |
|---|
| WILLIAM SANDHAM ET AL.: "Blood glucose prediction for diabetes therapy using a recurrent artificial neural network", 《CONF PROC EUSIPCO》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112017778A (en) * | 2020-08-06 | 2020-12-01 | 东北大学 | Clinical-oriented multi-level blood glucose abnormity early warning method |
| CN112017778B (en) * | 2020-08-06 | 2024-04-05 | 东北大学 | Clinical multi-level blood glucose abnormality early warning method |
| CN112116967A (en) * | 2020-09-25 | 2020-12-22 | 苏州市立医院 | Information processing method and device for improving urine sample collection efficiency |
| WO2022083124A1 (en) * | 2020-10-22 | 2022-04-28 | 平安科技（深圳）有限公司 | Personalized diabetes health management system and device, and storage medium |
| CN113679348A (en) * | 2021-08-26 | 2021-11-23 | 平安国际智慧城市科技股份有限公司 | Blood glucose prediction method, blood glucose prediction device, blood glucose prediction apparatus, and storage medium |
| CN113679348B (en) * | 2021-08-26 | 2024-02-06 | 深圳平安智慧医健科技有限公司 | Blood glucose prediction method, blood glucose prediction device, apparatus, and storage medium |
| CN113951879A (en) * | 2021-12-21 | 2022-01-21 | 苏州百孝医疗科技有限公司 | Blood glucose prediction method and device and system for monitoring blood glucose level |
| CN115662616A (en) * | 2022-10-24 | 2023-01-31 | 重庆联芯致康生物科技有限公司 | Critical patient intelligent blood sugar management system based on CGM and management method thereof |
| CN117612737A (en) * | 2024-01-24 | 2024-02-27 | 胜利油田中心医院 | Intelligent optimization method for diabetes care data |
| CN117612737B (en) * | 2024-01-24 | 2024-05-03 | 胜利油田中心医院 | Intelligent optimization method for diabetes care data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111329491A (en) | Blood glucose prediction method and device, electronic equipment and storage medium | |
| Li et al. | GluNet: A deep learning framework for accurate glucose forecasting | |
| Martinsson et al. | Blood glucose prediction with variance estimation using recurrent neural networks | |
| US11544610B2 (en) | Adaptive analytical behavioral and health assistant system and related method of use | |
| Ali et al. | Continuous blood glucose level prediction of type 1 diabetes based on artificial neural network | |
| Daskalaki et al. | Model-free machine learning in biomedicine: Feasibility study in type 1 diabetes | |
| US10420489B2 (en) | System coordinator and modular architecture for open-loop and closed-loop control of diabetes | |
| Kheifets et al. | Mice take calculated risks | |
| US20220257199A1 (en) | System and method for online domain adaptation of models for hypoglycemia prediction in type 1 diabetes | |
| van Swieten et al. | Modeling the effects of motivation on choice and learning in the basal ganglia | |
| Fitzgerald et al. | Incorporating real-world evidence into the development of patient blood glucose prediction algorithms for the ICU | |
| Jacobs et al. | Artificial intelligence and machine learning for improving glycemic control in diabetes: best practices, pitfalls and opportunities | |
| Erbas et al. | Embodied imitation-enhanced reinforcement learning in multi-agent systems | |
| Zhu et al. | Population-specific glucose prediction in diabetes care with transformer-based deep learning on the edge | |
| US20220359079A1 (en) | Systems, methods and devices for predicting personalized biological state with model produced with meta-learning | |
| CN113076486B (en) | Drug information pushing method, device, computer equipment and storage medium | |
| Hettiarachchi et al. | G2P2C—A modular reinforcement learning algorithm for glucose control by glucose prediction and planning in Type 1 Diabetes | |
| Goyal et al. | CarbMetSim: A discrete-event simulator for carbohydrate metabolism in humans | |
| CN112037932A (en) | Patient medication behavior intervention method and device, server and storage medium | |
| Avila et al. | A simple method for recommending specialized specifications for diabetes monitoring | |
| US20220375615A1 (en) | Method and system for predicting health risk | |
| Gustin et al. | Diabetes management through artificial intelligence and gamification: blood glucose prediction models and mHealth design considerations | |
| Ginsberg et al. | Meeting Report: The Fourth Artificial Pancreas Workshop: Testing and Adoption of Current and Emerging Technologies | |
| JP2024097132A (en) | Processing system, processing method, and program | |
| Breton et al. | Artificial intelligence and machine learning for improving glycemic control in diabetes: best practices, pitfalls and opportunities |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |