Non-invasive methods, devices, equipment, and storage media for detecting abnormal blood glucose concentrations

Abstract

This invention provides a non-invasive method, device, equipment, and storage medium for detecting abnormal blood glucose concentrations. It acquires electrocardiogram (ECG) signal data and extracts features from the ECG signal data using a tensor parallel factor decomposition algorithm to obtain ECG signal feature parameters. Then, by comparing and analyzing the ECG signal feature parameters with preset parameters, it obtains the analysis results of abnormal blood glucose concentration states. This invention is based on the principle that abnormal fluctuations in blood glucose may cause greater damage to myocardial cells than an overall increase in blood glucose levels, affecting the function of the cardiac conduction system and leading to abnormal ECGs. It analyzes whether there are abnormalities in the ECG features of the target blood glucose level and reduces the computational complexity of feature extraction through a tensor parallel factor decomposition algorithm, achieving low-cost, high-sensitivity, and high-accuracy non-invasive detection of abnormal blood glucose concentrations.

Description

Technical Field

[0001] This application relates to the field of blood glucose detection, and in particular to a non-invasive method, apparatus, device, and storage medium for detecting abnormal blood glucose concentrations. Background Technology

[0002] Currently, the mainstream methods for self-monitoring blood glucose are generally invasive. While invasive methods are sensitive and effective, they are not only costly in terms of consumables, but repeated sampling also carries the risk of infection. Therefore, non-invasive blood glucose testing technology is of significant research importance.

[0003] Currently, non-invasive blood glucose testing methods include electrical, optical, and acoustic methods. However, while these methods are non-invasive and have relatively fast response times, their sensitivity is not high. In practical applications, target groups who need to frequently monitor their blood glucose levels often require timely detection of abnormal blood glucose levels. Therefore, if non-invasive blood glucose testing methods cannot guarantee sensitivity to abnormal blood glucose states, it will pose health risks to those using these methods. Summary of the Invention

[0004] This invention provides a non-invasive method, device, equipment, and storage medium for detecting abnormal blood glucose concentrations. By extracting features from electrocardiogram signal data, it achieves low-cost, high-sensitivity, and high-accuracy non-invasive detection of abnormal blood glucose concentrations.

[0005] In a first aspect, the present invention provides a non-invasive method for detecting abnormal blood glucose concentrations, comprising the following steps:

[0006] Acquire electrocardiogram (ECG) signal data;

[0007] The ECG signal data is subjected to feature extraction using the tensor parallel factorization algorithm to obtain ECG signal feature parameters.

[0008] By comparing and analyzing the characteristic parameters of the electrocardiogram signal with the preset parameters, the analysis results of abnormal blood glucose concentration are obtained.

[0009] Furthermore, the feature extraction of the electrocardiogram signal data includes the following steps:

[0010] The electrocardiogram signal data is divided into multiple sub-modules to construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix.

[0011] By using cubic tensor factorization, the sample covariance matrices of the multiple sub-modules are synthesized into a third-order tensor.

[0012] The third-order tensor is expanded into a tensor matrix and then iterated using the least squares method to obtain the sample weight matrix.

[0013] Based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method.

[0014] Furthermore, the step of dividing the electrocardiogram signal data into multiple sub-modules and constructing a sample covariance matrix includes the following steps:

[0015] Based on the electrocardiogram signal data, a sample covariance matrix is ​​constructed, and the expression for the sample covariance matrix is ​​as follows:

[0016] E[x(t)x H(t) ] = AR s A

[0017] Where x(t) represents the electrocardiogram signal data at time t, x H (t) denotes the conjugate of x(t), E[·] denotes the expectation operation, and R S =E[s(t)s H [(t)] represents the sample core matrix, s(t) represents the ECG signal data features at time t, and A represents the sample weight matrix.

[0018] Further, obtaining the sample weight matrix includes the following steps:

[0019] The electrocardiogram signal data is divided into P sub-modules, and the expression for the sample covariance matrix of the P sub-modules is as follows:

[0020]

[0021] The Combined into a third-order tensor R x Define matrix C for the third-order tensor R. x Perform tensor matrix expansion to obtain the R x The matrix expression is

[0022] R x =[A⊙A*]·C T

[0023] Where ⊙ represents the Khatri-Rao product;

[0024] The sample weight matrix equation is constructed using singular value decomposition, and the expression for the sample weight matrix equation is as follows:

[0025] R x =U∑V H

[0026]

[0027] Based on the sample weight matrix equation, the sample weight matrix is ​​obtained by iteratively applying the least squares method. The iterative process further optimizes the convergence speed of the least squares method using an exact line search algorithm.

[0028] Furthermore, based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method, including the following steps:

[0029] Sort the sample weight matrix to obtain adaptive weights;

[0030] Based on the sample weight matrix and the adaptive weights, features of the electrocardiogram signal are extracted through sparse reconstruction. s The features s The expression is

[0031] s = Π -1 (s)A T [AΠ -1 (s)A T ] -1 x

[0032] Where, Π(s)=diag(|s1| -1 ,…,|s n | -1 ).

[0033] Furthermore, the electrocardiogram signal data includes at least one of the following:

[0034] The duration of the P-wave, the amplitude of the P-wave, the duration of the PQ interval, the duration of the QRS composite wave, the duration of the QT interval, and the amplitude of the T-wave.

[0035] Secondly, the present invention also provides a non-invasive blood glucose concentration abnormality detection device, comprising:

[0036] The data acquisition module is used to acquire electrocardiogram (ECG) signal data;

[0037] The feature parameter acquisition module is used to extract features from the electrocardiogram signal data using a tensor parallel factorization algorithm to obtain electrocardiogram signal feature parameters.

[0038] The abnormal result analysis module is used to obtain the abnormal blood glucose concentration analysis results by comparing and analyzing the characteristic parameters of the electrocardiogram signal with the preset parameters.

[0039] Furthermore, the feature parameter acquisition module further includes:

[0040] The sample matrix construction unit is used to divide the electrocardiogram signal data into multiple sub-modules and construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix.

[0041] A synthesis tensor unit is used to synthesize the sample covariance matrices of the multiple sub-modules into a third-order tensor through cubic tensor parallel factorization.

[0042] The weight matrix acquisition unit is used to perform tensor matrix expansion on the third-order tensor and iterate through the least squares method to obtain the sample weight matrix.

[0043] The electrocardiogram (ECG) signal feature acquisition unit is used to acquire ECG signal data features based on the sample weight matrix using a sparse reconstruction method.

[0044] Thirdly, the present invention also provides a computer device, comprising:

[0045] At least one memory and at least one processor;

[0046] The memory is used to store one or more programs;

[0047] When the one or more programs are executed by the at least one processor, the at least one processor performs the steps of a non-invasive method for detecting abnormal blood glucose concentrations as described in the first aspect.

[0048] Fourthly, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a non-invasive method for detecting abnormal blood glucose concentrations as described in the first aspect.

[0049] This invention acquires electrocardiogram (ECG) signal data and extracts features from the ECG signal data using a tensor parallel factor decomposition algorithm to obtain ECG signal feature parameters. Then, by comparing and analyzing these ECG signal feature parameters with preset parameters, it obtains the analysis results of abnormal blood glucose concentration. Based on the principle that abnormal fluctuations in blood glucose may cause greater damage to myocardial cells than an overall increase in blood glucose levels, affecting the function of the cardiac conduction system and leading to abnormal ECGs, this invention analyzes whether there are abnormalities in the ECG features of the target blood glucose level. By using a tensor parallel factor decomposition algorithm to reduce the computational complexity of feature extraction, it achieves low-cost, high-sensitivity, and high-accuracy non-invasive detection of abnormal blood glucose concentration.

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Attached Figure Description

[0051] Figure 1 This is a schematic diagram illustrating the application environment of a non-invasive method for detecting abnormal blood glucose concentrations in an exemplary embodiment.

[0052] Figure 2 This is a flowchart of a non-invasive method for detecting abnormal blood glucose concentrations in an exemplary embodiment.

[0053] Figure 3 This is a schematic diagram of experimental verification data for a non-invasive method for detecting abnormal blood glucose concentrations in an exemplary embodiment.

[0054] Figure 4 This is a schematic diagram of a non-invasive blood glucose concentration abnormality detection device provided in an exemplary embodiment;

[0055] Figure 5 This is an internal structural diagram of a computer device provided in one exemplary embodiment;

[0056] Figure 6 This is an internal structural diagram of a computer device provided in one exemplary embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0058] It should be understood that the described embodiments are merely some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of the embodiments of this application.

[0059] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0060] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0061] Furthermore, in the description of this application, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0062] Current research indicates a correlation between electrocardiogram (ECG) signals and blood glucose levels. Therefore, abnormal blood glucose levels can be indirectly determined by detecting ECG signals. Since abnormal fluctuations in blood glucose may cause greater damage to myocardial cells than an overall increase in blood glucose levels, affecting the function of the cardiac conduction system and leading to ECG abnormalities, and since ECG signals are waveforms that comprehensively reflect cardiac activity, their detection and analysis of specific traces left on the ECG signal, such as heart rate variability and ST segment changes, can help infer whether a patient's blood glucose level is abnormal.

[0063] Based on the above considerations, this application provides a non-invasive method for detecting abnormal blood glucose concentrations. This method is applied to a non-invasive blood glucose meter, such as... Figure 1 As shown, Figure 1 In an exemplary example, a non-invasive method for detecting abnormal blood glucose concentration is described. The non-invasive blood glucose meter 100 includes an electrocardiogram (ECG) signal acquisition device 101 and a processor 102. The ECG signal acquisition device 101 is connected to the processor 102. The ECG signal acquisition device 101 is used to acquire ECG signal data of the target and send it to the processor 102. The processor 102 is used to receive the data sent by the ECG signal acquisition device 101 and process the data using the non-invasive method for detecting abnormal blood glucose concentration provided in this embodiment.

[0064] The steps of a non-invasive method for detecting abnormal blood glucose concentration provided in this application are as follows: Figure 2 As shown, it includes:

[0065] S201: Acquire electrocardiogram signal data.

[0066] Specifically, electrocardiogram (ECG) signal data refers to the data information contained in an electrocardiogram. An ECG is a common method for recording the electrical activity of the heart; it detects the electrical currents generated by the heart muscle by placing electrodes on the body surface. In the embodiments of this application, the ECG signal data may include the duration of the P wave, the amplitude of the P wave, the duration of the PQ interval, the duration of the QRS complex, the duration of the QT interval, and the amplitude of the T wave.

[0067] Specifically, the P wave occurs because the sinoatrial node is located at the junction of the right atrium and the superior vena cava. Therefore, the excitation of the sinoatrial node is first conducted to the right atrium, then through the interatrial bundle to the left atrium, forming the P wave on the electrocardiogram. The P wave represents atrial excitation; the first half represents right atrial excitation, and the second half represents left atrial excitation. When the atria enlarge and conduction between the two atria becomes abnormal, the P wave may appear as a tall, peaked or biphasic P wave.

[0068] The QRS complex is formed when an impulse travels downwards through the His bundle and the left and right bundle branches, simultaneously exciting the left and right ventricles to form a QRS complex. The QRS complex represents ventricular depolarization. When there is conduction block in the left and right bundle branches of the heart, ventricular enlargement, or hypertrophy, the QRS complex becomes widened, deformed, and prolonged.

[0069] The T wave represents ventricular repolarization. In leads where the main QRS wave is upward, the T wave should be in the same direction as the main QRS wave. Changes in the T wave on an electrocardiogram are influenced by various factors. For example, myocardial ischemia can manifest as a flattened and inverted T wave. Tall T waves can be seen in hyperkalemia, the hyperacute phase of acute myocardial infarction, etc.

[0070] The QT interval represents the time it takes for the ventricles to depolarize and repolarize. A normal QT interval is 0.44 seconds. Because the QT interval is affected by heart rate, the concept of a corrected QT interval (QTC) has been introduced. Prolongation of the QT interval is often associated with the occurrence of malignant arrhythmias.

[0071] The PQ interval is the time interval between the start of the P wave and the start of the QRS complex. In cardiac electrical activity, the P wave represents atrial depolarization (contraction), while the QRS complex represents ventricular depolarization (contraction). The PQ interval is also known as the PR interval. The normal range for the PQ interval is 0.12 to 0.20 seconds (120 to 200 milliseconds), encompassing the process of atrial conduction impulses passing through the atrioventricular node and reaching the ventricular myocytes. Variations in the PQ interval can reflect the functional status of the cardiac conduction system. Abnormally shortened or lengthened PQ intervals may indicate atrioventricular conduction delay or block, or other problems with the cardiac conduction system.

[0072] S202: The ECG signal data is subjected to feature extraction using the tensor parallel factorization algorithm to obtain ECG signal feature parameters.

[0073] Tensor Parallel Factorization (TPF) is a method for decomposing high-order tensors. It effectively reduces the complexity of high-dimensional tensors and extracts their latent structural information. The basic idea of ​​TPF is to represent a high-order tensor as a product of multiple lower-dimensional factors, thereby reducing the storage and computational complexity of the original tensor. Data preprocessing: The original data is preprocessed, such as through denoising and normalization, to ensure data quality and stability. Therefore, based on the above characteristics of the Tensor Parallel Factorization algorithm, this application uses this method to extract features from electrocardiogram (ECG) signal data, aiming to achieve both efficiency and accuracy in ECG feature extraction.

[0074] In a preferred example, feature extraction of electrocardiogram signal data further includes:

[0075] The electrocardiogram signal data is divided into multiple sub-modules to construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix.

[0076] By using cubic tensor factorization, the sample covariance matrices of the multiple sub-modules are synthesized into a third-order tensor.

[0077] The third-order tensor is expanded into a tensor matrix and then iterated using the least squares method to obtain the sample weight matrix.

[0078] Based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method.

[0079] In a preferred example, the electrocardiogram signal data is divided into multiple sub-modules, and a sample covariance matrix is ​​constructed, including the following steps:

[0080] Based on the electrocardiogram signal data, a sample covariance matrix is ​​constructed, and the expression for the sample covariance matrix is ​​as follows:

[0081] E[x(t)x H(t) ] = AR s A

[0082] Where x(t) represents the electrocardiogram signal data at time t, x H (t) denotes the conjugate of x(t), E[·] denotes the expectation operation, and R s =E[s(t)s H[(t)] represents the sample core matrix, s(t) represents the ECG signal data features at time t, and A represents the sample weight matrix.

[0083] In a preferred example, obtaining the sample weight matrix further includes the following steps:

[0084] The electrocardiogram signal data is divided into P sub-modules, and the expression for the sample covariance matrix of the P sub-modules is as follows:

[0085]

[0086] The Combined into a third-order tensor R x Define matrix C for the third-order tensor R. x Perform tensor matrix expansion to obtain the R x The matrix expression is

[0087] R x =[A⊙A*]·C T

[0088] Where ⊙ represents the Khatri-Rao product;

[0089] The sample weight matrix equation is constructed using singular value decomposition, and the expression for the sample weight matrix equation is as follows:

[0090] R x =U∑V H

[0091]

[0092] Based on the sample weight matrix equation, the sample weight matrix is ​​obtained by iteratively applying the least squares method. The iterative process further optimizes the convergence speed of the least squares method using an exact line search algorithm.

[0093] Specifically, the exact line search algorithm can accelerate the convergence speed of the least squares method. Least squares is typically used to fit model parameters by minimizing the sum of squared residuals between observed data and model predictions to select optimal parameters. The exact line search algorithm is a method for finding the optimal step size; combined with least squares, it can improve the efficiency of parameter updates and convergence speed. In this embodiment, the initial parameter values ​​for the least squares method, i.e., the sample weight matrix equation, are first determined. Then, the gradient of the loss function with respect to the parameters is calculated, i.e., the partial derivative of the loss function with respect to the parameters. This gradient vector indicates the direction of change of the loss function at the current parameter value. The exact line search algorithm is then used to find the step size that results in the fastest decrease in the loss function value along the gradient direction. This can be achieved using a linear search method or a binary search method. Based on the optimal step size obtained from the exact line search, gradient descent is used to update the parameters. This iterative process is repeated until a preset stopping condition is met. The preset stopping condition can be set as a maximum iteration threshold or a parameter change threshold, etc. By using the precise line search algorithm, the step size for each parameter update can be determined more accurately, avoiding blindly updating according to a fixed step size, thereby improving the convergence speed.

[0094] In a preferred example, based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method, including the following steps:

[0095] The sample weight matrix is ​​sorted to obtain adaptive weights, wherein the sorting method is calculated as follows:

[0096]

[0097] Based on the sample weight matrix and the adaptive weights, features s of the electrocardiogram signal are extracted through sparse reconstruction. The expression for feature s is:

[0098] s = Π -1 (s)A T [AΠ -1 (s)A T ] -1 x

[0099] Where, Π(s)=diaσ(|s1| -1 ,…,|s n | -1 ).

[0100] S203: By comparing and analyzing the electrocardiogram signal characteristic parameters with the preset parameters, the abnormal blood glucose concentration analysis results are obtained.

[0101] Specifically, when the characteristic parameters of the electrocardiogram signal exceed the standard range of the preset parameters, it is determined that the blood glucose concentration of the current test subject is abnormal. This abnormal state indicates that the blood glucose value of the test subject should be further tested through other invasive or non-invasive blood glucose concentration detection methods.

[0102] The accuracy of the non-invasive blood glucose concentration abnormality detection method described in this application can be verified by comparing it with an invasive detection method that most closely approximates the true blood glucose value.

[0103] Specifically, the blood glucose level of the target is detected according to the detection rules shown in the table below, wherein the non-invasive detection method described in this application is used.

[0104]

[0105]

[0106] The verification results are as follows Figure 4 As shown, the difference between the calibrated blood glucose concentration and the predicted blood glucose concentration is within the measurement error range. Therefore, the detection method provided in this application embodiment is relatively accurate.

[0107] This application provides a non-invasive method for detecting abnormal blood glucose concentration. It acquires electrocardiogram (ECG) signal data and extracts features from the ECG signal data using a tensor parallel factor decomposition algorithm to obtain ECG signal feature parameters. The method then compares these ECG signal feature parameters with preset parameters to obtain an analysis result of the abnormal blood glucose concentration state. This invention is based on the principle that abnormal fluctuations in blood glucose may cause greater damage to myocardial cells than an overall increase in blood glucose levels, affecting the function of the cardiac conduction system and leading to abnormal ECG readings. It analyzes whether there are abnormalities in the ECG features of the target blood glucose level and reduces the computational complexity of feature extraction using a tensor parallel factor decomposition algorithm, thus achieving low-cost, high-sensitivity, and high-accuracy non-invasive detection of abnormal blood glucose concentration.

[0108] This application also provides a non-invasive blood glucose concentration abnormality detection device 300, such as... Figure 4 As shown, it includes:

[0109] Data acquisition module 301 is used to acquire electrocardiogram signal data;

[0110] The feature parameter acquisition module 302 is used to extract features from the electrocardiogram signal data using a tensor parallel factorization algorithm to obtain electrocardiogram signal feature parameters.

[0111] The abnormal result analysis module 303 is used to obtain the abnormal blood glucose concentration state analysis result by comparing and analyzing the characteristic parameters of the electrocardiogram signal with the preset parameters.

[0112] In one exemplary example, the feature parameter acquisition module 302 further includes:

[0113] The sample matrix construction unit is used to divide the electrocardiogram signal data into multiple sub-modules and construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix.

[0114] A synthesis tensor unit is used to synthesize the sample covariance matrices of the multiple sub-modules into a third-order tensor through cubic tensor parallel factorization.

[0115] The weight matrix acquisition unit is used to perform tensor matrix expansion on the third-order tensor and iterate through the least squares method to obtain the sample weight matrix.

[0116] The electrocardiogram (ECG) signal feature acquisition unit is used to acquire ECG signal data features based on the sample weight matrix using a sparse reconstruction method.

[0117] It should be noted that both the non-invasive blood glucose concentration abnormality detection device and the non-invasive blood glucose concentration abnormality detection method originate from the same inventive concept. For the relevant explanation of the non-invasive blood glucose concentration abnormality detection device, please refer to the embodiments in the non-invasive blood glucose concentration abnormality detection method, which will not be repeated here.

[0118] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When executed by the processor, the computer program implements a non-invasive method for detecting abnormal blood glucose concentrations.

[0119] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a non-invasive method for detecting abnormal blood glucose concentrations. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0120] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a non-invasive method for detecting abnormal blood glucose concentration as described in any of the above embodiments.

[0121] This invention can take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, 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 technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0122] It should be understood that the embodiments of this application are not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from their scope. The scope of the embodiments of this application is limited only by the appended claims.

[0123] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the embodiments of this application, and these all fall within the protection scope of the embodiments of this application.

Claims

1. A non-invasive method for detecting abnormal blood glucose concentration, characterized in that, Includes the following steps: Acquire electrocardiogram (ECG) signal data; The ECG signal data is subjected to feature extraction using the tensor parallel factorization algorithm to obtain ECG signal feature parameters. By comparing and analyzing the characteristic parameters of the electrocardiogram signal with preset parameters, the analysis results of abnormal blood glucose concentration are obtained; The feature extraction of the electrocardiogram signal data includes the following steps: The electrocardiogram signal data is divided into multiple sub-modules to construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix. The sample covariance matrices of the multiple sub-modules are combined into a third-order tensor; The third-order tensor is expanded into a tensor matrix and then iterated using the least squares method to obtain the sample weight matrix. Based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method.

2. The non-invasive method for detecting abnormal blood glucose concentration according to claim 1, characterized in that, The step of dividing the electrocardiogram signal data into multiple sub-modules and constructing a sample covariance matrix includes the following steps: Based on the electrocardiogram signal data, a sample covariance matrix is ​​constructed, and the expression for the sample covariance matrix is ​​as follows: E[x(t)x H (t)]=AR s A H in, express ECG signal data at any given time express The transpose and conjugate of , This represents the expectation operation. Represents the core matrix of the samples. express The characteristics of the electrocardiogram signal data at any given time, where A represents the sample weight matrix.

3. The non-invasive method for detecting abnormal blood glucose concentration according to claim 2, characterized in that, Obtaining the sample weight matrix includes the following steps: The electrocardiogram signal data is divided into P sub-modules, and the expression for the sample covariance matrix of the P sub-modules is as follows: The Combined into a third-order tensor Define matrix C for the third-order tensor. Perform tensor matrix expansion to obtain the... The matrix expression is in, Represents the Khatri-Rao product; The sample weight matrix equation is constructed using singular value decomposition, and the expression for the sample weight matrix equation is as follows: Based on the sample weight matrix equation, the sample weight matrix is ​​obtained by iteratively applying the least squares method. The iterative process further optimizes the convergence speed of the least squares method using an exact line search algorithm.

4. The non-invasive method for detecting abnormal blood glucose concentration according to claim 3, characterized in that, Based on the sample weight matrix, the electrocardiogram signal data features are obtained through a sparse reconstruction method, including the following steps: Sort the sample weight matrix to obtain adaptive weights; Based on the sample weight matrix and the adaptive weights, features s of the electrocardiogram signal are extracted through sparse reconstruction. The expression for feature s is: in, .

5. The non-invasive method for detecting abnormal blood glucose concentration according to claim 4, characterized in that, The electrocardiogram signal data includes at least one of the following: The duration of the P-wave, the amplitude of the P-wave, the duration of the PQ interval, the duration of the QRS composite wave, the duration of the QT interval, and the amplitude of the T-wave.

6. A non-invasive blood glucose concentration abnormality detection device, characterized in that, include: The data acquisition module is used to acquire electrocardiogram (ECG) signal data; The feature parameter acquisition module is used to extract features from the electrocardiogram signal data using a tensor parallel factorization algorithm to obtain electrocardiogram signal feature parameters. The abnormal result analysis module is used to obtain the abnormal blood glucose concentration analysis results by comparing and analyzing the characteristic parameters of the electrocardiogram signal with preset parameters. The feature parameter acquisition module further includes: The sample matrix construction unit is used to divide the electrocardiogram signal data into multiple sub-modules and construct a sample covariance matrix, wherein the sample covariance matrix includes a sample core matrix and a sample weight matrix. The synthesis tensor unit synthesizes the sample covariance matrices of the multiple sub-modules into a third-order tensor; The weight matrix acquisition unit is used to perform tensor matrix expansion on the third-order tensor and iterate through the least squares method to obtain the sample weight matrix. The electrocardiogram (ECG) signal feature acquisition unit is used to acquire ECG signal data features based on the sample weight matrix using a sparse reconstruction method.

7. A computer device, characterized in that, include: At least one memory and at least one processor; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the at least one processor implements the steps of a non-invasive method for detecting abnormal blood glucose concentrations as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of a non-invasive method for detecting abnormal blood glucose concentration as described in any one of claims 1 to 5.

Images