Method and device for constructing chronic obstructive pulmonary disease prediction model based on millimeter wave radar

Abstract

The invention relates to a method and system for constructing a chronic obstructive pulmonary disease prediction model based on a millimeter-wave radar, and the method comprises the steps: determining the motion state of a target in real time based on an echo signal of the millimeter-wave radar which carries out the step test of the target according to the echo signal; adjusting a filtering method of the echo signal according to the motion state of the target; through a variational mode decomposition method and a spatial spectrum estimation method, sign data of the target are extracted from the filtered echo signals, and the sign data comprise the speed, the acceleration, the respiratory frequency and the heartbeat frequency; constructing a chronic obstructive pulmonary disease data set based on the sign data of the multiple targets; and based on the chronic obstructive pulmonary disease data set, constructing a chronic obstructive pulmonary disease prediction model through a random forest method. According to the method, unattended sign data measurement is realized through echoes of the millimeter wave radar, and the prediction model is constructed through the random forest, so that the accuracy of model prediction is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of millimeter-wave radar and physiological data, and particularly relates to a method and device for constructing a chronic obstructive pulmonary disease (COPD) prediction model based on millimeter-wave radar. Background Art

[0002] Chronic obstructive pulmonary disease is the third leading cause of death in China, which can lead to dyspnea and has problems such as low awareness rate and low diagnosis rate. There are several main methods for the initial detection of existing COPD patients, including patients filling out questionnaires based on their own symptoms and exercise ability, and the risk of disease can be evaluated through scores, but the sensitivity and specificity are insufficient; in the early stage of COPD, small airway lesions and alveolar destruction will occur, and chest CT can identify and quantify them, but it involves radiation exposure, and frequent examinations may increase the risk of patients receiving radiation, and the examinations are relatively expensive; measuring the forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) can evaluate the condition of airflow limitation, and the FEV1 / FVC ratio is the gold standard for the diagnosis of COPD, which can be measured by a spirometer, but the measurement requires the patient's close cooperation, such as forced exhalation, etc., which may be a challenge for some patients (such as children), and the requirements for operators are relatively high. Summary of the Invention

[0003] To solve the problems raised in the background art, in the first aspect of the present invention, a method for constructing a COPD prediction model based on millimeter-wave radar is provided, including: based on the echo signal of a millimeter-wave radar for a target to perform a step test, and determining the motion state of the target in real time according to the echo signal; adjusting the filtering method of the echo signal according to the motion state of the target; extracting the physical sign data of the target from the filtered echo signal through the variational mode decomposition method and the spatial spectrum estimation method, where the physical sign data includes speed, acceleration, respiratory rate, and heart rate; constructing a COPD disease dataset based on the physical sign data of multiple targets; and constructing a COPD disease prediction model through the random forest method based on the COPD disease dataset.

[0004] In some embodiments of the present invention, the adjusting the filtering method of the echo signal according to the motion state of the target includes: if the target is stationary, filtering the phase and frequency of the echo signal.

[0005] In some embodiments of the present invention, the extracting the physical sign data of the target from the filtered echo signal through the variational mode decomposition method and the spatial spectrum estimation method includes: extracting the respiratory signal and heart signal of the target from the filtered echo signal through the variational mode decomposition method; and extracting the respiratory rate and heart rate of the target from the filtered echo signal through the spatial spectrum estimation method.

[0006] Further, extracting the respiration signal and heartbeat signal of the target from the filtered echo signal by the variational mode decomposition method includes: decomposing the filtered echo signal into multiple components, and establishing a constrained variational model according to the multiple components; through the quadratic penalty term and the Lagrange multiplier, transforming the constrained variational model into an unconstrained variational model; based on the quadratic penalty term and the Lagrange multiplier, iterating the unconstrained variational model multiple times until a preset convergence condition is reached, obtaining the optimal solution of the unconstrained variational model; based on the optimal solution, separating the respiration signal and the heartbeat signal.

[0007] Further, extracting the respiration frequency and heartbeat frequency of the target from the filtered echo signal by the spatial spectrum estimation method includes: constructing a covariance matrix of the filtered echo signal, and performing eigenvalue decomposition on the covariance matrix to obtain a signal subspace and a noise subspace; based on the signal subspace and the noise subspace, constructing a cost function; determining the respiration frequency and the heartbeat frequency through the peak value of the cost function.

[0008] In the above embodiment, determining the motion state of the target according to the echo signal in real time includes: mixing and filtering the echo signal and the transmitted signal of the millimeter radar wave to obtain an intermediate frequency signal; performing Fourier transform and spectrum analysis on the intermediate frequency signal to obtain the distance data and velocity data of the target; based on the distance data and the velocity data, determining the motion state of the target.

[0009] In the second aspect of the present invention, there is provided a system for constructing a chronic obstructive pulmonary disease prediction model based on a millimeter wave radar, including: a determination module, configured to determine the motion state of a target in real time based on the echo signal of the millimeter wave radar for the target to perform a step test, and according to the echo signal; an adjustment module, configured to adjust the filtering method of the echo signal according to the motion state of the target; an extraction module, configured to extract the physical sign data of the target from the filtered echo signal by the variational mode decomposition method and the spatial spectrum estimation method, where the physical sign data includes velocity, acceleration, respiration frequency, and heartbeat frequency; a construction module, configured to construct a chronic obstructive pulmonary disease data set based on the physical sign data of multiple targets; and based on the chronic obstructive pulmonary disease data set, constructing a chronic obstructive pulmonary disease prediction model by the random forest method.

[0010] In the third aspect of the present invention, there is provided an electronic device, including: one or more processors; a storage device, configured to store one or more programs, and when the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the method for constructing a chronic obstructive pulmonary disease prediction model based on a millimeter wave radar provided in the first aspect of the present invention.

[0011] In a fourth aspect of the present invention, there is provided a computer-readable medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the method for constructing a prediction model based on millimeter-wave radar and chronic obstructive pulmonary disease provided by the present invention in the first aspect.

[0012] The beneficial effects of the present invention are as follows: The present disclosure reduces the complexity of collecting chronic obstructive pulmonary disease data, improves test safety, and at the same time inputs millimeter-wave radar information back into the voice module for broadcasting, which can timely feedback the test progress; after the tester completes all exercise cycle tests, he / she rests at a set position for two minutes, observes the changes in heart rate and respiration values, collects all information during the test, and inputs it into a pre-trained random forest classification model, which can identify the result in a relatively short time; on the other hand, it can be carried out without human supervision, and patients can test themselves while ensuring the safety of the tester. After multiple verification tests, the accuracy rate of the model in identifying patients with chronic obstructive pulmonary disease reaches 92%. At the same time, the test results are reported to the test platform, and the tester can further check according to the test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic flowchart of the basic process of the method for constructing a prediction model based on millimeter-wave radar and chronic obstructive pulmonary disease in some embodiments of the present invention; Figure 2 It is one of the schematic flowcharts of the random forest method in some embodiments of the present invention; Figure 3 It is another schematic flowchart of the random forest method in some embodiments of the present invention; Figure 4 It is a schematic structural diagram of the system for constructing a prediction model based on millimeter-wave radar and chronic obstructive pulmonary disease in some embodiments of the present invention; Figure 5 It is a schematic structural diagram of an electronic device in some embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0015] Refer to Figure 1, in the first aspect of the present invention, a method for constructing a chronic obstructive pulmonary disease (COPD) prediction model based on a millimeter-wave radar is provided, including: S100. Based on the echo signal of the millimeter-wave radar for the target to perform a step test, and determine the motion state of the target in real time according to the echo signal; S200. Adjust the filtering method of the echo signal according to the motion state of the target; S300. Extract the physical sign data of the target from the filtered echo signal through the variational mode decomposition method and the spatial spectrum estimation method, where the physical sign data includes speed, acceleration, respiratory rate, and heart rate; S400. Based on the physical sign data of multiple targets, construct a COPD disease dataset; based on the COPD disease dataset, construct a COPD disease prediction model through the random forest method.

[0016] Schematically, the basic process of the step test is as follows: The tester stands behind the starting line. After the radar detects the target, the voice module issues an instruction: "The test starts. Each time you go up and down the stairs is regarded as an activity cycle. You need to complete ten activity cycles. Please keep walking during the test and pay attention to your safety under your feet." Then the tester starts climbing the stairs. The radar detects it as a motion mode, can monitor the motion trajectory of the target, and at the same time calculates the speed and acceleration information and uploads it to the random forest model as input. The tester will receive a voice prompt every time an odd number of activity cycles are completed: "You have completed 1 (3, 5, 7, 9) activity cycles. Come on!" After every even number of activity cycles, rest for thirty seconds at the designated position and receive a voice prompt: "You have completed 2 (4, 6, 8) activity cycles. Please stand in the red area and rest in place for thirty seconds." "You have completed 10 activity cycles. Please stand in the red area and rest in place for two minutes." When resting, the radar detects it as a stationary mode, measures the heart rate and respiratory value of the test target, and uploads it to the random forest model as input, and at the same time feeds it back to the voice module. According to the preset threshold, it is judged whether the tester can continue the test. If so, the voice broadcasts: "You can continue the test." If not, the voice broadcasts: "Please pause the test!" At this time, it is identified as an abnormal situation and uploaded to the test platform. According to the actual situation, the threshold is set to 60 breaths per minute for respiration and 160 beats per minute for heart rate; repeat the test until ten activity cycles are completed. During this period, the voice module randomly broadcasts "Not bad, keep it up." "You're doing great." "Very good! Hold on a little longer." After the test is completed, the voice module issues an instruction: "You have completed the test. Please leave the test site." The collected test data is classified to determine whether the tester is a COPD patient, and the result is uploaded to the test platform. The tester further checks according to the test result.

[0017] In step S100 of some embodiments of the present invention, the real-time determination of the motion state of the target according to the echo signal includes: S101. Mix and filter the echo signal and the transmitted signal of the millimeter radar wave to obtain an intermediate frequency signal; Specifically, a millimeter-wave radar based on frequency modulated continuous wave (FMCW) is adopted. The antenna serves as a transceiver. The radar module transmits a linear frequency modulation pulse signal (TX) whose frequency changes linearly with time. After being reflected by the target, an echo signal (RX) is obtained. The original signal and the echo signal can obtain an intermediate frequency signal (IF) through a mixer and a low-pass filter.

[0018] S102. Perform Fourier transform and spectral analysis on the intermediate frequency signal to obtain the distance data and speed data of the target; Specifically, the frequency of the intermediate frequency signal is the frequency difference between TX and RX, and the initial phase is the phase difference at the current moment. The ADC (analog-to-digital conversion) sampling of the intermediate frequency signal obtains the time-domain signal of the radar. The original data of the signal sampled by ADC is subjected to multi-pulse non-coherent accumulation processing for each Chirp to form a matrix with a size of N×M dimensions. The distance dimension FFT is performed on the fast-time sampling points of M frames for spectral analysis to obtain the distance characteristics. The frequency corresponding to the peak extracted after Fourier transform is proportional to the distance, and the frequency can be converted into distance. Since the time of only one frame is too short, the speed information of a certain distance is selected for calculation. It is set that every 0.5 m of movement is a node. The distance characteristics combined with the frame time and the number of frames can calculate the speed. Finally, the acceleration is obtained by differentiating the speed data with respect to time. The acceleration is calculated every 5 m of movement.

[0019] S103. Based on the distance data and the speed data, determine the motion state of the target.

[0020] Specifically, the distance characteristics combined with the speed characteristics can determine whether the target is in a stationary state or a motion state. When the speed is 0 and the distance is less than 1 m (within the red area), the radar determines it as the stationary mode and tests the heart rate and breathing information of the target. When the distance is greater than 2 m (outside the starting line), the radar determines it as the motion mode and tests the speed and acceleration information of the target.

[0021] In step S200 of some embodiments of the present invention, the method for adjusting the filtering method of the echo signal according to the motion state of the target includes: S201. If the target is stationary, filter the phase and frequency of the echo signal; When the subject is in a static state, the human heartbeat and breathing information can be extracted from the intermediate frequency signal, and the position of the target can be determined by the extracted distance features. However, since the displacement change is very small during chest vibration, the chest displacement range caused by breathing is 1-12 mm, and the chest displacement caused by the heartbeat is 0.1-0.5 mm, which cannot be calculated by the intermediate frequency signal frequency information, and the phase is greatly affected by it. Therefore, the phase is processed to extract the physical sign information. The specific phase is obtained through the arctangent function. When the chest moves greatly, the absolute value of the phase difference between adjacent sampling points is greater than π. The phase unwrapping method of adding or subtracting 2π is used to process the discontinuous phase, and then the continuous phase values are subtracted by phase difference to enhance the target signal and suppress phase drift; Specifically, an FIR band-pass filter is used to screen out the breathing signal and the heartbeat signal. Since both breathing and heart rate are higher than normal after exercise, and the breathing and heart rate of patients are also higher than those of non-patients, the passband frequency ranges are respectively set to 0.2-1.2 Hz and 1.0-3.0 Hz; the rest time is set to 30 s, and the adaptive variational mode decomposition algorithm is selected to decompose the breathing signal and the heartbeat signal. The intrinsic mode components can be regarded as intrinsic mode signals with a central frequency and a finite bandwidth. By continuously iterating and updating, the final intrinsic mode function components can be obtained, which can well avoid the mode mixing phenomenon; In step S300 of some embodiments of the present invention, the extracting the physical sign data of the target from the filtered echo signal by the variational mode decomposition method and the spatial spectrum estimation method includes: S301. Extract the breathing signal and the heartbeat signal of the target from the filtered echo signal by the variational mode decomposition method; Further, in step S301, extracting the breathing signal and the heartbeat signal of the target from the filtered echo signal by the variational mode decomposition method includes: S3011. Decompose the filtered echo signal into multiple components, and establish a constrained variational model according to the multiple components; Specifically, the original signal is decomposed into k components, and a constrained variational model is established: , wherein, is the unit impulse function; to solve the constrained optimization problem, the constrained variational problem is transformed into an unconstrained variational problem, and by utilizing the advantages of the quadratic penalty term and the Lagrange multiplier method, an augmented Lagrangian function is introduced: S3012. Transform the constrained variational model into an unconstrained variational model through the quadratic penalty term and the Lagrange multiplier; Specifically, ; wherein, is the penalty factor, is the Lagrange multiplier; through the iteration of the alternating direction multiplier algorithm, for all , update the functional , until the convergence condition is satisfied.

[0022] S3013. Based on the quadratic penalty term and the Lagrange multiplier, perform multiple iterations on the unconstrained variational model until the preset convergence condition is reached, and obtain the optimal solution of the unconstrained variational model; S3014. Based on the optimal solution, separate the respiration signal and the heartbeat signal.

[0023] Specifically, the convergence condition: , , where the superscript n is the number of iterations, is the relative error of the component, is the absolute error of the component. When both equations are satisfied, the saddle point of the augmented Lagrangian model is obtained, the optimal solution of the model is calculated, and according to the of the component, the corresponding signal is determined to realize the separation of the respiration signal and the heartbeat signal.

[0024] S302. Through the spatial spectrum estimation method, extract the respiration frequency and heartbeat frequency of the target from the filtered echo signal.

[0025] It can be understood that the VMD algorithm determines the frequency and amplitude of each mode function through continuous optimization to achieve effective signal decomposition. Decompose a real signal into a series of mode functions , each 's spectrum is concentrated near its central frequency . The algorithm updates iteratively, including updating the mode signal and the central frequency, so that the sum of all mode functions has the minimum error with the original signal, while maintaining the smoothness of the mode signal. VMD requires the mode function to meet two conditions: the sum of the mode bandwidths is the smallest, and the sum of the modes is equal to the original signal; VMD transforms the objective function into a constrained Lagrangian function and optimizes it through the variational method.

[0026] Furthermore, the extracting the respiration frequency and heartbeat frequency of the target from the filtered echo signal through the spatial spectrum estimation method includes: S3021. Construct the covariance matrix of the filtered echo signal, and perform eigenvalue decomposition on the covariance matrix to obtain the signal subspace and the noise subspace; Specifically, truncate the input signal S(t) into a signal X(t) with a length of N*M, that is, construct an N*M data matrix, that is , ， , where A is the array direction matrix and N(t) is the noise matrix; the covariance matrix R is obtained; , where is the signal covariance matrix, is the noise covariance matrix; S3022. Based on the signal subspace and the noise subspace, construct a cost function; Specifically, perform eigenvalue decomposition on the covariance matrix to obtain the signal subspace US and the noise subspace UN, divide the angle [0, 2π] into Nω angle points, and obtain the angle search vector A( ):

[0027] Construct the cost function as: , where is 's conjugate matrix, is 's conjugate matrix; S3023. Determine the breathing frequency and the heart rate through the peak value of the cost function.

[0028] Specifically, using the orthogonality of the signal space and the noise space, we can obtain 's maximum value, and find the frequency corresponding to the peak angle through the peak value of the cost function, that is, the breathing frequency and the heart rate.

[0029] It can be understood that by performing eigen-decomposition on the covariance matrix of the received data, according to the signal autocorrelation, it is decomposed into eigen-signals with strong autocorrelation and eigen-signals with weak autocorrelation. The strong autocorrelation signals are regarded as the main eigen-signals, while the weak autocorrelation signals are regarded as noise signals. Utilize the orthogonality of the separated signal subspace and noise subspace to determine the angle, form the spatial domain and perform global search for spectral peaks, thereby realizing the parameter estimation of the signal.

[0030] Reference Figure 2 and Figure 3 , in step S400 of some embodiments of the present invention, based on the physical sign data of multiple targets, construct a chronic obstructive pulmonary disease dataset; based on the chronic obstructive pulmonary disease dataset, construct a chronic obstructive pulmonary disease prediction model through the random forest method.

[0031] Specifically, the random forest ensemble learning method is adopted to improve the prediction accuracy by constructing multiple decision trees and outputting the average result. The inputs are four features: speed, acceleration, heart rate, and respiratory value, and the outputs are two categories: patients and non-patients. The missing values and outliers in the collected data are processed; then, the features are normalized to improve the model training efficiency and performance; the feature importance is evaluated to determine the feature weights, that is, 20% for speed and acceleration each, and 30% for heart rate and respiratory value each, to optimize the model performance; the dataset is divided into a training set and a test set, with 80% for the training set and 20% for the test set. The training set is used to train the model, and the parameters of the random forest, such as the number of trees and the maximum depth, are adjusted to find the optimal parameter combination and the highest accuracy; the model consists of multiple decision trees, and each tree uses a randomly selected subset of features during training. The final classification result is determined by the voting results of each tree. Finally, the trained model is used for classification to achieve the purpose of screening patients with chronic obstructive pulmonary disease (COPD).

[0032] Specifically, the random forest ensemble learning method is mainly divided into three processes: training set generation, decision tree training to form a random forest, and test set testing, that is, the formation and testing of the random forest.

[0033] The input training dataset is the 100 non-COPD patients and 100 COPD patients collected, and the features include speed, acceleration, heart rate value, and respiratory value. The Bootstrap method is used to randomly sample from the original training set with replacement. Each node at each level represents a specific feature of the input data, and the leaf nodes represent the categories to which the data belongs. This process starts from the root node, and according to the splitting criterion of the node, we divide based on a certain feature of the input object. Through such steps, we will finally reach a specific leaf node to obtain the prediction result of the decision tree.

[0034] The splitting algorithm selects the CART algorithm. CART is called Classification and Regression Tree, which is a binary decision tree. When splitting nodes, the principle of the minimum Gini index is adopted to calculate the impurity of the set (P). The Gini index selection criterion is that each child node reaches the highest purity. The Gini coefficient is: , where represents the probability of the category in the set, calculate the coefficient after splitting. When divided into two categories, obtain the Gini coefficient for this time: , where is the number of samples that meet the conditions, is the number of samples that meet the conditions, is the total number of sets. The feature with the lowest Gini index is selected for splitting, and this feature is considered the best choice in the current splitting step. Different splitting methods or combinations of each feature are evaluated to determine the best splitting point for each feature. Then, by comparing the optimal splitting points of each feature, multiple classification trees are generated by allowing each decision tree to grow maximally without any pruning, and these multiple classification trees form a random forest, and finally the best classification is determined.

[0035] Example 2 Reference Figure 4 , in the second aspect of the present invention, there is provided a system 1 for constructing a prediction model of chronic obstructive pulmonary disease based on a millimeter-wave radar, including: a determination module 11, configured to determine the motion state of a target in real time based on the echo signal of the millimeter-wave radar for the target to perform a step test, and according to the echo signal; an adjustment module 12, configured to adjust the filtering method of the echo signal according to the motion state of the target; an extraction module 13, configured to extract the physical sign data of the target from the filtered echo signal through a variational mode decomposition method and a spatial spectrum estimation method, where the physical sign data includes speed, acceleration, respiratory rate, and heart rate; a construction module 14, configured to construct a chronic obstructive pulmonary disease data set based on the physical sign data of multiple targets; and construct a chronic obstructive pulmonary disease prediction model through a random forest method based on the chronic obstructive pulmonary disease data set.

[0036] Further, the extraction module 13 includes: a first extraction unit, configured to extract the respiratory signal and heart signal of the target from the filtered echo signal through a variational mode decomposition method; a second extraction unit, configured to extract the respiratory rate and heart rate of the target from the filtered echo signal through a spatial spectrum estimation method.

[0037] Example 3 Reference Figure 5 , in the third aspect of the present invention, there is provided an electronic device, including: one or more processors; a storage device, configured to store one or more programs, when the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the method for constructing a prediction model of chronic obstructive pulmonary disease based on a millimeter-wave radar in the first aspect of the present invention.

[0038] The electronic device 500 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 501, which may perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 502 or the program loaded from the storage device 508 into the random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. The input / output (I / O) interface 505 is also connected to the bus 504.

[0039] Typically, the following devices can be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a hard disk, etc.; and a communication device 509. The communication device 509 can allow the electronic device 500 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 5 the electronic device 500 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices can be implemented or had. Figure 5 Each block shown in

[0040] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program includes program code for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-described functions defined in the methods of the embodiments of the present disclosure are performed. It should be noted that the computer-readable medium described in the embodiments of the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiments of the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the embodiments of the present disclosure, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0041] The above computer-readable medium can be included in the above electronic device; or can exist separately without being assembled into the electronic device. The above computer-readable medium carries one or more computer programs, and when the above one or more programs are executed by the electronic device, the electronic device is caused to: Computer program code for performing the operations of the embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, Python, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0042] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should be noted that each block in the block diagram and / or flowchart, and the combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0043] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar, characterized in that: include: Based on the echo signal of the millimeter wave radar performing step test on the target, the motion state of the target is determined in real time according to the echo signal; Adjust the filtering method of the echo signal according to the target's motion state; Extracting target vital sign data from the filtered echo signal by using a variational mode decomposition method and a spatial spectrum estimation method, wherein the vital sign data includes speed, acceleration, respiratory rate and heart rate; Based on the vital sign data of multiple targets, a COPD dataset was constructed. Based on the COPD dataset, a COPD prediction model was constructed using the random forest method.

2. The method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 1, characterized in that: The filtering method for adjusting the echo signal according to the motion state of the target includes: If the target is stationary, the phase and frequency of the echo signal are filtered.

3. The method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 1, characterized in that: The step of extracting the target's vital sign data from the filtered echo signal by using a variational mode decomposition method and a spatial spectrum estimation method comprises: The target's breathing signal and heartbeat signal are extracted from the filtered echo signal by using the variational mode decomposition method; The target's breathing rate and heart rate are extracted from the filtered echo signal through the spatial spectrum estimation method.

4. The method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 3, characterized in that: By using the variational mode decomposition method, the target's breathing signal and heartbeat signal are extracted from the filtered echo signal, including: Decomposing the filtered echo signal into multiple components, and establishing a constrained variational model based on the multiple components; The constrained variational model is transformed into an unconstrained variational model through quadratic penalty terms and Lagrange multipliers; Based on the quadratic penalty term and Lagrange multiplier, the unconstrained variational model is iterated multiple times until the preset convergence condition is reached, and the optimal solution of the unconstrained variational model is obtained; Based on the optimal solution, the breathing signal and the heartbeat signal are separated.

5. The method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 3, characterized in that: The method of extracting the target's breathing frequency and heart rate from the filtered echo signal by using the spatial spectrum estimation method includes: Construct the covariance matrix of the filtered echo signal, and perform eigendecomposition on the covariance matrix to obtain the signal subspace and the noise subspace; Constructing a cost function based on the signal subspace and the noise subspace; The breathing frequency and the heart rate are determined by the peak value of the cost function.

6. The method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 1, characterized in that: Determining the motion state of the target in real time according to the echo signal comprises: Mix and filter the echo signal and the transmission signal of the millimeter radar wave to obtain an intermediate frequency signal; Performing Fourier transform and spectrum analysis on the intermediate frequency signal to obtain distance data and speed data of the target; Based on the distance data and the speed data, a motion state of the target is determined.

7. A system for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 1, characterized in that: include: A determination module, used for performing a step test on the target based on the echo signal of the millimeter-wave radar, and determining the motion state of the target in real time according to the echo signal; An adjustment module, used for adjusting the filtering method of the echo signal according to the motion state of the target; An extraction module, used to extract the target's vital sign data from the filtered echo signal by using a variational mode decomposition method and a spatial spectrum estimation method, wherein the vital sign data includes speed, acceleration, respiratory rate and heart rate; A construction module is used to construct a COPD dataset based on the vital sign data of multiple targets; based on the COPD dataset, a COPD prediction model is constructed using the random forest method.

8. The system for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter wave radar according to claim 7, characterized in that: The extraction module comprises: A first extraction unit is used to extract the target's breathing signal and heartbeat signal from the filtered echo signal by using a variational mode decomposition method; The second extraction unit is used to extract the target's breathing frequency and heart rate from the filtered echo signal by using a spatial spectrum estimation method.

9. An electronic device, comprising: one or more processors; A storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter-wave radar as described in any one of claims 1 to 6.

10. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method for constructing a prediction model for chronic obstructive pulmonary disease based on millimeter-wave radar and chronic obstructive pulmonary disease is implemented as described in any one of claims 1 to 6.

Images