Hospital infusion monitoring and alarming method based on millimeter wave radar and electronic equipment

Abstract

The invention provides a hospital infusion monitoring and alarming method based on millimeter-wave radar and electronic equipment. The method comprises the steps that echo signals of the millimeter-wave radar of the longitudinal section of an infusion bottle are collected; de-noising processing is carried out on the echo signal through a Wiener filter; calculating angle information in the de-noised echo signal; performing image reconstruction on the de-noised echo signals to obtain an image array; determining edge points of the image array; the edge points of the image array are combined with the electromagnetic wave velocity of the millimeter wave radar and the angle information in the echo signal to obtain the coordinate information of the edge points; fitting the edge point coordinates to obtain an edge function; collecting height information of the liquid level of the infusion bottle; according to the edge function and the height information of the liquid level of the infusion bottle, the infusion amount and the infusion speed within the metering time are determined; the infusion amount and the infusion rate are compared with preset values, and when the infusion amount or the infusion rate deviates from the preset values, an alarm prompt is given; the system has the advantage of improving infusion safety and is suitable for the field of medical electronic information.

Description

Technical Field

[0001] The present application relates to the technical field of medical electronic information, and in particular to a hospital infusion monitoring and alarm method and electronic equipment based on millimeter wave radar. Background Art

[0002] Before giving patients intravenous infusions in the hospital, nurses will explain the infusion volume to the patients and their caregivers according to the doctor's medication arrangements. Some drugs, such as mannitol, will be infused in a quantitative manner. Caregivers and nurses need to monitor the liquid level of the infusion bottle during the infusion process to prevent blood return caused by overinfusion or completion of infusion. If handled improperly, it can easily lead to accidents.

[0003] Currently, in addition to manual monitoring, infrared sensors or pressure sensors cannot accurately measure the infusion volume, nor can they quantitatively warn of the completion of the infusion in advance, leaving no buffer time for nurses to change dressings or remove needles.

[0004] In the prior art, although image recognition technology based on artificial intelligence technology can solve the above problems, for example, Chinese patent number: CN115252962A discloses an intelligent infusion alarm method and system based on machine vision, but this method needs to collect image data in an environment based on surveillance cameras. As we all know, patients in the hospital are mostly infused in wards. In order to protect the privacy of patients, hospitals will not install surveillance cameras in wards. In addition, in the case of reflection on the bottle surface or weak light at night, the surveillance camera cannot collect the infusion image, and it is also impossible to achieve accurate monitoring and warning. Summary of the invention

[0005] In order to solve one of the above-mentioned technical defects, an embodiment of the present application provides a hospital infusion monitoring alarm method and electronic equipment based on millimeter wave radar.

[0006] According to a first aspect of an embodiment of the present application, a hospital infusion monitoring and alarm method based on millimeter wave radar is provided, comprising:

[0007] S1, the millimeter wave radar installed just above the longitudinal section of the infusion bottle slowly rotates 90° at a constant speed to emit a radar detection signal;

[0008] S2, collecting the echo signal of the millimeter-wave radar of the longitudinal section of the infusion bottle;

[0009] S3, denoising the echo signal through a Wiener filter;

[0010] S4, calculating the angle information in the echo signal after denoising;

[0011] S5, performing image reconstruction on the denoised echo signal to obtain an image array;

[0012] S6, determining edge points of the image array by using a Canny operator and a cross-correlation operation;

[0013] S7, combining the edge points of the image array with the electromagnetic wave velocity of the millimeter wave radar to obtain distance information of the edge points; combining the distance information of the edge points with the angle information in the echo signal to obtain coordinate information of the edge points; and fitting all discrete edge point coordinates to obtain an edge function;

[0014] S8, collecting height information of the liquid level of the infusion bottle through the FFT algorithm;

[0015] S9, determining the infusion volume and infusion rate within the measurement time through the edge function and the height information of the liquid level of the infusion bottle;

[0016] S10, comparing the infusion volume and infusion rate with the preset values, and when the infusion volume or infusion rate deviates from the preset values, an alarm is issued.

[0017] According to a second aspect of an embodiment of the present application, there is provided an electronic device, including:

[0018] memory; processor; and computer program;

[0019] The computer program is stored in the memory and is configured to be executed by the processor to implement the method as described above.

[0020] The embodiment of the present application adopts the above technical solution, which has the following technical effects:

[0021] The embodiment of the present application provides a hospital infusion monitoring alarm method and electronic equipment based on millimeter wave radar. The millimeter wave radar arranged just above the longitudinal section of the infusion bottle slowly rotates 90 degrees at a uniform speed to emit a radar detection signal. During this process, the echo signal of the millimeter wave radar of the longitudinal section of the infusion bottle is collected; then, the noise mixed in the echo signal is filtered out by a Wiener filter, and the angle information in the denoised echo signal is calculated; at the same time, by removing the noise mixed in the echo signal, the amplitude of the echo signal can be enhanced to protect the edge information, so that the overall transition of the layer is smooth and natural; the denoised echo signal is determined by the Canny operator and cross-correlation operation. The edge points of the column are then combined with the edge points and the angle information to obtain the coordinates of the edge points, and the edge point coordinates are fitted to obtain the edge function. In this embodiment, during the monitoring of infusion in the hospital, the height information of the liquid level of the infusion bottle is collected in real time, and the infusion volume and infusion rate within the measuring time are determined through the edge function and the height information of the liquid level of the infusion bottle; and the infusion volume and infusion rate are compared with the preset values. When the infusion volume or the infusion rate deviates from the preset value, an alarm is issued to remind the nurse to carry out subsequent dressing changes or needle removal work. This embodiment can monitor and warn the infusion volume and infusion rate in real time, and can improve the safety of infusion. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0023] Figure 1 A schematic diagram of a flow chart of a hospital infusion monitoring and alarm method based on millimeter wave radar provided in an embodiment of the present application;

[0024] Figure 2 This is a flow chart of step S4 in the embodiment of the present application;

[0025] Figure 3 This is a flow chart of step S6 in the embodiment of the present application;

[0026] Figure 4 It is a schematic diagram of the structure in which the gradient angle is discretized into four sectors of a circle in an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the technical solutions and advantages in the embodiments of the present application more clearly understood, the exemplary embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0028] Embodiment 1

[0029] like Figure 1 As shown, a hospital infusion monitoring and alarm method based on millimeter wave radar is provided in Embodiment 1 of the present application, comprising:

[0030] S1, the millimeter wave radar installed just above the longitudinal section of the infusion bottle slowly rotates 90° at a constant speed to emit a radar detection signal;

[0031] S2, collecting the echo signal of the millimeter-wave radar of the longitudinal section of the infusion bottle;

[0032] S3, denoising the echo signal through a Wiener filter;

[0033] S4, calculating the angle information in the echo signal after denoising;

[0034] S5, performing image reconstruction on the denoised echo signal to obtain an image array;

[0035] S6, determining edge points of the image array by using a Canny operator and a cross-correlation operation;

[0036] S7, combining the edge points of the image array with the electromagnetic wave velocity of the millimeter wave radar to obtain distance information of the edge points; combining the distance information of the edge points with the angle information in the echo signal to obtain coordinate information of the edge points; and fitting all discrete edge point coordinates to obtain an edge function;

[0037] S8, collecting height information of the liquid level of the infusion bottle through the FFT algorithm;

[0038] S9, determining the infusion volume and infusion rate within the measurement time through the edge function and the height information of the liquid level of the infusion bottle;

[0039] S10, comparing the infusion volume and infusion rate with the preset values, and when the infusion volume or infusion rate deviates from the preset values, an alarm is issued.

[0040] In this embodiment, a Wiener filter can be constructed by minimizing the difference between the output and the expected response, thereby filtering out the noise mixed in the original signal and enhancing the amplitude information of the echo signal.

[0041] In this embodiment, the echo information of the longitudinal section of the infusion bottle is collected, the noise mixed in the echo signal is filtered out by a Wiener filter, and the angle information in the denoised echo signal is calculated; at the same time, by removing the noise mixed in the echo signal, the amplitude of the echo signal can be enhanced and the edge information can be protected, so that the overall transition of the layer is smooth and natural; the denoised echo signal is used to determine the edge point of the image array through the Canny operator and cross-correlation operation; then, the coordinates of the edge point are obtained by combining the edge point and the angle information, and the edge point coordinates are fitted to obtain the edge function; in this embodiment, during the monitoring of infusion in the hospital, the height information of the liquid level of the infusion bottle is collected in real time, and the infusion volume and infusion rate within the metering time are determined by the edge function and the height information of the liquid level of the infusion bottle; and the infusion volume and infusion rate are compared with the preset value. When the infusion volume or the infusion rate deviates from the preset value, an alarm prompt is issued to remind the nurse to carry out subsequent dressing changes or needle removal work.

[0042] like Figure 2 As shown, the S4, calculating the angle information in the echo signal after denoising, comprises:

[0043] S41, constructing a sparse basis matrix equation based on the echo signal;

[0044] The expression of the sparse basis matrix equation is:

[0045] y(t)=Ax(t)+O(t)

[0046] x(t)=[x1,x2,···x N ] T

[0047]

[0048] In the above formula, y(t) represents the received signal at time t, x(t) represents the echo signal at time t; O(t) represents the noise at time t, represents the steering vector at each angle, and A represents the direction matrix;

[0049] Wherein, x(t)=[x1(t),x2(t),x3(t),...,x N (t)], n∈1,2···N, t∈1,2···T; x1(t) represents the first echo signal received at time t, and its angle value corresponds to x N (t) represents the Nth echo signal received at time t, and its angle value corresponds to

[0050] The angle values ​​of all echo signals are recorded as:

[0051] S42, solving the sparse basis matrix equation based on the multipath matching pursuit compressed sensing algorithm, and obtaining the estimated value α of the echo signal multipath parameter, where: α=[α0,α1,α2,···α n-1 ].

[0052] In this embodiment, y(t)=Ax(t)+O(t) is simplified to: y=Ax+O.

[0053] In this embodiment, the compressed sensing algorithm based on multipath matching pursuit solves the sparse basis matrix equation, takes the received signal y as the observation signal, and takes the direction matrix A as the sensing matrix Φ; the compressed sensing algorithm in this embodiment is to find K non-zero sparse positions in the N×1 sparse vector, that is, to obtain a combination of these K positions in the N optional positions; KN.

[0054] The compressed sensing algorithm in this implementation retains L index sets in each iteration, and each index set is expanded in the subsequent iteration process to generate L index sets, thereby generating multiple paths.

[0055] The compressed sensing algorithm based on multi-path matching pursuit generates a set, each element in the set corresponds to a possible optimal path, and each candidate support set corresponds to a residual; in the atomic matching stage, for a candidate support set Calculate residuals The correlation coefficient between the atoms is selected and the atoms with the largest correlation coefficient are added to the candidate support set. In the process, we get L new support sets; after K iterations, we will generate a set S consisting of several candidate support sets. K , select the one with the smallest residual as the final reconstructed support set.

[0056] The specific process may include:

[0057] S42-0, input the observation value y at time t, the sensing matrix Φ, the sparsity K and the multipath extension factor L;

[0058] S42-1: Before iteration, initialize the number of iterations, residual and support set;

[0059] Start the iteration, let i = j = 1, the number of support sets m = 0, and the support set factor S k =φ;

[0060] S42-2: Calculate the correlation between the residual and the perception matrix Φ corresponding to the i-th support set based on k-1 iterations, and the vector η corresponding to the L positions when it is the strongest.

[0061] S42-3: Calculate the i-th support set and η based on k-1 iterations j The union of

[0062] S42-4: If s tep Does not belong to the support set factor S k ; Then, i=i+1,

[0063] S42-5: Solving the set based on the least squares method get Residual margin

[0064] S42-6: j=j+1, if j≤L, return to S42-1, otherwise execute S42-7;

[0065] S42-7: i=i+1, if i≤|S k-1 |, return to S42-1, otherwise execute S42-8;

[0066] S42-8: k=k+1, if k<K, return to S42-1, otherwise execute S42-9;

[0067] S42-9: After the loop is completed, the support set corresponding to the minimum value of the support set residual margin is determined as the final selected support set s;

[0068] The support set s is the iterative estimate x of the original signal, and then we get The spatial spectrum P(n) in the direction:

[0069] P(n)=Σ|(x) n | 2 ,n=1,···N;

[0070] In this embodiment, the one-to-one correspondence between the signal and the angle is used to estimate the angle of the echo signal, and the angle of each column of data is obtained as α=[α0, α1, α2, ···α n-1 ].

[0071] like Figure 3 As shown, in S6, the edge points of the image array are determined by using the Canny operator, including:

[0072] S61, smoothing the image array based on the first-order derivative of the two-dimensional Gaussian function; specifically, S61, smoothing the image array based on the first-order derivative of the two-dimensional Gaussian function, includes:

[0073] Assume that the point coordinates of the image array are: (i, j), the size of the Gaussian filter kernel is (2k+1)×(2k+1), and k is an integer; then, the calculation expression for smoothing is:

[0074]

[0075] S62, based on the Sobel operator, convolution operations are performed on each pixel of the smoothed image array in the horizontal and vertical directions in turn to obtain the first-order partial derivatives f of the corresponding pixel points. x (i,j),f y (i, j); in the process of performing convolution operations on each pixel of the image array in the horizontal and vertical directions, the horizontal convolution operator is G x , the vertical convolution operator is G y ; Its calculation expression is:

[0076]

[0077] S63, based on the first-order partial derivative f of the pixel position x (i,j),f y (i, j), calculate the gradient magnitude M(i, j) and gradient direction θ(i, j) of the pixel; the calculation expression is:

[0078]

[0079] The gradient direction is perpendicular to the edge direction;

[0080] S64, performing non-maximum suppression on the gradient amplitude M(i,j) to obtain a processed image array E1; specifically comprising:

[0081] S64-1, using the Canny operator to discretize the gradient angle into one of four sectors of a circle, where the four sectors are numbered from 0 to 3;

[0082] like Figure 4 As shown, in this implementation, the gradient is divided into 8 directions, namely E, NE, N, NW, W, SW, S, SE, where 0 represents 0 to 45 degrees, 1 represents 45 to 90 degrees, 2 represents -90 to 45 degrees, and 3 represents -45 to 0 degrees;

[0083] S64-2, using a 3×3 window for suppression operation; including:

[0084] When the gradient amplitude M of the central pixel of the 3×3 template p Greater than the gradient magnitude M of two locations with the same label p1 and M p2 When , the gradient magnitude M of the central pixel of the 3×3 template is retained pAs the local maximum gradient value; otherwise, suppress the 3×3 template center pixel M p The gradient amplitude is 0, and the gradient linear interpolation of pixel points P1 and P2 is:

[0085] M p1 =(1-tan(β))×E+tan(β)×NE

[0086] M p2 =(1-tan(β))×W+tan(β)×SW;

[0087] Among them, E, W, SW, and NE are abbreviations of directions, representing east, west, southwest, and northeast respectively; β is the angle between the pixel point P1 and the direction E;

[0088] S65, suppressing pseudo edge pixels in the image array E1 by a double threshold detection method, and retaining strong edge pixels with high gradient values; specifically including:

[0089] S65-1, set the high threshold to Th1 and the low threshold to Th2;

[0090] S65-2, according to the gradient amplitude M of the current edge pixel p The relationship between these two thresholds determines the strength of the edge; specifically:

[0091] If the gradient amplitude M of the current edge pixel p ≥Th1, the current edge pixel is marked as a strong edge and retained;

[0092] If the gradient amplitude of the current edge pixel Th2≤M p <Th1, the current edge pixel is marked as a weak edge; if the weak edge is adjacent to a strong edge pixel, the point is retained; if the weak edge is not adjacent to a strong edge pixel, the point is suppressed;

[0093] If the gradient amplitude M of the current edge pixel p <Th2, the current edge pixel is suppressed.

[0094] In this embodiment, since there are still some pseudo-edge pixels caused by noise, bubbles and other factors remaining in the image array after non-maximum suppression processing, the double threshold detection method can suppress the pseudo-edge pixels and retain the strong edge pixels with high gradient values.

[0095] S66, performing a cross-correlation operation on the column data of the undetected edge points in the image array E2 to obtain the image array E3 and the seed point of each column data of the image array E3; including:

[0096] S66-1, taking the point where the strong edge of the image array E2 is located as the initial seed point, and determining the width n of the search window in the column where the initial seed point is located with the initial seed point as the center, the n-point data vector X=[x0,x1,x2,···x n-1 ];

[0097] S66-2, determine the range of the sliding window 2n+1 with the same point in the next column of the initial seed point as the center, and the 2n+1 points constitute a new data vector Y=[y0,y1,y2,...y 2n ];

[0098] S66-3, completing the cross-correlation operation of XY, and determining the window position corresponding to the maximum cross-correlation value of XY during the sliding process, and determining it as the seed point of the current column data;

[0099] S66-4, starting from the initial seed point, cross-correlation operations are performed in sequence to the left and right sides until the cross-correlation operations of all columns of data on both sides are completed, thereby obtaining the seed point of each column of data in the image array E3.

[0100] S67, combining the strong edge pixel points of the image array E2 with the seed points of the image array E3 to obtain complete edge points of the image array E4.

[0101] In this embodiment, since the adjacent sequence events of the echo signals of each dielectric layer in the infusion bottle have gradual amplitude variation and continuity in waveform, based on this feature, the cross-correlation analysis can realize automatic tracking and detection of edge information.

[0102] Assume that x(n) and y(n) are two discrete random sequences with a sampling interval of T. s , then the cross-correlation expression is:

[0103]

[0104] In fact, the echo signal collected by the millimeter-wave radar is a stationary random process, and the previous statistical average can be replaced by the time average. Therefore, the cross-correlation function expression of the discrete time series x(n) and y(n) is:

[0105]

[0106] Here, m represents the delay, and k=-N, -N+1, ...0, 1, ...N.

[0107] In this embodiment, the edge points of the image array are calculated by using the Canny operator and the cross-correlation operation. The two algorithms can obtain complete edge point information in a complementary form, provide data support for the construction of the edge function, make the fitting accuracy of the edge function higher, and thus improve the calculation accuracy of the infusion volume.

[0108] In this embodiment, the step S9, determining the infusion volume and infusion rate within the measurement time through the edge function and the height information of the liquid level of the infusion bottle, includes:

[0109] S9-1, calculate the initial solution volume of the infusion bottle; the calculation expression is:

[0110]

[0111] Wherein, V0 represents the initial solution volume of the infusion bottle, x1 represents the horizontal coordinate corresponding to the edge function of the initial solution of the infusion bottle, F(x,y) is the edge function, and x and y represent the horizontal coordinate and vertical coordinate of each edge point respectively;

[0112] S9-2, after the infusion starts, collect the real-time liquid level h of the infusion bottle, determine the infusion volume V and infusion rate v within the time T; the calculation expression is:

[0113]

[0114] Among them, x h Indicates the corresponding horizontal coordinate value when the vertical coordinate of the edge function is h.

[0115] In this embodiment, the liquid level height of the infusion bottle can be collected in real time through the FFT algorithm.

[0116] In this embodiment, the volume change of the liquid in the infusion bottle is obtained by integrating the marginal function, and then the real-time infusion speed is obtained. In the specific implementation process, the bed information and the infusion volume V can be input into the infusion monitoring and early warning system of the nurse station before the infusion starts. th , infusion rate v th When the real-time infusion volume or infusion rate deviates from the preset value, it indicates that the infusion is abnormal, such as compression of the infusion tube or distortion of the infusion needle. The alarm is used to remind the nurse to carry out subsequent dressing changes or needle removal.

[0117] The present application also provides an electronic device, including:

[0118] memory; processor; and computer program;

[0119] The computer program is stored in the memory and is configured to be executed by the processor to implement the method as described above.

[0120] An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored; the computer program is executed by a processor to implement the method described above.

[0121] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0125] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0126] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A hospital infusion monitoring and alarm method based on millimeter wave radar, characterized in that: include: S1, the millimeter wave radar installed just above the longitudinal section of the infusion bottle slowly rotates 90° at a constant speed to emit a radar detection signal; S2, collecting the echo signal of the millimeter-wave radar of the longitudinal section of the infusion bottle; S3, denoising the echo signal through a Wiener filter; S4, calculating the angle information in the echo signal after denoising; S5, performing image reconstruction on the denoised echo signal to obtain an image array; S6, determining edge points of the image array by using a Canny operator and a cross-correlation operation; S7, combining the edge points of the image array with the electromagnetic wave velocity of the millimeter wave radar to obtain distance information of the edge points; combining the distance information of the edge points with the angle information in the echo signal to obtain coordinate information of the edge points; and fitting all discrete edge point coordinates to obtain an edge function; S8, collecting height information of the liquid level of the infusion bottle through the FFT algorithm; S9, determining the infusion volume and infusion rate within the measurement time through the edge function and the height information of the liquid level of the infusion bottle; S10, comparing the infusion volume and infusion rate with the preset values, and when the infusion volume or infusion rate deviates from the preset values, an alarm is issued.

2. A hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 1, characterized in that: The step S4, calculating the angle information in the denoised echo signal, comprises: S41, constructing a sparse basis matrix equation according to the echo signal; The expression of the sparse basis matrix equation is: y(t)=Ax(t)+O(t) x(t)=[x1,x2,···x N ] T A=[a(θ1),a(θ2),···a(θ N )] T In the above formula, y(t) represents the received signal at time t, x(t) represents the echo signal at time t; O(t) represents the noise at time t, a(θ n ) represents the steering vector at each angle, and A represents the direction matrix; Wherein, x(t)=[x1(t),x2(t),x3(t),...,x N (t)], n∈1,2···N, t∈1,2···T; x1(t) represents the first echo signal received at time t, and its angle value corresponds to θ1; x N (t) represents the Nth echo signal received at time t, and its angle value corresponds to θ N ; The angle values ​​of all echo signals are recorded as: θ=[θ1,θ2,θ3,···θ N ]; S42, solving the sparse basis matrix equation based on the compressed sensing algorithm of multipath matching pursuit, and obtaining the estimated value α of the multipath parameter of the echo signal.

3. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 2 is characterized in that: In S6, edge points of the image array are determined by using a Canny operator and a cross-correlation operation, including: S61, smoothing the image array based on the first-order derivative of the two-dimensional Gaussian function; S62, based on the Sobel operator, convolution operations are performed on each pixel of the smoothed image array in the horizontal and vertical directions, and the first-order partial derivatives f of the corresponding pixel are obtained respectively. x (i,j),f y (i,j); S63, based on the first-order partial derivative of the pixel position f x (i,j),f y (i, j), calculate the gradient magnitude M(i, j) and gradient direction θ(i, j) of the pixel; the calculation expression is: The gradient direction is perpendicular to the edge direction; S64, performing non-maximum suppression on the gradient amplitude M(i,j) to obtain a processed image array E1; S65, suppressing pseudo edge pixels in the image array E1 by a double threshold detection method, and retaining strong edge pixels with high gradient values, to obtain an image array E2; S66, performing a cross-correlation operation on the column data of the undetected edge points in the image array E2 to obtain the image array E3 and the seed point of each column data of the image array E3; S67, combining the strong edge pixel points of the image array E2 with the seed points of the image array E3 to obtain complete edge points of the image array E4.

4. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 3 is characterized in that: The step S64, performing non-maximum suppression on the gradient amplitude M(i,j), includes: S64-1, using the Canny operator to discretize the gradient angle into one of four sectors of a circle, where the four sectors are numbered from 0 to 3; S64-2, using a 3×3 window for suppression operation; including: When the gradient magnitude M of the central pixel of the 3×3 template p Greater than the gradient magnitude M of two locations with the same label p1 and M p2 When , the gradient magnitude M of the central pixel of the 3×3 template is retained p As the local maximum gradient value; otherwise, suppress the 3×3 template center pixel M p The gradient amplitude is 0, and the gradient linear interpolation of pixel points P1 and P2 is: M p1 (1-tan(β))×E+tan(β)×NE M p2 (1-tan(β))×W+tan(β)×SW Among them, E, W, SW, and NE are abbreviations of directions, representing east, west, southwest, and northeast respectively; β is the angle between the pixel point P1 and the direction E.

5. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 3 is characterized in that: The S65 includes: S65-1, set the high threshold to Th1 and the low threshold to Th2; S65-2, according to the gradient amplitude M of the current edge pixel p The relationship between these two thresholds determines the strength of the edge; specifically: If the gradient amplitude M of the current edge pixel p ≥Th1, the current edge pixel is marked as a strong edge and retained; If the gradient amplitude of the current edge pixel Th2≤M p <Th1, the current edge pixel is marked as a weak edge; if the weak edge is adjacent to a strong edge pixel, the point is retained; if the weak edge is not adjacent to a strong edge pixel, the point is suppressed; If the gradient amplitude M of the current edge pixel p <Th2, the current edge pixel is suppressed.

6. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 5 is characterized in that: S66, performing a cross-correlation operation on the column data of the undetected edge points in the image array E2 to obtain the image array E3 and the seed point of each column data of the image array E3; including: S66-1, taking the point where the strong edge of the image array E2 is located as the initial seed point, and determining the width n of the search window in the column where the initial seed point is located with the initial seed point as the center, the n-point data vector X=[x0,x1,x2,···x n-1 ]; S66-2, determine the range of the sliding window 2n+1 with the same point in the next column of the initial seed point as the center, and the 2n+1 points constitute a new data vector Y=[y0,y1,y2,...y 2n ]; S66-3, completing the cross-correlation operation of XY, and determining the window position corresponding to the maximum cross-correlation value of XY during the sliding process, and determining it as the seed point of the current column data; S66-4, starting from the initial seed point, cross-correlation operations are performed in sequence to the left and right sides until the cross-correlation operations of all columns of data on both sides are completed, thereby obtaining the seed point of each column of data in the image array E3.

7. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 2 is characterized in that: S9, determining the infusion volume and infusion rate within the measurement time through the edge function and the height information of the liquid level of the infusion bottle; include: S9-1, calculate the initial solution volume of the infusion bottle; the calculation expression is: Wherein, V0 represents the initial solution volume of the infusion bottle, x1 represents the horizontal coordinate corresponding to the edge function of the initial solution of the infusion bottle, F(x,y) is the edge function, and x and y represent the horizontal coordinate and vertical coordinate of each edge point respectively; S9-2, after the infusion starts, collect the real-time liquid level h of the infusion bottle, determine the infusion volume V and infusion rate v within the time T; the calculation expression is: Among them, x h Indicates the corresponding horizontal coordinate value when the vertical coordinate of the edge function is h.

8. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 2 is characterized in that: The step S61, performing smoothing processing on the image array based on the first-order derivative of the two-dimensional Gaussian function, comprises: Assume that the point coordinates of the image array E are: (i, j), the size of the Gaussian filter kernel is (2k+1)×(2k+1), and k is an integer; then, the calculation expression for the smoothing process is:

9. The hospital infusion monitoring and alarm method based on millimeter wave radar according to claim 2 is characterized in that: In the process of performing convolution operations on each pixel point of the image array in the horizontal and vertical directions in S62, the horizontal direction convolution operator is G x , the vertical convolution operator is G y ; Its calculation expression is:

10. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the method according to any one of claims 1 to 9.

Images