Intensive care unit video visiting management method and system based on image acquisition

Abstract

The invention discloses an intensive care unit video visiting management method and system based on image acquisition, and relates to the technical field of video visiting, the intensive care unit video visiting management method based on image acquisition comprises the following steps: acquiring an intensive care unit image in real time, and preprocessing the intensive care unit image; identifying a patient action image by using an intensive care unit image identification method, and analyzing whether the patient actions are the same or not; judging the state of the patient according to an analysis result, adjusting visiting times of the medical staff, and generating a visiting management table; and storing the generated visiting management table to a video platform, and updating in real time. According to the method, whether the actions of the patient are the same or not is analyzed, the visiting times of the medical personnel are flexibly adjusted, and resource allocation is performed according to the stability of the state of the patient, so that the optimization of visiting management is facilitated, the manpower resource and time cost is effectively reduced, the working efficiency is improved, and the medical personnel are ensured to more accurately pay attention to the needs of the patient.

Description

Technical Field

[0001] The present invention relates to the technical field of video visiting, and in particular to a method and system for managing video visiting in an intensive care unit based on image acquisition. Background Art

[0002] The intensive care unit is equipped with advanced equipment and professional medical staff. It is a special place dedicated to treating critically ill patients. Since the critically ill patients in the intensive care unit have complex conditions and low immunity, they are prone to become a high-risk group for hospital infection. Excessive flow of people will increase the chance of cross-infection among patients. For this reason, the hospital prohibits visiting the room, and can only visit through the glass window. It also adopts a semi-restricted room visiting system, stipulates the number of people and time for visiting the room, and requires visitors to wear isolation clothing, shoes, hats and masks.

[0003] However, this visiting method not only wastes the time of medical staff, but also wastes disposable consumables and cannot guarantee a clean environment indoors. In order to obtain patient information in a timely manner, meet patients' visiting needs, and prevent nosocomial infections, many hospitals have launched video visiting projects through information technology. Through video visiting, medical staff can conveniently visit patients through video, which not only achieves the purpose of obtaining patient information in a timely manner, but also effectively prevents the risk of cross infection in the hospital.

[0004] In the existing ICU video visits, due to the large number of ICUs and patients, a visit management table needs to be developed according to the characteristics of each patient's condition and treatment needs. However, there are human factors in this process, which may lead to errors or omissions in information entry. Due to the different conditions of patients, the development of different visit management tables will make management complicated, especially when constant updates and adjustments are required, which is prone to confusion and inconsistency. This may lead to uneven resource allocation. Some patients may receive too much attention and visits, while some patients may be neglected or have a low visit frequency, thus affecting the rational use of medical resources.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes a video visiting management method and system for an intensive care unit based on image acquisition to overcome the above-mentioned technical problems existing in the existing related art.

[0007] To this end, the specific technical solution adopted by the present invention is as follows: According to one aspect of the present invention, a method for managing video visitation in an intensive care unit based on image acquisition is provided. The method for managing video visitation in an intensive care unit based on image acquisition comprises the following steps: S1, real-time acquisition of intensive care unit images and preprocessing of intensive care unit images; S2, using the intensive care unit image recognition method to identify the patient's action images and analyze whether the patient's actions are the same; S3. Determine the patient's status based on the analysis results, adjust the number of visits by medical staff, and generate a visit management form; S4. Store the generated visitation management form on the video platform and update it in real time.

[0008] Furthermore, real-time acquisition of intensive care unit images and preprocessing of the intensive care unit images include the following steps: S11, using sensors to collect images of the intensive care unit in real time; S12, graying the collected image using a mean filtering method; S13, performing histogram equalization processing on the grayscale processed image; S14. Use Gaussian filtering technology to perform noise reduction on the image after histogram equalization processing.

[0009] Furthermore, the construction of the ICU image library includes the following steps: Collect several analyzed images from the intensive care unit and use feature extraction to obtain the environmental features and patient motion features of the images; The intensive care unit labeling method is used to classify the categories of environmental features and patient motion features, and a database based on the environment and patient motion is constructed to obtain the intensive care unit image library, which is also updated regularly.

[0010] Furthermore, using the intensive care unit image recognition method to identify the patient's action image and analyzing whether the patient's actions are the same includes the following steps: S21, extracting a patient motion image by using a patient motion detection algorithm according to the preprocessed intensive care unit image; S22, obtaining historical images from the constructed intensive care unit image library, preprocessing the historical images, and extracting historical action images through a patient action detection algorithm; S23. Compare the patient's motion image with the historical motion image to analyze whether the patient's motion is the same.

[0011] Further, according to the preprocessed ICU image, extracting the patient motion image by using the patient motion detection algorithm comprises the following steps: S211, using a two-dimensional generalized Laplace Gaussian filter to process the intensive care unit image to obtain an image action template; S212, performing convolution calculation on the intensive care unit image and the image action template, and dividing the calculation result into a plurality of grids; S213, selecting the maximum value from a number of grids as a feature point, and calculating the direction; S214, voting for the pixel points located in the angle above the feature point, and determining the position of the vanishing point according to the voting result; S215, determining a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter, and acquiring the patient's motion contour within the selection interval; S216. Extracting a corresponding patient motion image from the intensive care unit image according to the acquired patient motion contour.

[0012] Further, determining the selection interval of the patient's motion contour according to the position of the vanishing point and the preset range parameter, and obtaining the patient's motion contour within the selection interval includes the following steps: S2151, determining a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter; S2152, generating an initial action profile group within the selected interval, and initializing the action profile group; S2153, constructing a complete action contour independently from each action contour by using edge detection; S2154: If all the action profile groups are constructed, the action profiles of the patient are obtained; if all the action profile groups are not constructed, step S2153 is continued until all the action profile groups are constructed.

[0013] Furthermore, comparing the patient's motion image with the historical motion image to analyze whether the patient's motion is the same includes the following steps: S231, comparing the patient's action image with the historical action image by using an improved similarity algorithm to obtain a similarity value; S232, analyzing whether the patient action image and the historical patient image represent the same action based on the obtained similarity value and a preset similarity threshold; S233. If the similarity value is greater than the similarity threshold, the patient action image and the historical patient image represent the same action; if the similarity value is less than or equal to the similarity threshold, the patient action image and the historical patient image represent different actions.

[0014] Furthermore, by comparing the patient's action image with the historical action image through an improved similarity algorithm, a similarity value calculation formula in the similarity value is obtained, which includes the following steps: ; In the formula, T It is expressed as the similarity value between the patient action image and the historical patient image; n It is expressed as the number of patient action images; k It is expressed as the number of historical patient images; Expressed as i Patient action images in historical patient images k The weight of Expressed as j Patient action images in historical patient images k The weight of .

[0015] Furthermore, judging the patient's status according to the analysis results, adjusting the number of visits by medical staff, and generating a visit management form include the following steps: S31. If the analysis results indicate the same action, the patient's state is determined to be stable, and the number of visits by medical staff is reduced. If the analysis results indicate different actions, the patient's state is determined to be abnormal, and the number of visits by medical staff is increased. S32. Combine the analysis results with the work arrangements of medical staff to generate a visit management form.

[0016] Furthermore, combining the analysis results with the work arrangements of medical staff, generating a visit management form includes the following steps: S321. Based on the analysis results, assess the patient's visitation needs; S322. Obtaining medical staff scheduling information according to the medical staff's work arrangements; S323, generating an initial visitation management form in combination with visitation demand and medical staff scheduling information, and calculating the objective function value of the initial visitation management form in combination with a preset objective function; S324, setting the initial temperature according to the temperature drop strategy, randomly adjusting the visitation arrangement based on the current temperature to generate a new visitation management table, and calculating the objective function value of the new visitation management table; S325, calculating the objective function variance of the objective function value of the initial visitation management table and the objective function value of the new visitation management table; S326. If the variance of the objective function is less than zero, the new solution is accepted as the current solution. If the variance of the objective function is greater than zero, the new solution is accepted with a preset probability. S327. Repeat steps S324 to S325 until a preset threshold is reached, and generate a visitation management table.

[0017] According to another aspect of the present invention, there is also provided an intensive care unit video visit management system based on image acquisition, the intensive care unit video visit management system based on image acquisition comprises an image collection and processing module, a patient action recognition module, a visit management table generation module and a visit management table storage module; The image collection and processing module, the patient action recognition module, the visit management form generation module and the visit management form storage module are connected in sequence; An image collection and processing module is used to collect intensive care unit images in real time and perform preprocessing on the intensive care unit images; Patient action recognition module, used to identify patient action images using intensive care unit image recognition method and analyze whether the patient actions are the same; The visit management form generation module is used to determine the patient's status based on the analysis results, adjust the number of visits by medical staff, and generate a visit management form; The visitation management table storage module is used to store the generated visitation management table to the video platform and update it in real time.

[0018] The beneficial effects of the present invention are: 1. The present invention analyzes whether the patient's movements are the same, flexibly adjusts the number of visits by medical staff, and allocates resources according to the stability of the patient's condition, which helps to optimize visit management, reduce unnecessary visit frequency, and effectively reduce human resources and time costs. It not only improves work efficiency, but also ensures that medical staff pay more attention to the needs of patients and realizes the intelligent use of resources.

[0019] 2. By generating a visiting management form, the present invention can organically combine the patient's visiting needs and the work arrangements of medical staff, thereby effectively organizing and managing visiting tasks, which not only ensures the timely completion and recording of visiting tasks, but also provides a clear framework so that medical staff can better plan and execute visiting work and improve the efficiency and quality of medical work.

[0020] 3. The present invention can realize the sharing and updating of visit management information by storing the generated visit management form on the video platform and updating it in real time, so that medical staff can understand the patient's visit plan and historical record at any time, thereby improving the collaboration efficiency and information sharing level; at the same time, through the storage and update function of the video platform, the reliability and accessibility of the information are ensured, providing a convenient information management platform for the medical team, which helps to better coordinate work and improve the quality and efficiency of medical services. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1is a flow chart of a method for managing video visitation in an intensive care unit based on image acquisition according to an embodiment of the present invention; Figure 2 The invention is a block diagram of a video visit management system for an intensive care unit based on image acquisition according to an embodiment of the invention.

[0023] In the figure: 1. Image collection and processing module; 2. Patient action recognition module; 3. Visiting management form generation module; 4. Visiting management form storage module. DETAILED DESCRIPTION

[0024] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0025] According to an embodiment of the present invention, a method and system for managing video visitation in an intensive care unit based on image acquisition are provided.

[0026] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an intensive care unit video visit management method based on image acquisition according to an embodiment of the present invention, the intensive care unit video visit management method based on image acquisition includes the following steps: S1. Acquire intensive care unit images in real time and preprocess the intensive care unit images.

[0027] The real-time acquisition of intensive care unit images and preprocessing of the intensive care unit images include the following steps: S11, using sensors to collect images of the intensive care unit in real time; S12, graying the collected image using a mean filtering method; It should be noted that the mean filtering method replaces the grayscale value of the target pixel with the average grayscale value of the pixels in the neighborhood around the target pixel, thereby achieving image smoothing. In the mean filtering method, the collected image data is separated into three color channels: red, green, and blue. The color channel values ​​at each pixel position are averaged to obtain the grayscale value of the pixel. The mean filter is applied to each pixel of the image, that is, the grayscale values ​​of the pixel and its surrounding pixels are averaged.

[0028] S13, performing histogram equalization processing on the grayscale processed image; It should be noted that histogram equalization is a commonly used image enhancement technology used to improve the contrast and brightness distribution of an image. By redistributing the gray levels of an image, the gray levels in the image can be more evenly distributed throughout the gray range.

[0029] S14. Use Gaussian filtering technology to perform noise reduction on the image after histogram equalization processing.

[0030] It should be noted that a weight matrix is ​​generated using a Gaussian function, and then the weight matrix is ​​applied to each pixel of the image to smooth the image and reduce noise.

[0031] The construction of the intensive care unit image library includes the following steps: Collect several analyzed images from the intensive care unit and use feature extraction to obtain the environmental features and patient motion features of the images; The intensive care unit labeling method is used to classify the categories of environmental features and patient motion features, and a database based on the environment and patient motion is constructed to obtain the intensive care unit image library, which is also updated regularly.

[0032] S2. Use the intensive care unit image recognition method to identify the patient's action images and analyze whether the patient's actions are the same.

[0033] Among them, using the intensive care unit image recognition method to identify the patient's action image and analyze whether the patient's actions are the same includes the following steps: S21. Extracting a patient motion image through a patient motion detection algorithm based on the preprocessed intensive care unit image.

[0034] The method of extracting the patient motion image by using the patient motion detection algorithm according to the preprocessed intensive care unit image includes the following steps: S211. Use a two-dimensional generalized Laplace Gaussian filter to process the intensive care unit image to obtain an image action template.

[0035] It should be noted that the two-dimensional generalized Laplace Gaussian filter is defined as: ; In the formula, Z represents the function value of the two-dimensional generalized Laplace of Gaussian filter; represents the parameters controlling the two-dimensional generalized Laplace of Gaussian filter; G Represents the function value of a two-dimensional generalized Gaussian filter; x Represents the shape value of the Gaussian kernel; yRepresents the direction value of the Gaussian kernel.

[0036] S212, performing convolution calculation on the intensive care unit image and the image action template, and dividing the calculation result into a plurality of grids.

[0037] It should be noted that the convolution calculation result is obtained by sliding the image action template on the ICU image and multiplying it element-wise with each part of the ICU image, and then adding the results.

[0038] In addition, it should be noted that convolution calculations are performed on the intensive care unit image and 18 groups of image action templates respectively, and the size and shape of the grid are defined, and then the calculation result intensive care unit image is divided according to preset rules, where each grid represents the convolution result of a specific area in the image.

[0039] S213. Select the maximum value in several grids as a feature point and calculate the direction.

[0040] It should be added that the maximum value is selected as the feature point in several grids. The feature point represents the strongest point of the intensive care unit image in the sub-grid area. These points are used as feature points and the direction is calculated. The largest group of results represents the texture direction of the feature point.

[0041] S214: Voting for the pixels located in the angle above the feature point, and determining the position of the vanishing point according to the voting result.

[0042] It should be noted that the specific formula for voting for pixels within a certain angle range above the feature point is: ; In the formula, T represents the function value of voting; exp represents the expected value of the vote; d(p,v) Yes means p and v The Euclidean distance between p The numerical value representing the feature point; v Indicates the value of a pixel; Represents the angle between the feature point and the pixel point; Indicates the length of the diagonal line; Indicates a preset value.

[0043] S215 . Determine a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter, and obtain the patient's motion contour within the selection interval.

[0044] The process of determining the selection interval of the patient's motion contour according to the position of the vanishing point and the preset range parameters, and obtaining the patient's motion contour within the selection interval includes the following steps: S2151, determining a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter; It should be noted that the preset range parameters include the width and height of the selected interval and the relative position relationship with the vanishing point.

[0045] S2152, generating an initial action profile group within the selected interval, and initializing the action profile group; It should be noted that each action contour is identified and classified within the selected interval; the identified and classified action contours are used as the initial action contour group; the initialized action contour group is screened and optimized to remove inappropriate contours and retain contours that may represent valid actions; and attributes such as position, size, movement direction, etc. are initialized for each action contour group to facilitate subsequent action analysis and tracking.

[0046] S2153, constructing a complete action contour independently from each action contour by using edge detection; It should be noted that obtaining the patient's motion contour within the selected interval includes using the Canny edge detection algorithm to detect the contour within the selected interval. The contour is usually an obvious edge or connected area in the intensive care unit image; filtering and screening the detected contour according to the shape, size, area and other characteristics of the contour, removing the contours that do not meet the conditions, and retaining the contours that may represent the patient's motion; performing contour smoothing optimization on the retained contour to obtain a more accurate and complete patient motion contour; and extracting the main contour representing the patient's motion based on the optimized contour.

[0047] S2154: If all the action profile groups are constructed, the action profiles of the patient are obtained; if all the action profile groups are not constructed, step S2153 is continued until all the action profile groups are constructed.

[0048] It should be noted that the construction of the motion profile group can be regarded as an optimization problem, and obtaining the patient's motion profile is the ultimate goal of this problem; through continuous iteration and optimization, when all the motion profile groups are constructed, it means that the optimal solution that meets the conditions has been found, that is, the patient's motion profile.

[0049] S216. Extracting a corresponding patient motion image from the intensive care unit image according to the acquired patient motion contour.

[0050] It should be noted that, based on the acquired patient motion contour, the position of the patient motion contour in the intensive care unit image is determined; the shape of the patient motion contour is matched to ensure that the shape of the patient motion contour has a high degree of matching with the corresponding area in the intensive care unit image; based on the matched patient motion contour, the corresponding patient motion image is extracted in the intensive care unit image; and the extracted patient motion image is corrected and processed.

[0051] S22. Acquire historical images from the constructed intensive care unit image library, preprocess the historical images, and extract historical motion images through a patient motion detection algorithm.

[0052] It should be noted that historical images are obtained from the constructed intensive care unit image library, and grayscale, denoising, edge enhancement and other operations are performed on the historical images; a two-dimensional generalized Laplace Gaussian filter is used to process the historical images to obtain historical image action templates; convolution calculations are performed on the historical images and the historical image action templates, and the calculation results are divided into several grids; the maximum value is selected as the feature point in several grids, and the direction is calculated; the pixel points located within the angle above the feature point are voted, and the position of the vanishing point is determined according to the voting results; according to the position of the vanishing point and the pre-set range parameters, the selection interval of the historical action contour is determined, and the historical action contour is obtained within the selection interval; according to the obtained historical action contour, the corresponding historical action image is extracted from the historical image.

[0053] S23. Compare the patient's motion image with the historical motion image to analyze whether the patient's motion is the same.

[0054] Comparing the patient's motion image with the historical motion image and analyzing whether the patient's motion is the same includes the following steps: S231, comparing the patient's action image with the historical action image by using an improved similarity algorithm to obtain a similarity value; Among them, comparing the patient action image with the historical action image through the improved similarity algorithm, and obtaining the similarity value calculation formula in the similarity value includes the following steps: ; In the formula, T It is expressed as the similarity value between the patient's motion image and the historical motion image; n It is expressed as the number of patient action images; k It is expressed as the number of historical action images; Expressed as iPatient motion images in historical motion images k The weight of Expressed as j Patient motion images in historical motion images k The weight of .

[0055] S232, analyzing whether the patient action image and the historical patient image represent the same action based on the obtained similarity value and a preset similarity threshold; It should be noted that the similarity threshold is a boundary value used to determine the degree of similarity between two images. It is set according to the application scenario and needs of the intensive care unit. It can be a fixed value or dynamically adjusted according to actual conditions.

[0056] In addition, it should be noted that the selection of the similarity threshold needs to take into account the error tolerance of the video visitation management method, the requirements for accuracy, and the characteristics of the actual data; if the similarity threshold is set too low, it may lead to misjudgment; and if it is set too high, some relevant information may be missed; therefore, experiments and tuning are usually required to find the similarity threshold that best suits the intensive care unit application scenario.

[0057] S233. If the similarity value is greater than the similarity threshold, the patient action image and the historical patient image represent the same action; if the similarity value is less than or equal to the similarity threshold, the patient action image and the historical patient image represent different actions.

[0058] S3. Determine the patient's condition based on the analysis results, adjust the number of visits by medical staff, and generate a visit management form.

[0059] Among them, judging the patient's status according to the analysis results, adjusting the number of visits by medical staff, and generating a visit management form include the following steps: S31. If the analysis results indicate the same action, the patient's state is determined to be stable, and the number of visits by medical staff is reduced. If the analysis results indicate different actions, the patient's state is determined to be abnormal, and the number of visits by medical staff is increased. It should be noted that if the analysis results indicate the same action, it means that the patient's condition is relatively stable and does not require frequent visits from medical staff. The effective use of medical resources can be improved by reducing the visiting frequency, while ensuring that the patient is properly monitored and cared for in a relatively good condition. If the analysis results are different actions, it means that the patient's condition may be abnormal and requires more frequent visits from medical staff to promptly discover and deal with any potential health problems. Increasing the number of visits can help provide timely medical intervention and support to ensure that patients receive the best care.

[0060] S32. Combine the analysis results with the work arrangements of medical staff to generate a visit management form.

[0061] Among them, combining the analysis results with the work arrangements of medical staff, generating a visit management form includes the following steps: S321. Based on the analysis results, evaluate the patient's visiting needs.

[0062] It should be added that, based on the analysis results, the visiting needs of each patient were assessed, and patients who needed frequent visits and those in a stable state were distinguished.

[0063] S322. Obtain the medical staff's shift information according to the medical staff's work arrangements.

[0064] S323. Generate an initial visit management form based on visit demand and medical staff scheduling information, and calculate the objective function value of the initial visit management form based on a preset objective function.

[0065] It should be noted that the objective function formula is: In the formula, f represents the objective function; ω 1 represents non-negative weights; D A measure of visitation need satisfaction, with higher values ​​indicating better need satisfaction; ω 2 represents non-negative weights; W A measure of the workload of medical staff, with lower values ​​indicating a more balanced workload; ω 3 represents non-negative weights; M A measure of the degree of match between visiting hours and medical staff schedules, with higher values ​​indicating a better match; ω 4 represents non-negative weights; R A measure of the adequacy of rest time for medical staff. The higher the value, the more sufficient the rest time.

[0066] also, ω 1 , ω 2 , ω 3 , ω 4 are non-negative weights, their sum is usually 1, used to balance the relative importance of different factors.

[0067] S324. Set the initial temperature according to the temperature drop strategy, randomly adjust the visitation arrangement based on the current temperature to generate a new visitation management table, and calculate the objective function value of the new visitation management table.

[0068] S325, calculating the objective function variance of the objective function value of the initial visitation management table and the objective function value of the new visitation management table.

[0069] S326. If the variance of the objective function is less than zero, the new solution is accepted as the current solution. If the variance of the objective function is greater than zero, the new solution is accepted with a preset probability.

[0070] It should be noted that the default probability is e −Δf / T , where Δf represents the variance of the objective function and T represents the current temperature.

[0071] S327. Repeat steps S324 to S325 until a preset threshold is reached, and generate a visitation management table.

[0072] It should be noted that appropriate visiting time periods should be determined based on the patient's visiting needs and the work arrangements of medical staff; a visiting management form should be drawn up based on the determined visiting time periods and patient information; the visiting management form should include the patient's name, visiting date, visiting time period, number of visits, visiting medical staff information, etc., to facilitate the hospital's registration and management of visits.

[0073] S4. Store the generated visitation management form on the video platform and update it in real time.

[0074] It should be noted that storing the generated visitation management form on the video platform includes the following steps: Step 1: Use lossless compression to compress the generated visitation management form.

[0075] The visitation management table is decomposed into blocks of different directions using color wavelet transform, and wavelet coefficients are obtained through them.

[0076] In addition, the visitation management table is divided into red blocks and black blocks according to the horizontal and vertical intersections, and into blue blocks and yellow blocks according to the diagonal intersections; the black block is predicted by the four neighboring red blocks in the horizontal and vertical directions to obtain the predicted value, and the actual value of the black block is replaced by the difference between the actual value of the black block and the predicted value to obtain the horizontal and vertical wavelet coefficients; the yellow block is predicted by the four diagonal neighboring blue blocks to obtain the predicted value, and the actual value of the yellow block is replaced by the difference between the actual value of the yellow block and the predicted value to obtain the diagonal wavelet coefficient; the wavelet coefficients in the horizontal and vertical directions and the diagonal direction are combined to obtain the wavelet coefficients.

[0077] Find the maximum value among the wavelet coefficients and use coefficient normalization to normalize the wavelet coefficients to the preset range.

[0078] In addition, the wavelet coefficients are traversed to find the maximum value and normalized. The normalized calculation formula is: ; In the formula, is represented by the normalized wavelet coefficients, X is represented as the original wavelet coefficients, X min is expressed as the minimum value of the wavelet coefficients, X max It is expressed as the maximum value of the wavelet coefficients.

[0079] It should be noted that the preset interval is [-1,1].

[0080] The wavelet coefficients normalized to the preset interval are linearly expanded to obtain unsigned integer wavelet coefficients.

[0081] In addition, multiply each coefficient by LH max =2 len-1 Then round off and convert the normalized wavelet coefficients into unsigned integers for further processing; on the one hand, integers are easy to encode; on the other hand, after the data is expanded, the truncation error caused by the limitation of the number of encoding bits can be reduced; at the same time, len The compression ratio can be controlled by the quantization step size.

[0082] The data encoding algorithm is used to compress the unsigned integer wavelet coefficients.

[0083] In addition, the wavelet coefficients of unsigned integers are scanned, the frequency of occurrence of each unsigned integer wavelet coefficient is counted, and a data tree is constructed according to its frequency; based on the data tree, a data coding table is generated using a recursive algorithm; combined with the data coding table, the number of occurrences of the wavelet coefficients of the unsigned integers is counted; based on the number of occurrences of the wavelet coefficients of the unsigned integers, the optimal code is selected, and encoding compression is performed based on the optimal code.

[0084] Step 2: Use security protocols and encryption technology to transmit the compressed visitation management form to the video platform.

[0085] In addition, the security protocol refers to a series of specifications and protocols used to ensure the security of the visitation management form during the transmission of video visits and visitation management forms. The security protocol involves encryption, authentication, integrity protection and other aspects of the visitation management form to ensure the security and privacy of communications; encryption technology is an important part of the field of information security and is used to protect the confidentiality and security of the visitation management form. Encryption technology converts the original visitation management form into ciphertext so that only people with the corresponding decryption key can interpret and access the data content; security protocols and encryption technology play a key role in network communications and visitation management form transmission. They are used to establish a secure connection between sensors and video platforms to protect the confidentiality and integrity of visitation management forms.

[0086] Step 3: Perform a decompression operation on the video platform to restore the compressed visitation management form received.

[0087] In addition, the compressed visitation management form is decompressed using a corresponding decompression tool (such as WinRAR, 7-Zip, etc.); the decompressed visitation management form usually contains multiple data files or data sequences, which are processed and restored through a programming language (such as the OpenCV library in Python).

[0088] Step 4: Establish a visitation management table storage mechanism to store the decompressed visitation management table.

[0089] In addition, it is necessary to determine the requirements such as the stored data attributes, data volume, storage period, etc.; according to the storage requirements, establish a visitation management table storage mechanism in the video platform to organize and manage data; store the decompressed visitation management table in the video platform and store it according to the visitation management table storage mechanism.

[0090] like Figure 2 As shown, according to another embodiment of the present invention, there is also provided an intensive care unit video visit management system based on image acquisition, the intensive care unit video visit management system based on image acquisition comprises an image collection and processing module 1, a patient action recognition module 2, a visit management table generation module 3 and a visit management table storage module 4; The image collection and processing module 1, the patient action recognition module 2, the visit management table generation module 3 and the visit management table storage module 4 are connected in sequence; The image collection and processing module 1 is used to collect intensive care unit images in real time and perform preprocessing on the intensive care unit images; The patient action recognition module 2 is used to recognize the patient action image using the intensive care unit image recognition method and analyze whether the patient actions are the same; The visit management form generation module 3 is used to determine the patient's status according to the analysis results, adjust the number of visits by medical staff, and generate a visit management form; The visitation management table storage module 4 is used to store the generated visitation management table in the video platform and update it in real time.

[0091] In summary, with the help of the above technical solution of the present invention, by analyzing whether the patient's actions are the same, the number of visits of medical staff can be flexibly adjusted, and resources can be allocated according to the stability of the patient's state, which is helpful to optimize the visit management, reduce unnecessary visit frequency, effectively reduce human resources and time costs, not only improve work efficiency, but also ensure that medical staff pay more attention to the needs of patients, and realize the intelligent use of resources. By generating a visit management form, the patient's visit needs and the work arrangements of medical staff can be organically combined, so as to effectively organize and manage the visit tasks, which not only ensures the timely completion and recording of the visit tasks, but also provides a clear framework so that medical staff can better plan and perform visit work and improve the efficiency and quality of medical work. By storing the generated visit management form on the video platform and updating it in real time, the sharing and updating of visit management information can be realized, so that medical staff can understand the patient's visit plan and historical records at any time, thereby improving the collaboration efficiency and information sharing level; at the same time, through the storage and update function of the video platform, the reliability and accessibility of the information are ensured, and a convenient information management platform is provided for the medical team, which helps to better coordinate work and improve the quality and efficiency of medical services.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A video visit management method for intensive care units based on image acquisition, characterized in that: The intensive care unit video visit management method based on image acquisition comprises the following steps: S1, real-time acquisition of intensive care unit images and preprocessing of intensive care unit images; S2, using the intensive care unit image recognition method to identify the patient's action images and analyze whether the patient's actions are the same; S3. Determine the patient's status based on the analysis results, adjust the number of visits by medical staff, and generate a visit management form; S4. Store the generated visitation management form in the video platform and update it in real time; The method of using the intensive care unit image recognition method to identify the patient's action images and analyzing whether the patient's actions are the same includes the following steps: S21, extracting a patient motion image by using a patient motion detection algorithm according to the preprocessed intensive care unit image; S22, obtaining historical images from the constructed intensive care unit image library, preprocessing the historical images, and extracting historical action images through a patient action detection algorithm; S23, comparing the patient's motion image with the historical motion image, and analyzing whether the patient's motion is the same; The method of extracting the patient motion image by using the patient motion detection algorithm according to the preprocessed ICU image comprises the following steps: S211, using a two-dimensional generalized Laplace Gaussian filter to process the intensive care unit image to obtain an image action template; S212, performing convolution calculation on the intensive care unit image and the image action template, and dividing the calculation result into a plurality of grids; S213, selecting the maximum value from a number of grids as a feature point, and calculating the direction; S214, voting for the pixel points located in the angle above the feature point, and determining the position of the vanishing point according to the voting result; S215, determining a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter, and acquiring the patient's motion contour within the selection interval; S216. Extracting a corresponding patient motion image from the intensive care unit image according to the acquired patient motion contour.

2. The method for managing intensive care unit video visits based on image acquisition according to claim 1, characterized in that: The real-time acquisition of intensive care unit images and preprocessing of the intensive care unit images include the following steps: S11, using sensors to collect images of the intensive care unit in real time; S12, using the mean filtering method to perform grayscale processing on the collected image; S13, performing histogram equalization processing on the grayscale processed image; S14. Use Gaussian filtering technology to perform noise reduction on the image after histogram equalization processing.

3. The method for managing intensive care unit video visits based on image acquisition according to claim 1, characterized in that: The construction of the intensive care unit image library includes the following steps: Collect several analyzed images from the intensive care unit and use feature extraction to obtain the environmental features and patient motion features of the images; The intensive care unit labeling method is used to classify the categories of environmental features and patient motion features, and a database based on the environment and patient motion is constructed to obtain the intensive care unit image library, which is also updated regularly.

4. The method for managing intensive care unit video visits based on image acquisition according to claim 1, characterized in that: Determining the selection interval of the patient's motion contour according to the position of the vanishing point and the preset range parameters, and obtaining the patient's motion contour within the selection interval comprises the following steps: S2151, determining a selection interval of the patient's motion contour according to the position of the vanishing point and a preset range parameter; S2152, generating an initial action profile group within the selected interval, and initializing the action profile group; S2153, constructing a complete action contour independently from each action contour by using edge detection; S2154: If all the action profile groups are constructed, the action profiles of the patient are obtained; if all the action profile groups are not constructed, step S2153 is continued until all the action profile groups are constructed.

5. The method for managing intensive care unit video visits based on image acquisition according to claim 1, characterized in that: The comparing of the patient's motion image with the historical motion image to analyze whether the patient's motion is the same comprises the following steps: S231, comparing the patient's action image with the historical action image by using an improved similarity algorithm to obtain a similarity value; S232, analyzing whether the patient action image and the historical patient image represent the same action based on the obtained similarity value and a preset similarity threshold; S233. If the similarity value is greater than the similarity threshold, the patient action image and the historical patient image represent the same action; if the similarity value is less than or equal to the similarity threshold, the patient action image and the historical patient image represent different actions.

6. The method for managing intensive care unit video visits based on image acquisition according to claim 5, characterized in that: The method of comparing the patient's action image with the historical action image by using the improved similarity algorithm to obtain a similarity value calculation formula includes the following steps: ； In the formula, T It is expressed as the similarity value between the patient action image and the historical patient image; n It is expressed as the number of patient action images; k It is expressed as the number of historical patient images; Expressed as i Patient action images in historical patient images k The weight of Expressed as j Patient action images in historical patient images k The weight of .

7. The method for managing intensive care unit video visits based on image acquisition according to claim 1, characterized in that: The method of judging the patient's status according to the analysis results, adjusting the number of visits by medical staff, and generating a visit management form includes the following steps: S31. If the analysis results indicate the same action, the patient's state is determined to be stable, and the number of visits by medical staff is reduced. If the analysis results indicate different actions, the patient's state is determined to be abnormal, and the number of visits by medical staff is increased. S32. Combine the analysis results with the work arrangements of medical staff to generate a visit management form.

8. The method for managing intensive care unit video visits based on image acquisition according to claim 7, characterized in that: The generation of a visit management form by combining the analysis results with the work arrangement of medical staff includes the following steps: S321. Based on the analysis results, assess the patient's visitation needs; S322. Obtaining medical staff scheduling information according to the work arrangements of medical staff; S323, generating an initial visitation management form in combination with visitation demand and medical staff scheduling information, and calculating the objective function value of the initial visitation management form in combination with a preset objective function; S324, setting the initial temperature according to the temperature drop strategy, randomly adjusting the visitation arrangement based on the current temperature to generate a new visitation management table, and calculating the objective function value of the new visitation management table; S325, calculating the objective function variance of the objective function value of the initial visitation management table and the objective function value of the new visitation management table; S326. If the variance of the objective function is less than zero, the new solution is accepted as the current solution. If the variance of the objective function is greater than zero, the new solution is accepted with a preset probability. S327. Repeat steps S324 to S325 until a preset threshold is reached, and generate a visitation management table.

9. An intensive care unit video visit management system based on image acquisition, used to implement the intensive care unit video visit management method based on image acquisition according to any one of claims 1 to 8, characterized in that: The intensive care unit video visit management system based on image acquisition includes an image collection and processing module, a patient action recognition module, a visit management form generation module and a visit management form storage module; The image collection and processing module, the patient action recognition module, the visit management table generation module and the visit management table storage module are connected in sequence; The image collection and processing module is used to collect intensive care unit images in real time and perform preprocessing on the intensive care unit images; The patient motion recognition module is used to recognize patient motion images using an intensive care unit image recognition method and analyze whether the patient motions are the same; The visit management form generation module is used to determine the patient status according to the analysis results, adjust the number of visits by medical staff, and generate a visit management form; The visitation management table storage module is used to store the generated visitation management table in the video platform and update it in real time.

Images