Emergency intensive care unit nursing management system and method based on cloud computing

Abstract

The invention discloses an emergency intensive care unit nursing management system and method based on cloud computing, and relates to the field of nursing management, and the emergency intensive care unit nursing management system comprises a management and control module which is used for editing and executing operation instructions of all function modules; the acquisition unit is used for acquiring personal information of a current patient and acquiring physique data, action data and audio data; the preprocessing module is used for performing cleaning, standardization and key feature extraction on the acquired data; the model construction module is used for constructing an anomaly prediction model through a random forest algorithm, and performing model training by taking historical acquisition data as a training sample; by integrating multi-dimensional data and utilizing an anomaly prediction model to automatically learn and adapt to the characteristics of different patients, intelligent anomaly detection is performed based on currently acquired physical data, actions and sounds of the patients, and the configuration of medical personnel and medical equipment is automatically optimized according to the anomaly levels of the patients.

Description

Technical Field

[0001] The present invention relates to the field of nursing management technology, and in particular to an emergency intensive care unit nursing management system and method based on cloud computing. Background Art

[0002] The rapid development of cloud computing technology, especially its advantages in data storage, processing and analysis, has enabled efficient and convenient management of medical data. IoT devices such as sensors, monitoring instruments and mobile devices are becoming increasingly popular in hospital clinical environments, allowing real-time collection of important physiological and environmental data. As the workload of nurses and doctors increases, the medical industry faces a shortage of human resources, which requires more intelligent management methods to improve work efficiency, especially in emergency and ICU environments, where patient safety is the most important task; Traditional nursing management systems usually rely on manual monitoring or timed data collection, which may miss sudden changes in health status and lack comprehensive monitoring and analysis of patients' physical data, movements and sounds. Existing technologies often have data silos, with information scattered across different devices and systems and lacking effective integration, resulting in a lack of comprehensive information support for medical staff when making decisions. It is difficult to quickly and effectively predict abnormal patient conditions, which often relies on the subjective judgment of doctors. Medical staff may face the dilemma of complicated information and difficulty in making decisions quickly, which may cause information delays and medical errors in high-pressure emergency environments. In traditional emergency care, the allocation of medical resources often relies on the experience of medical staff, and it is impossible to quickly and effectively evaluate patient priorities, which can easily lead to key resources failing to reach the patients who need them most in time. Summary of the invention

[0003] 1. Technical issues to be resolved In view of the above-mentioned shortcomings of the prior art, the present invention provides an emergency intensive care unit nursing management system and method based on cloud computing, which can effectively solve the problems of the prior art.

[0004] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention discloses a cloud computing-based emergency intensive care unit nursing management system, comprising: The control module is used to edit and execute the operation instructions of each functional module; The collection unit is used to obtain the personal information of the current patient, and collect physical data, motion data and audio data; The preprocessing module is used to clean, standardize and extract key features of the collected data; The model building module is used to build an anomaly prediction model using the random forest algorithm and train the model using historically collected data as training samples; The data analysis module is used to receive the abnormal prediction model trained by the model building module, input the newly extracted key features into the abnormal prediction model, and calculate the abnormal coefficient of the patient in real time; A determination module is used to obtain the abnormal coefficient output by the data analysis module, compare it based on a number of preset health status thresholds, and determine the patient's current abnormal level; The scheduling module is used to evaluate the priority of required medical resources according to the abnormality level of the patient and generate personnel scheduling instructions and medical equipment scheduling instructions; The feedback module is used to obtain follow-up treatment data based on the scheduling results of the scheduling module, generate regular medical reports, and evaluate the effectiveness of personnel scheduling instructions and medical equipment scheduling instructions as a reference for adjusting the model building module.

[0005] Furthermore, the management and control module is interactively connected to a data storage module via a wireless network, and the data storage module is used to establish a data warehouse on the cloud platform, store patient information from different data sources, and set different access rights for different users.

[0006] Furthermore, the acquisition unit is deployed with submodules at the lower level, and the submodules include: a physical acquisition module, a motion acquisition module and a voice acquisition module, wherein: Physical fitness collection module, used to obtain the patient's heart rate, blood pressure and blood oxygen saturation physical fitness data in real time; The motion acquisition module is used to capture the patient's body movements, body position changes and facial expressions through cameras and sensors to form a motion database; The voice acquisition module records the patient's voice and tone changes through a recording device and stores them as audio data.

[0007] Furthermore, the determination module is interactively connected to a threshold module via a wireless network, and the threshold module is used to obtain historical patient data as a reference, set different abnormal coefficients and corresponding health status thresholds, and submit them to the determination module for configuration.

[0008] Furthermore, the determination module is interactively connected to an alarm module via a wireless network, and the alarm module is used to automatically trigger an alarm message to be submitted to the management and control module when it is found that the abnormal coefficient of the patient exceeds a set threshold, and to locally store the triggered alarm message.

[0009] Furthermore, the attributes of the medical report generated by the feedback module include: patient status changes, abnormal event records, medical staff information, medical equipment information and treatment results.

[0010] Furthermore, the control module is interactively connected with the collection unit, the preprocessing module, the model building module, the data analysis module, the determination module, the scheduling module and the feedback module via a wireless network.

[0011] The cloud computing-based emergency intensive care unit nursing management method includes the following steps: Step 1: Connect to the patient's monitoring device, obtain access rights, capture the patient's basic physical data, capture the patient's body movements, body position changes and facial expressions, perform real-time analysis on the collected videos and images, extract key action data, capture the patient's voice, apply natural language processing technology, transcribe and perform sentiment analysis on the patient's voice, and identify potential pain or discomfort signals; Step 2: Clean and standardize the different types of data obtained in step 1, extract the heart rate change and blood pressure fluctuation characteristics from the physical data, extract the movement frequency and duration characteristics from the movement data, and extract the tone change and sound intensity characteristics from the sound data; Step 3: Build an abnormality prediction model through the random forest algorithm, use historical patient data to train the model, convert the current features extracted in step 2 into an input format acceptable to the abnormality prediction model, and the abnormality prediction model calculates and outputs the patient abnormality coefficient of the current cycle based on the input features; Step 4: Define the range of abnormal coefficients and their corresponding health status, obtain the abnormal coefficients of patients in the current cycle, and determine whether there is an abnormality in the current cycle; Step 5: According to the historical abnormality coefficient, the patient's abnormal conditions are divided into different levels, and corresponding response personnel scheduling strategies and resource allocation plans are formulated for each level. According to the current abnormality coefficient and judgment results, the corresponding personnel scheduling strategies and resource allocation plans are matched; Step 6: Develop and apply a unique treatment plan based on the matched personnel scheduling strategy, resource allocation plan, and patient specific conditions; Step 7: Obtain the trend of the patient's condition after applying the treatment plan, evaluate the effectiveness of the treatment plan, and adjust the abnormality prediction model based on the evaluation results.

[0012] Furthermore, the calculation formula of the abnormal prediction model in step 3 is: ; In the formula, Represents the abnormal coefficient of the current patient, Represents the total number of decision trees in a random forest. Usually, a random forest uses multiple trees to improve the robustness and accuracy of the model. The feature vector representing the patient is composed of several key features, including physical data, motion data, and sound features. Represents the nth decision tree, which is a basic classifier in the entire random forest model. Each tree evaluates the input features through a series of logical judgments. Representative decision tree For the feature vector The predicted output of , represents the tree's evaluation of the patient's status, Represents the bias term, which is used to correct the comprehensive value of the decision tree output to the preset abnormal coefficient range. , able to control the baseline level of output.

[0013] (III) Beneficial effects Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: 1. By acquiring the patient's physical data, movements and sounds in real time, providing timely health status updates, integrating information from different sensors and data sources, and integrating multi-dimensional data, it can identify and analyze potential problems in the patient's condition, improve the accuracy of diagnosis, and build an abnormal prediction model to automatically learn and adapt to the characteristics of different patients, so as to perform intelligent abnormality detection based on the currently collected patient physical data, movements and sounds.

[0014] 2. By automatically optimizing the configuration of medical personnel and equipment according to the patient's abnormal level and urgency, it ensures that key resources reach the corresponding patients when they are most needed, thereby reducing resource waste, improving the work efficiency of medical staff, and ensuring that critically ill patients receive timely and appropriate care. It can also formulate personalized medical plans for patients based on their specific conditions and combined with historical data to improve treatment effects.

[0015] 3. Through automated monitoring and alarm mechanisms, potential critical situations can be discovered and responded to in a timely manner, significantly improving patient safety, ensuring that every patient receives timely and appropriate care, reducing the incidence of complications and adverse events, and reducing waste of medical resources and unnecessary treatments through real-time monitoring, optimizing resource allocation, and improving treatment outcomes, thereby reducing overall medical costs. By analyzing data to generate visual reports, scientific basis can be provided for clinical decision-making by medical staff, reducing the risk of subjective judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1It is a schematic diagram of the framework of the emergency intensive care unit nursing management system based on cloud computing in the present invention; Figure 2 It is a flowchart of the emergency intensive care unit nursing management method based on cloud computing in the present invention.

[0018] The numbers in the figure represent respectively, 1. control module; 2. acquisition unit; 21. physical acquisition module; 22. motion acquisition module; 23. voice acquisition module; 3. preprocessing module; 4. model building module; 5. data analysis module; 6. judgment module; 7. scheduling module; 8. feedback module; 9. data storage module; 10. threshold module; 11. alarm module. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0020] The present invention will be further described below in conjunction with the embodiments.

[0021] Example 1: The emergency intensive care unit nursing management system based on cloud computing in this example is as follows: Figure 1 As shown, including: The control module 1 is used to edit and execute the operation instructions of each functional module; the control module 1 is interactively connected to the data storage module 9 through a wireless network, and the data storage module 9 is used to establish a data warehouse on the cloud platform, store patient information from different data sources, and set different access rights for different users; The acquisition unit 2 is used to obtain the personal information of the current patient, collect physical data, motion data and audio data; by obtaining the physical data and other relevant data of the patient in real time, it provides basic data for a comprehensive assessment of the patient's health status; The collection unit 2 is deployed with submodules, including: a physical collection module 21, a motion collection module 22 and a voice collection module 23, wherein: The physical condition acquisition module 21 is used to obtain the patient's heart rate, blood pressure and blood oxygen saturation physical condition data in real time; The motion collection module 22 is used to capture the patient's body movements, body position changes and facial expressions through cameras and sensors to form a motion database; The voice collection module 23 records the patient's voice and tone changes through a recording device and stores them as audio data; Preprocessing module 3 is used to clean, standardize and extract key features of the collected data; cleaning and feature extraction of the collected data improves the data quality and ensures the accuracy of subsequent analysis; Model building module 4 is used to build an abnormal prediction model through random forest algorithm and train the model by using historical collected data as training samples; build an abnormal prediction model through random forest algorithm and train it by using historical data to enable the system to have strong learning and prediction capabilities; The data analysis module 5 is used to receive the abnormal prediction model trained by the model building module 4, input the newly extracted key features into the abnormal prediction model, and calculate the abnormal coefficient of the patient in real time; the real-time calculation of the abnormal coefficient can quickly identify the abnormal changes of the patient and improve the response speed; The determination module 6 is used to obtain the abnormal coefficient output by the data analysis module 5, compare it based on a number of preset health status thresholds, and determine the current abnormal level of the patient; the determination module 6 is interactively connected to the threshold module 10 through a wireless network, and the threshold module 10 is used to obtain historical patient data as a reference, set different abnormal coefficients and corresponding health status thresholds, and submit them to the determination module 6 for configuration; by comparing and determining the abnormal coefficients, the patient's risk level can be accurately identified according to the health status threshold, which is convenient for timely medical intervention. In addition, it is connected to the historical data reference to enhance the scientific nature of the judgment; The determination module 6 is interactively connected to the alarm module 11 via a wireless network. The alarm module 11 is used to automatically trigger an alarm message to be submitted to the control module 1 when the abnormal coefficient of the patient exceeds a set threshold value, and the triggered alarm message is stored locally; Scheduling module 7 is used to evaluate the priority of medical resources required according to the abnormality level of the patient, and generate personnel scheduling instructions and medical equipment scheduling instructions; according to the abnormality level of the patient, the priority of medical resources is evaluated, and medical staff and equipment are reasonably scheduled to improve the efficiency of medical resource utilization; The feedback module 8 is used to obtain follow-up treatment data based on the scheduling results of the scheduling module 7, generate regular medical reports, and evaluate the effectiveness of personnel scheduling instructions and medical equipment scheduling instructions. As a reference for adjusting the model building module 4, the attributes of the medical report generated by the feedback module 8 include: patient status changes, abnormal event records, medical staff information, medical equipment information and treatment results. By regularly generating medical reports to record patient status changes and treatment effects, it helps with follow-up care and resource optimization, and provides data support for medical decision-making.

[0022] As a preferred implementation in this embodiment, Figure 1As shown, the control module 1 is interactively connected with the collection unit 2, the preprocessing module 3, the model building module 4, the data analysis module 5, the determination module 6, the scheduling module 7 and the feedback module 8 through a wireless network.

[0023] In the specific implementation of this embodiment, the overall control of the global function module is performed through the control module 1, all the collected data and analysis data are stored through the data storage module 9, the patient's physical data is collected through the physical acquisition module 21, the patient's action state data is collected through the action acquisition module 22, and the patient's voice data is collected through the voice acquisition module 23. The data collected by the physical acquisition module 21, the action acquisition module 22 and the voice acquisition module 23 are pre-processed by the pre-processing module 3, and key features are extracted. The abnormal prediction model is constructed by the model construction module 4, the key features are input, the abnormal coefficient is output, and the abnormal coefficient is judged by the judgment module 6; When the determination module 6 determines that there is an abnormality, the specific abnormality level of the abnormality coefficient is classified, and the scheduling module 7 retrieves the medical personnel and medical resources corresponding to the abnormal level, analyzes the priority, formulates and applies the scheduling plan. When the determination module 6 determines that there is an abnormality, the alarm module 11 is triggered to alarm the control module 1. The feedback module 8 evaluates the implementation results of the current scheduling plan and adjusts the parameters of the model building module 4. The threshold module 10 sets the threshold for the judgment standard of the determination module 6. This embodiment obtains the patient's physical data, movements and sounds in real time to provide timely health status updates, integrates information from different sensors and data sources, and can identify and analyze potential problems in the patient's status by integrating multi-dimensional data, thereby improving the accuracy of diagnosis. The abnormal prediction model can automatically learn and adapt to the characteristics of different patients, thereby realizing intelligent abnormality detection based on the currently collected patient's physical data, movements and sounds, and automatically optimizing the configuration of medical personnel and equipment according to the patient's abnormality level and urgency, ensuring that key resources reach the corresponding patient at the most needed time, thereby reducing resource waste, improving the work efficiency of medical staff, and ensuring that critically ill patients receive timely and appropriate care. According to the patient's specific condition, combined with historical data, a personalized medical plan can be formulated for the patient to improve the treatment effect; Through automated monitoring and alarm mechanisms, potential critical situations can be discovered and responded to in a timely manner, significantly improving patient safety, ensuring that every patient receives timely and appropriate care, and reducing the incidence of complications and adverse events. Through real-time monitoring, optimized resource allocation, and improved treatment outcomes, it can reduce waste of medical resources and unnecessary treatments, and overall reduce medical costs. By analyzing data to generate visual reports, it can provide a scientific basis for medical staff's clinical decision-making and reduce the risk of subjective judgment.

[0024] Embodiment 2: In other aspects, this embodiment also provides a cloud computing-based emergency intensive care unit nursing management method, such as Figure 2 As shown, the following steps are included: Step 1: Connect to the patient's monitoring device, obtain access rights, capture the patient's basic physical data, capture the patient's body movements, body position changes and facial expressions, perform real-time analysis on the collected videos and images, extract key action data, capture the patient's voice, apply natural language processing technology, transcribe and perform sentiment analysis on the patient's voice, and identify potential pain or discomfort signals; Step 2: Clean and standardize the different types of data obtained in step 1, extract the heart rate change and blood pressure fluctuation characteristics from the physical data, extract the movement frequency and duration characteristics from the movement data, and extract the tone change and sound intensity characteristics from the sound data; Step 3: Build an abnormality prediction model through the random forest algorithm, use historical patient data to train the model, convert the current features extracted in step 2 into an input format acceptable to the abnormality prediction model, and the abnormality prediction model calculates and outputs the patient abnormality coefficient of the current cycle based on the input features; Step 4: Define the range of abnormal coefficients and their corresponding health status, obtain the abnormal coefficients of patients in the current cycle, and determine whether there is an abnormality in the current cycle; Step 5: According to the historical abnormality coefficient, the patient's abnormal conditions are divided into different levels, and corresponding response personnel scheduling strategies and resource allocation plans are formulated for each level. According to the current abnormality coefficient and judgment results, the corresponding personnel scheduling strategies and resource allocation plans are matched; Step 6: Develop and apply a unique treatment plan based on the matched personnel scheduling strategy, resource allocation plan, and patient specific conditions; Step 7: Obtain the trend of the patient's condition after applying the treatment plan, evaluate the effectiveness of the treatment plan, and adjust the abnormality prediction model based on the evaluation results.

[0025] Compared with existing technologies, multi-source data collection methods can be used to achieve comprehensive monitoring of patients' real-time status, improve the grasp of patients' health status, and use random forest algorithm for abnormal prediction, so that abnormal changes in patients' status can be warned in advance, providing support for clinical decision-making. It can efficiently dispatch medical resources, optimize the configuration of medical staff and equipment, improve treatment efficiency and resource utilization, and regularly generate medical reports, which not only help the medical team summarize the treatment effect, but also serve as a reference for subsequent optimization of the system.

[0026] Embodiment 3: In this embodiment, a calculation formula of an abnormal prediction model is provided, which is specifically: ; In the formula, Represents the abnormal coefficient of the current patient, Represents the total number of decision trees in a random forest. Usually, a random forest uses multiple trees to improve the robustness and accuracy of the model. The feature vector representing the patient is composed of several key features, including physical data such as heart rate and blood pressure, motion data such as limb movement frequency, and sound features such as voice changes. Represents the nth decision tree, which is a basic classifier in the entire random forest model. Each tree evaluates the input features through a series of logical judgments. Representing a decision tree For the feature vector The predicted output of , represents the tree's evaluation of the patient's status, Represents the bias term, which is used to correct the comprehensive value of the decision tree output to the preset abnormal coefficient range. , able to control the baseline level of output; The working logic of the calculation formula is: obtain the latest patient characteristics from real-time monitoring and form the patient's feature vector, input the new feature vector into the abnormality prediction model constructed by random forest, and for each tree, calculate the predicted value through the tree path, use the formula to average the output of all trees, and add the bias term to get the final abnormality coefficient; This embodiment makes the output of the abnormal prediction model more in line with the needs of specific application scenarios. Through the above complex and detailed formula, the patient's abnormal coefficient can be calculated in real time through the trained random forest model and the newly extracted key features, providing strong support for clinical decision-making. This real-time computing capability can effectively improve the emergency response capabilities of first aid and critical care, thereby helping to improve the treatment effect of patients.

[0027] In summary, the cloud computing-based method realizes real-time data collection and analysis, can monitor the patient's physical data, movements and voices in a timely manner, improves the response speed to changes in the patient's health status, realizes the aggregation and unified management of multi-source data, and can more comprehensively analyze the patient's health status. By building an abnormal prediction model and combining the patient's movement, voice and vital sign data, it realizes automatic and real-time abnormal prediction. By deeply analyzing the patient's physical, voice and behavior data, it not only optimizes real-time monitoring, but also provides support for subsequent improvements and research by mining the potential patterns of historical data, thereby improving the accuracy and timeliness of judgment. By clarifying the abnormal level of the abnormal coefficient output by the abnormal prediction model, it helps to prioritize medical personnel and equipment, thereby making more efficient use of medical resources. Through data analysis and report generation, it provides scientific and personalized medical solutions to assist medical staff in making more effective clinical decisions.

[0028] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. Emergency intensive care unit nursing management system based on cloud computing, characterized by: include: The control module (1) is used to edit and execute the operation instructions of each functional module; A collection unit (2) is used to obtain personal information of the current patient, and collect physical data, motion data and audio data; A preprocessing module (3) is used to clean, standardize and extract key features of the collected data; A model building module (4) is used to build an abnormal prediction model using a random forest algorithm and train the model using historical collected data as training samples; The data analysis module (5) is used to receive the abnormality prediction model trained by the model building module (4), input the newly extracted key features into the abnormality prediction model, and calculate the abnormality coefficient of the patient in real time; A determination module (6) is used to obtain the abnormal coefficient output by the data analysis module (5), compare it based on a number of preset health status thresholds, and determine the patient's current abnormal level; A scheduling module (7), used to evaluate the priority of required medical resources according to the abnormality level of the patient, and generate personnel scheduling instructions and medical equipment scheduling instructions; The feedback module (8) is used to obtain follow-up treatment data based on the scheduling results of the scheduling module (7), generate regular medical reports, and evaluate the effectiveness of personnel scheduling instructions and medical equipment scheduling instructions as a reference for adjustment of the model building module (4).

2. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The control module (1) is interactively connected to a data storage module (9) via a wireless network. The data storage module (9) is used to establish a data warehouse on a cloud platform, store patient information from different data sources, and set different access rights for different users.

3. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The acquisition unit (2) is provided with submodules at the lower level, and the submodules include: a physical condition acquisition module (21), an action acquisition module (22) and a voice acquisition module (23), wherein: A physical condition acquisition module (21) is used to obtain the patient's heart rate, blood pressure and blood oxygen saturation physical condition data in real time; A motion acquisition module (22) is used to capture the patient's body movements, body position changes and facial expressions through a camera and a sensor to form a motion database; The voice collection module (23) records the patient's voice and tone changes through a recording device and stores them as audio data.

4. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The determination module (6) is interactively connected to a threshold module (10) via a wireless network. The threshold module (10) is used to obtain historical patient data as a reference, set different abnormal coefficients and corresponding health status thresholds, and submit them to the determination module (6) for configuration.

5. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The determination module (6) is interactively connected to the alarm module (11) via a wireless network. The alarm module (11) is used to automatically trigger alarm information to be submitted to the control module (1) when it is found that the abnormal coefficient of the patient exceeds a set threshold, and to locally store the triggered alarm information.

6. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The attributes of the medical report generated by the feedback module (8) include: patient status changes, abnormal event records, medical staff information, medical equipment information and treatment results.

7. The cloud computing-based emergency intensive care unit nursing management system according to claim 1, characterized in that: The control module (1) is interactively connected with the collection unit (2), the preprocessing module (3), the model building module (4), the data analysis module (5), the determination module (6), the scheduling module (7) and the feedback module (8) via a wireless network.

8. A cloud computing-based emergency intensive care unit nursing management method, the method being an implementation method of a cloud computing-based emergency intensive care unit nursing management system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Connect to the patient's monitoring device, obtain access rights, capture the patient's basic physical data, capture the patient's body movements, body position changes and facial expressions, perform real-time analysis on the collected videos and images, extract key action data, capture the patient's voice, apply natural language processing technology, transcribe and perform sentiment analysis on the patient's voice, and identify potential pain or discomfort signals; Step 2: Clean and standardize the different types of data obtained in step 1, extract the heart rate change and blood pressure fluctuation characteristics from the physical data, extract the movement frequency and duration characteristics from the movement data, and extract the tone change and sound intensity characteristics from the sound data; Step 3: Build an abnormality prediction model through the random forest algorithm, use historical patient data to train the model, convert the current features extracted in step 2 into an input format acceptable to the abnormality prediction model, and the abnormality prediction model calculates and outputs the patient abnormality coefficient of the current cycle based on the input features; Step 4: Define the range of abnormal coefficients and their corresponding health status, obtain the abnormal coefficients of patients in the current cycle, and determine whether there is an abnormality in the current cycle; Step 5: According to the historical abnormality coefficient, the patient's abnormal conditions are divided into different levels, and corresponding response personnel scheduling strategies and resource allocation plans are formulated for each level. According to the current abnormality coefficient and judgment results, the corresponding personnel scheduling strategies and resource allocation plans are matched; Step 6: Develop and apply a unique treatment plan based on the matched personnel scheduling strategy, resource allocation plan, and patient specific conditions; Step 7: Obtain the trend of the patient's condition after applying the treatment plan, evaluate the effectiveness of the treatment plan, and adjust the abnormality prediction model based on the evaluation results.

9. The cloud computing-based emergency intensive care unit nursing management method according to claim 8, characterized in that: The calculation formula of the abnormal prediction model in step 3 is: ； In the formula, Represents the abnormal coefficient of the current patient, represents the total number of decision trees in the random forest, The feature vector representing the patient is composed of several key features. Represents the nth decision tree, which is a basic classifier in the entire random forest model. Representative decision tree For the feature vector The predicted output is Represents the bias term, which is used to correct the comprehensive value output by the decision tree to the preset abnormal coefficient range.

Images