Emergency patient information data matching method and system based on Internet

Abstract

The invention discloses an emergency patient information data matching method and system based on the Internet, and the method comprises the steps: obtaining multi-source heterogeneous data in real time according to an electronic medical record, image data and real-time vital sign data of an emergency patient, and generating a time-synchronized multi-source data set through a self-adaptive data format conversion algorithm; for the multi-source data set, using a feature extraction network to extract illness state features of patients, and generating patient priority scores; adjusting a resource allocation scheme in real time by adopting a medical resource matching algorithm based on a graph neural network according to the patient priority score and the illness state feature, and generating a preliminary medical resource matching result; and for a medical resource matching result, generating a final emergency patient information data matching scheme through a differential privacy protection mechanism by adopting a data security transmission protocol based on homomorphic encryption. According to the embodiment of the invention, the method can meet the requirements of quick, safe and effective processing of data in an emergency scene, and improves the quality and efficiency of emergency medical service.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data matching, and in particular to an emergency patient information data matching method and system based on the Internet. BACKGROUND

[0002] In modern medicine, timely treatment of emergency patients is crucial, but current emergency patient information management faces many challenges, including data source diversity, format heterogeneity, high real-time requirements, and security issues. Emergency patient information usually comes from electronic medical records, image data, and real-time vital sign data. These data formats are different, and traditional management methods are difficult to efficiently integrate, affecting the speed of real-time information acquisition and processing. At the same time, the method of evaluating patient condition characteristics and generating priority scores lacks flexibility, and cannot accurately reflect the severity of the condition and the urgency of treatment, leading to inaccurate treatment decisions. In addition, medical resource matching often lacks dynamic adjustment capability, and cannot timely optimize resource allocation according to the patient's emergency level and real-time situation, affecting treatment efficiency. At the same time, the sensitivity of patient information makes secure transmission in the Internet environment a major challenge, and existing transmission protocols face the risk of data leakage. SUMMARY

[0003] The purpose of the present application is to provide an emergency patient information data matching method and system based on the Internet to solve the problems in the prior art and meet the needs of fast, safe and effective data processing in emergency scenarios, and to improve the quality and efficiency of emergency medical services.

[0004] One embodiment of the present application provides an emergency patient information data matching method based on the Internet, which comprises: According to the electronic medical records, image data and real-time vital sign data of emergency patients, a data acquisition framework based on edge computing is adopted to acquire multi-source heterogeneous data in real time, and an adaptive data format conversion algorithm is used to convert data from different sources into a standardized data format, generating a time-synchronized multi-source data set; For the time-synchronized multi-source data set, a feature extraction network based on multi-head attention mechanism is used to extract patient condition characteristics, and a priority evaluation model based on fuzzy logic is used to generate patient priority scores by combining the severity of the condition and the urgency of treatment; According to the patient priority score and the condition characteristics, a medical resource matching algorithm based on graph neural network is used to dynamically match the optimal medical resources by combining hospital departments, doctor expertise and equipment availability information, and a distributed task scheduling framework is used to adjust the resource allocation scheme in real time, generating a preliminary medical resource matching result; For the matching result of medical resources, a data security transmission protocol based on homomorphic encryption is adopted to ensure the safe transmission of patient information in the Internet environment, and a differential privacy protection mechanism is adopted to add noise to sensitive data for processing to generate the final emergency patient information data matching scheme.

[0005] Optionally, according to the electronic medical record, image data and real-time vital sign data of the emergency patient, a data acquisition framework based on edge computing is adopted to acquire multi-source heterogeneous data in real time, different sources of data are uniformly converted into a standardized data format through an adaptive data format conversion algorithm, and a time-synchronized multi-source data set is generated, including: According to the electronic medical record, image data and real-time vital sign data of the emergency patient, a data acquisition framework based on edge computing is adopted to acquire multi-source heterogeneous data in real time through distributed edge nodes, each edge node is equipped with a lightweight data caching mechanism to ensure the real-time and continuity of data acquisition; For multi-source heterogeneous data, a format recognition model based on deep learning is adopted to identify the format type of different data sources, and an adaptive parsing algorithm is used to parse data of different formats into a structured intermediate representation to generate a preliminary parsed data set; For the preliminary parsed data set, a standardized conversion algorithm based on the combination of a rule engine and machine learning is adopted to uniformly convert data of different sources into a standardized data format, and a context-aware mapping rule library is used to dynamically adjust the conversion logic to generate a standardized data set; For the standardized data set, a time synchronization algorithm based on dynamic time warping is adopted to eliminate the timestamp differences of different data sources, and an interpolation filling method is used to supplement missing data to generate a time-synchronized multi-source data set.

[0006] Optionally, for the time-synchronized multi-source data set, a feature extraction network based on a multi-head attention mechanism is adopted to extract patient condition features, and a priority evaluation model based on fuzzy logic is adopted to combine the severity of the patient's condition and the urgency of treatment to generate a patient priority score, including: For the time-synchronized multi-source data set, a feature extraction network based on a multi-head attention mechanism is adopted to extract symptom description text features, visual features of image data and time sequence features of vital sign data of the patient, respectively, and a cross-modal attention mechanism is adopted to weight and fuse features of different modalities to generate multi-modal condition feature representation; For the multi-modal condition feature representation, a condition severity quantification model based on a deep neural network is adopted to calculate the severity score of the patient's condition in combination with clinical medical guidelines and expert experience, wherein an explainability module of the condition severity quantification model is adopted to generate a quantification result of the severity of the condition and the basis thereof; According to the severity score of the disease, combined with the real-time vital sign data of the patient, a time series analysis-based urgency assessment algorithm is used to predict the time window of disease deterioration, and a dynamic weight adjustment mechanism is used to generate a treatment urgency score; According to the severity score of the disease and the treatment urgency score, a fuzzy logic-based priority assessment model is used to dynamically calculate the priority score of the patient in combination with the current status of the hospital emergency resources and the treatment capacity, and a multi-objective optimization method is used to generate the final priority score and its sorting result.

[0007] Optionally, according to the patient priority score and the disease characteristics, a medical resource matching algorithm based on graph neural network is used to dynamically match the optimal medical resource in combination with the hospital department, doctor's expertise and equipment availability information, and a distributed task scheduling framework is used to real-time adjust the resource allocation scheme to generate the preliminary medical resource matching result, including: According to the hospital department, doctor's expertise and equipment availability information, a medical resource graph structure is constructed, wherein the nodes of the medical resource graph represent the departments, doctors or equipment, and the edges of the medical resource graph represent the cooperation relationship or dependency relationship between resources, and the node and edge information of the graph is dynamically maintained by introducing a resource state real-time update mechanism to generate a real-time medical resource graph; According to the patient priority score and the disease characteristics, a matching algorithm based on graph neural network is used to associate the patient demand with the nodes in the medical resource graph, and an attention mechanism is used to calculate the matching degree between the patient and the resource node to generate a preliminary patient-resource matching relationship; For the preliminary patient-resource matching relationship, a dynamic adjustment method based on a distributed task scheduling framework is used to solve resource allocation conflicts in combination with real-time resource availability and task priority, and a Nash equilibrium algorithm in game theory is used to optimize the resource allocation scheme to generate a conflict-free medical resource scheduling plan; For the conflict-free medical resource scheduling plan, a matching result generation method based on visualization technology is used to real-time feedback the matching result to medical staff, and a feedback correction mechanism is used to dynamically adjust the matching algorithm parameters to generate the final medical resource matching result.

[0008] Optionally, for the medical resource matching result, a homomorphic encryption-based data security transmission protocol is used to ensure the secure transmission of patient information in an Internet environment, and a differential privacy protection mechanism is used to add noise to sensitive data to generate a final emergency patient information data matching scheme, including: For the patient information in the medical resource matching result, a homomorphic encryption-based data security transmission protocol is used to encrypt sensitive data into ciphertext, and a lightweight key management mechanism is used to ensure the secure transmission of encrypted data in an Internet environment to generate an encrypted data transmission package; For the sensitive field in the encrypted data transmission packet, a differential privacy protection method based on Laplace mechanism is adopted, the noise addition amount is dynamically calculated according to the data sensitivity level, the adaptive noise allocation algorithm is used to add noise to the key field, and the privacy protected data packet is generated; For the privacy protected data packet, a homomorphic decryption algorithm is used at the receiving end to restore the data, an integrity verification mechanism based on a hash function is used to verify whether the data has been tampered with during transmission, and the decrypted complete data is generated; For the decrypted complete data, a matching result generation method based on visualization technology is used to integrate the medical resource matching result and the patient information, and a dynamic permission control mechanism is used to ensure that only authorized personnel can access sensitive information, and the final emergency patient information data matching scheme is generated.

[0009] Another embodiment of the application provides an Internet-based emergency patient information data matching system, which comprises: The acquisition module is used to acquire multi-source heterogeneous data in real time based on the electronic medical record, image data and real-time vital sign data of the emergency patient, and convert different sources of data into a standardized data format through an adaptive data format conversion algorithm, to generate a time-synchronized multi-source data set. The extraction module is used to extract the patient's condition characteristics from the time-synchronized multi-source data set by using a feature extraction network based on a multi-head attention mechanism, and generate a patient priority score by combining the severity of the condition and the urgency of treatment through a priority evaluation model based on fuzzy logic. The matching module is used to dynamically match the optimal medical resources based on the patient priority score and the condition characteristics by using a medical resource matching algorithm based on a graph neural network, combining hospital departments, doctor expertise and equipment availability information, and real-time adjusting the resource allocation scheme through a distributed task scheduling framework, to generate a preliminary medical resource matching result. The transmission module is used to ensure the secure transmission of patient information in an Internet environment by using a data security transmission protocol based on homomorphic encryption, and to generate the final emergency patient information data matching scheme by adding noise to sensitive data through a differential privacy protection mechanism.

[0010] Another embodiment of the application provides a storage medium having a computer program stored therein, wherein the computer program is configured to execute the method described in any of the above embodiments when running.

[0011] Still another embodiment of the present application provides an electronic device comprising a memory having a computer program stored therein and a processor configured to execute the computer program to perform the method described in any of the above.

[0012] Compared with the prior art, the emergency patient information data matching method based on the Internet provided by the present application can obtain multi-source heterogeneous data in real time according to the electronic medical record, image data and real-time vital sign data of the emergency patient, generate a time-synchronized multi-source data set through an adaptive data format conversion algorithm, extract the patient's condition characteristics through a feature extraction network to generate a patient priority score, adjust the resource allocation scheme in real time through a medical resource matching algorithm based on a graph neural network according to the patient priority score and the condition characteristics to generate a preliminary medical resource matching result, and generate a final emergency patient information data matching scheme through a data security transmission protocol based on homomorphic encryption and a differential privacy protection mechanism, so as to meet the demand for fast, safe and effective processing of data in an emergency scene and improve the quality and efficiency of emergency medical services. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 A hardware structure block diagram of a computer terminal of the emergency patient information data matching method based on the Internet provided by the embodiment of the present application is shown in the figure. Figure 2 A flowchart of the emergency patient information data matching method based on the Internet provided by the embodiment of the present application is shown in the figure. Figure 3 A structure diagram of the emergency patient information data matching system based on the Internet provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0014] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be explained as a limitation of the present application.

[0015] The embodiment of the present application first provides an emergency patient information data matching method based on the Internet, which can be applied to an electronic device such as a computer terminal, specifically a general computer, etc.

[0016] The following will be described in detail taking the computer terminal as an example. Figure 1 A hardware structure block diagram of a computer terminal of the emergency patient information data matching method based on the Internet provided by the embodiment of the present application is shown in the figure. Figure 1 As shown in the figure, the computer device comprises a processor, a memory and a network interface connected through a system bus, wherein the memory can comprise a non-volatile storage medium and an internal memory.

[0017] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions which, when executed, cause the processor to perform any one of the emergency patient information data matching methods based on the Internet.

[0018] The processor is configured to provide computing and control capabilities to support the operation of the entire computer device.

[0019] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium, which, when executed by the processor, causes the processor to perform any one of the emergency patient information data matching methods based on the Internet.

[0020] The network interface is configured to perform network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that, Figure 1 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0021] It should be understood that the processor can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0022] Referring to Figure 2 The embodiments of the present application provide an emergency patient information data matching method based on the Internet, which can include the following steps: S201, according to the electronic medical record, image data and real-time vital sign data of the emergency patient, using a data acquisition framework based on edge computing, real-time acquisition of multi-source heterogeneous data, through an adaptive data format conversion algorithm, different sources of data are uniformly converted into a standardized data format, and a time-synchronized multi-source data set is generated; The method first acquires the electronic medical records, imaging data and vital sign data of emergency patients in real time through an edge computing-based data acquisition framework. This framework uses distributed edge nodes for data collection, which can process information close to the data source, eliminating the delay that may be caused by traditional centralized computing. Through an adaptive data format conversion algorithm, the system can convert data from different sources into a standardized format, ensuring that subsequent integration and analysis can proceed smoothly. At the same time, the time-synchronized multi-source dataset generated by this method provides a reliable data foundation for subsequent disease feature extraction, priority assessment and medical resource matching.

[0023] By acquiring and standardizing multi-source heterogeneous data in real time, this method can significantly improve the information processing efficiency in emergency medical care, ensuring that medical personnel can quickly obtain comprehensive information about patients. This not only provides data support for disease assessment and priority scoring, but also lays a solid foundation for the final medical resource matching. In emergency situations, rapid acquisition and processing of information is of great significance to improve decision-making efficiency, ensure patient safety and improve treatment outcomes.

[0024] Specifically, based on the electronic medical records, imaging data and real-time vital sign data of emergency patients, an edge computing-based data acquisition framework can be used to acquire multi-source heterogeneous data in real time through distributed edge nodes. Each edge node is equipped with a lightweight data caching mechanism to ensure real-time and continuity of data acquisition. This step can acquire various data in real time at the scene of emergency or in the clinical environment of patients by setting up distributed edge nodes. These edge nodes are connected to the main data center of the hospital and can quickly collect the electronic medical records, imaging data and real-time vital sign data of patients, and ensure data continuity through a lightweight data caching mechanism, so that data will not be lost even in unstable network conditions.

[0025] Through this step, emergency doctors can obtain the latest information about patients at a faster speed, especially in cases where the patient's condition changes rapidly. This data acquisition method can effectively improve the work efficiency of the emergency team, reduce the risk of patient treatment caused by data delay, and also reduce the waste of medical resources in the data acquisition stage.

[0026] In an emergency medical environment, edge computing devices are located near patient beds or emergency equipment, collecting vital signs data such as ECG, blood pressure, etc. in real time through sensors and network interfaces. These edge nodes can process data and upload it to a central database. To ensure the real-time nature of data during collection, edge nodes will set up a lightweight data caching mechanism according to the frequency and importance of data generation. For example, if the ECG monitor outputs heart rate data every second and the blood pressure monitoring device outputs blood pressure data every 5 seconds, the edge node will quickly cache these data and send them to the background system when the network bandwidth is sufficient.

[0027] In addition, the architecture of edge computing can effectively reduce network delay. Assuming that multiple emergency cases occur in the emergency room at the same time, and information collection of multiple patients is carried out simultaneously, if traditional centralized computing is used, data transmission from the scene to the central server and then return may cause delay, affecting the doctor's decision. Edge computing effectively solves this delay problem by performing preliminary data processing and storage closer to the data source. In this way, medical staff can obtain the latest patient data within a few seconds and make appropriate responses.

[0028] Finally, the lightweight data caching mechanism of edge nodes can persist data in the case of unstable network. For example, in the case of network fluctuations, the edge node will save the data locally and automatically upload it after the network is restored. This is equivalent to establishing a security line to ensure that critical patient information is not lost under any circumstances, ensuring the effectiveness and accuracy of medical decision-making.

[0029] For multi-source heterogeneous data, a deep learning-based format recognition model is used to identify the format type of different data sources, and through an adaptive parsing algorithm, data in different formats is parsed into structured intermediate representations to generate a preliminary parsed data set. This step uses deep learning technology to identify the data format from different data sources. Through model training, the system can automatically identify the data type and apply appropriate parsing strategies to convert these data into structured formats for analysis. This process ensures that data from any device or system can be effectively utilized in subsequent processing.

[0030] Implementing automatic identification and conversion of data formats can greatly simplify the workflow of data integration and reduce the need for manual intervention. This process not only improves the speed and accuracy of data processing, but also reduces the risk of information loss due to data format mismatch, thereby laying a good foundation for subsequent data analysis.

[0031] In this process, first, the multi-source heterogeneous data is pre-processed using a deep learning-based format recognition model. Assuming there are medical devices from different manufacturers in the hospital, the output data formats may vary. For example, an electrocardiogram device may output data in CSV file format, while a blood pressure monitor may use JSON format. The diversity of these data formats has caused difficulties for subsequent data processing. Through the trained deep learning model, the system can automatically identify the format of each data and select the appropriate parsing strategy for it.

[0032] Next, an adaptive parsing algorithm is used to convert each format of data into a unified structured intermediate representation. This conversion process will automatically generate corresponding database fields based on key information in the data. For example, electrocardiogram data can be parsed into multiple physiological parameters (such as heart rate, electrocardiogram waveform, etc.), while blood pressure data includes systolic pressure, diastolic pressure, etc. In this process, the system also considers the real-time and completeness of the data to ensure that important information is not missed in parsing.

[0033] After successfully generating the preliminary parsed data set, the system will regularly update the parsing model to adapt to new data formats. With the continuous updating of medical devices and the progress of technology, new data formats and types may appear, and the system maintains the efficiency and accuracy of data parsing through continuous learning and self-optimization. Such a mechanism ensures that the final generated data set is consistent and usable in any situation, providing strong support for subsequent data analysis and processing.

[0034] For the preliminary parsed data set, a standardized conversion algorithm based on rule engine and machine learning is used to unify the data from different sources into a standardized data format. Through a context-aware mapping rule library, the conversion logic is dynamically adjusted to generate a standardized data set. In this step, the rule engine and machine learning algorithm are combined to standardize the preliminary parsed data. This process ensures that important information is preserved during conversion, while adapting to the characteristics of different data sources, dynamically adjusting the conversion strategy to make the conversion result conform to the preset standard format.

[0035] The importance of this step lies in providing a unified standard format for the integration of multi-source data, making subsequent analysis and processing more efficient and consistent. This method not only improves data quality, but also enhances data usability, allowing medical teams to use standardized data for in-depth analysis and decision-making, thereby better serving patients.

[0036] This step first establishes a set of standardized data format rules through a rule engine. These rules cover data types from different medical devices and systems, such as for medical record information, standardized fields may include patient name, gender, age, admission time, medical record number, etc. Through the rule engine, the system can automatically detect and mark each field, ensuring that the output format is consistent regardless of the data source.

[0037] Next, by applying techniques combined with machine learning, the system can dynamically adjust rules during the conversion process. For example, when the system receives data from a new medical device, the machine learning model can identify and update the relevant rules to adapt to the new data type. For example, if a new blood glucose monitoring device is introduced, the system will learn and update the rules to convert blood glucose values into a standardized format. In this way, doctors can obtain consistent formats and information when viewing patient information, regardless of the data source, thereby reducing the time for parsing and understanding.

[0038] Finally, the standardized data set is updated in real time and stored in a central database. This database is not just a static data storage system, but a dynamically adjusted data management platform. The system can associate data based on context, such as combining a patient's medical record with real-time monitoring vital signs, so that medical personnel can obtain the most comprehensive information when making further treatment decisions. This real-time updating and standardization greatly improves the efficiency of medical personnel accessing patient information and decision-making speed.

[0039] For the standardized data set, a time synchronization algorithm based on dynamic time warping is used to eliminate the timestamp differences of different data sources, and an interpolation filling method is used to supplement missing data to generate a time-synchronized multi-source data set.

[0040] In this step, the Dynamic Time Warping (DTW) algorithm is used to process data from different sources to eliminate differences in their timestamps. Due to the time delay or frequency inconsistency when different devices collect data, the time series of the data may be misaligned, affecting the accuracy of subsequent analysis. Through the DTW algorithm, the system can rearrange the time series to align them on the time axis. For missing data caused by time alignment, an interpolation filling method will be used to supplement the missing values by interpolating the existing data points, thereby generating a time-synchronized multi-source data set.

[0041] This step ensures that data from different sources are consistent in time, allowing subsequent data analysis and model training to be aligned in the time dimension. This time synchronization not only improves the quality of data integration, but also provides strong support for dynamic monitoring of patient condition changes. In emergency situations, real-time and accurate time-synchronized data is crucial for making quick medical decisions.

[0042] The primary task of this step is to eliminate timestamp differences from different data sources. Suppose an emergency patient's electrocardiogram monitor and blood pressure monitor record data with different timestamps, with the electrocardiogram data updating every second and the blood pressure data updating every 5 seconds. This time difference can lead to ineffective comparison and correlation when analyzing vital signs. Through the dynamic time warping (DTW) algorithm, the system can optimally match the two time series, aligning them on the time axis for subsequent analysis.

[0043] Next, filling in missing data is an important step to ensure data integrity. For example, during synchronization, if missing blood pressure data is found in a certain time period, the system will use interpolation filling methods to predict the missing values. Assuming the blood pressure data at 10 seconds and 15 seconds is 120 mmHg and 130 mmHg respectively, the system can estimate the blood pressure data at 12 seconds to be 125 mmHg through linear interpolation. This process ensures data continuity and reasonableness while effectively addressing data loss caused by time synchronization.

[0044] Finally, the generated time-synchronized multi-source data set will provide a basis for subsequent analysis. This data set is not only accurate and comprehensive, but also provides strong data support for doctors' clinical decisions in emergency situations. Enhanced data availability is crucial in emergency situations, helping doctors quickly understand patients' real-time conditions and develop effective treatment plans. Through such processing, the medical team can better grasp the patient's condition changes, improve emergency efficiency, and optimize treatment effectiveness.

[0045] S202, on the time-synchronized multi-source data set, a feature extraction network based on multi-head attention mechanism is used to extract patient condition features, and a priority evaluation model based on fuzzy logic is used to combine the severity of the condition and the urgency of treatment to generate a patient priority score; In this step, a feature extraction network based on multi-head attention mechanism is employed to analyze the time-synchronized multi-source dataset. This network structure can effectively extract disease features from different modalities of data, including patient symptom description text, image data, and real-time vital sign data. The features of each modality data are weighted processed by their respective heads, and finally fused through a cross-modal attention mechanism to form a comprehensive multi-modal disease feature representation. This method allows the system to focus on key features in different types of data, enabling a comprehensive analysis from multiple dimensions when assessing the patient's condition.

[0046] The core of this process lies in the comprehensive analysis of the patient's diverse information, which can more comprehensively capture the patient's true condition, helping to improve the accuracy and reliability of the priority score. Through this multi-head attention mechanism-based feature extraction, medical personnel can quickly identify the key disease characteristics of emergency patients, which helps to make more effective treatment decisions in time-critical situations. This is particularly important in emergency scenarios, as rapid and accurate diagnosis can directly affect the treatment effect and survival rate of patients.

[0047] Specifically, a multi-head attention mechanism-based feature extraction network can be used to extract patient symptom description text features, image data visual features, and vital sign data time series features from a time-synchronized multi-source dataset. Through cross-modal attention mechanism, the features of different modalities are weighted and fused to generate a multi-modal disease feature representation. In this step, a feature extraction network based on multi-head attention mechanism is employed to analyze the time-synchronized multi-source dataset. This network structure can effectively extract disease features from different modalities of data, including patient symptom description text, image data, and real-time vital sign data. The features of each modality data are weighted processed by their respective heads, and finally fused through a cross-modal attention mechanism to form a comprehensive multi-modal disease feature representation. This method allows the system to focus on key features in different types of data, enabling a comprehensive analysis from multiple dimensions when assessing the patient's condition.

[0048] Through the weighted fusion of cross-modal attention mechanism, the system can integrate various types of data features to generate a comprehensive multi-modal disease feature representation. Such representation greatly improves the understanding of the patient's condition, making subsequent assessment more accurate. In practical applications, such feature fusion is crucial for clinical decision-making, providing doctors with more comprehensive information support to help them quickly determine the patient's condition.

[0049] In this step, we first need to extract various features from the multi-source dataset. This includes the patient's symptom description text, image data such as X-rays or CT images, and real-time vital sign data such as heart rate and blood pressure. The extraction of text features usually employs natural language processing techniques, such as using word embeddings to convert text into fixed-dimensional vector representations. Image data is processed through convolutional neural networks (CNN) to extract key visual features and form corresponding feature maps. Vital sign data is processed as time series, usually using recurrent neural networks (RNN) or long short-term memory networks (LSTM) to capture its temporal dynamic changes.

[0050] Next, using the multi-head attention mechanism, the system weights and fuses the features of different modalities. Specifically, in this process, multiple attention heads will independently weight each type of feature and learn the relevance of each modality in the process. For example, when processing image features, one attention head may focus on information in the lesion area, while another attention head may focus on the health status of surrounding tissues. In this way, the system can more comprehensively capture the patient's condition information and effectively integrate each feature.

[0051] Finally, after the weighted fusion of the cross-modal attention mechanism, a multi-modal condition feature representation is generated. This feature representation is a high-dimensional vector that can contain the comprehensive performance of various input information. For example, for a patient with difficulty breathing, the final feature vector not only contains "difficulty breathing" in the symptom description, but also includes lung infection features shown in the image data and real-time data of heart rate and blood pressure. Such comprehensive features will provide a solid foundation for subsequent disease severity scoring.

[0052] For the multi-modal condition feature representation, a disease severity quantification model based on deep neural networks is used to calculate the patient's disease severity score, combining clinical medical guidelines and expert experience. Through the explainability module of the disease severity quantification model, the quantification results and basis of the disease severity are generated; In this step, the disease severity quantification model based on deep neural networks is used to convert the multi-modal condition feature representation into a specific disease severity score. This model not only combines clinical medical guidelines, but also introduces the experience of medical experts to ensure the scientificity and clinical practicability of the score. The explainability module of the model aims to provide an analysis of the reasons behind the score, allowing medical personnel to understand the impact of each feature on the final score, thereby enhancing the transparency and credibility of decision-making.

[0053] By combining clinical guidelines and expert experience, the model not only provides a quantitative illness severity score, allowing emergency physicians to quickly assess the severity of a patient's condition, but also enhances the interpretability of the scoring process, making it easier for doctors to make clinical judgments and treatment plans based on it. This combination of quantification and interpretability makes medical decision-making more scientific, reduces the interference of subjective factors, and helps improve emergency response efficiency.

[0054] In this step, a deep neural network-based illness severity quantification model needs to be built first. This model will receive the generated multi-modal illness feature representation as input, process it through multiple fully connected layers and activation functions, and finally output an illness severity score. The design of the model needs to consider the integration of clinical medical guidelines and expert experience, such as the scoring criteria for specific diseases (such as heart failure, pneumonia, etc.) (such as SOFA score, APACHE II score) will be part of the model.

[0055] Secondly, the model should also include an interpretability module so that doctors can understand how the model arrives at the illness severity score. This module can be based on attention mechanisms and use visualization methods to show the importance of input features in scoring. For example, in the case of a score of 70, the model may explain that "vital signs" account for 40 points, image features account for 30 points, and patient symptom descriptions account for 30%. This interpretability not only improves the doctor's trust, but also helps medical personnel better understand the specific situation of the patient.

[0056] Finally, after model calculation, the illness severity score is generated, and the corresponding basis and explanation are also provided. This process can be achieved by comparing with clinical standards. For example, if a patient's score is higher than the critical value, the system will provide an alarm to remind medical staff to intervene as soon as possible. Doctors can make clinical decisions quickly based on these quantitative scores and explanations, improving the timeliness and effectiveness of treatment.

[0057] For illness severity scoring, real-time vital sign data is combined with a time series analysis-based urgency assessment algorithm to predict the time window of illness deterioration, and a dynamic weight adjustment mechanism is used to generate a treatment urgency score. The key to this step is to use real-time vital sign data combined with illness severity scoring and a time series analysis algorithm to predict the urgency of the illness. The system analyzes the trends of different vital signs (such as heart rate, blood pressure, and respiratory rate) and combines historical data models to predict the time window of possible illness deterioration in real time. The dynamic weight adjustment mechanism ensures that the model can quickly adapt to the influence of real-time data when features change, thus generating an accurate treatment urgency score.

[0058] This evaluation process is crucial as it helps medical teams identify high-risk patients promptly, ensuring that the most urgent cases are treated in the shortest time possible. This urgency score based on dynamic prediction is particularly important in emergency scenarios, helping to improve patient survival rates and treatment success rates, and providing strong support for doctors' clinical decisions.

[0059] In this step, the system first needs to monitor the patient's vital signs data in real-time, including heart rate, blood pressure, respiratory rate, etc. By establishing a time series analysis model (such as LSTM or GRU), the system can capture the dynamic characteristics of these vital signs over time. For example, during data collection, if a patient's heart rate suddenly rises while blood pressure drops, the system will immediately identify this change and issue a warning.

[0060] Then, combined with the generated severity score, the system will use a time series analysis-based urgency assessment algorithm to predict the time window of disease deterioration. By analyzing the trends and patterns of historical data, the system can generate an estimated time for the patient's condition to worsen. For example, if the model predicts that a patient may experience cardiac arrest within 15 minutes, such information can greatly help medical staff prepare and respond in advance.

[0061] Finally, implement a dynamic weight adjustment mechanism to ensure the adaptability and timeliness of the system. When changes in vital signs are detected, the system will adjust the calculation weights of the urgency score. For example, if the rate of change of a certain vital sign increases, the system will give it a higher weight in calculating the urgency score, thereby implementing a more accurate intervention strategy. Through this process, the system's treatment urgency score not only reflects the patient's current state, but also provides an important basis for medical staff's decision-making.

[0062] For the severity score and treatment urgency score, a fuzzy logic-based priority assessment model is used to dynamically calculate the priority score of each patient based on the current status of hospital emergency resources and treatment capacity, and through multi-objective optimization methods, the final priority score and its ranking results are generated.

[0063] In this step, the fuzzy logic-based priority assessment model integrates the severity score and treatment urgency score, and combines the real-time resource situation and treatment capacity of the hospital to dynamically calculate the priority score of each patient. The application of fuzzy logic enables the system to handle uncertainties in the data, reasonably reflect the actual treatment needs of patients, and also consider the existing resources of the hospital, ensuring that the priority assessment is accurate and reasonable.

[0064] Through dynamic calculation of priority scores, hospitals can optimize resource allocation in emergency environments, ensuring that patients in need of urgent treatment can be prioritized for treatment. This process is particularly important in responding to public health emergencies or high-volume emergency scenarios, effectively improving the hospital's response capacity and patient treatment efficiency under resource constraints.

[0065] In this step, a fuzzy logic-based priority assessment model is first constructed. The model obtains the severity score and urgency score, and converts them into fuzzy sets to handle uncertainty. For example, the model can define "high" priority and "low" priority as fuzzy sets, and convert specific scores to corresponding fuzzy levels by defining membership functions.

[0066] Next, the model dynamically calculates the priority score of each patient in combination with the current emergency resource status and medical capacity of the hospital. In this step, the system will obtain real-time resource data of the hospital, including available medical staff, equipment status and other information, and incorporate it into the priority calculation. For example, if a patient has a severe condition and high urgency, and the operating room of the hospital is currently idle, the system will dynamically increase the priority score of the patient according to this information, ensuring that it can receive the necessary treatment in a timely manner.

[0067] Finally, a multi-objective optimization algorithm is used to sort the priority scores and generate an optimized patient treatment list. Through the algorithm, patients with the highest priority can be quickly treated. In the case of limited resources in an emergency, this sorting effect can significantly improve the efficiency of the hospital, ensuring optimal allocation of clinical resources. For example, in an emergency room that receives multiple patients simultaneously, the system can help medical staff automatically identify patients for priority treatment, ensuring that those most in need of assistance can receive medical services in the shortest time, thereby maximizing the success rate of emergency treatment.

[0068] S203, according to the patient priority score and the condition characteristics, a medical resource matching algorithm based on graph neural network is used to dynamically match the optimal medical resources by combining the information of hospital departments, doctor expertise and equipment availability, and through a distributed task scheduling framework, the resource allocation scheme is adjusted in real time to generate a preliminary medical resource matching result; In this step, the system first constructs a medical resource graph structure, which consists of different nodes and edges. The nodes represent the various departments, doctors, and available equipment in the hospital, while the edges represent the collaborative relationships or dependencies between these resources. Using this information in the graph structure, the graph neural network can effectively identify the most suitable medical resources based on the patient's priority score and condition characteristics. Through the matching algorithm of the graph neural network, the system analyzes the correlation between the patient's needs and the hospital's resource nodes, and calculates the matching degree between the patient's needs and each resource node through the attention mechanism. For example, if a patient urgently needs cardiology medical services, the system can quickly find the doctors and equipment in the department through the graph neural network, and evaluate their availability to ensure dynamic matching of optimal resources.

[0069] The core of this process is to ensure that patients receive the most appropriate treatment options through dynamic matching of medical resources. By analyzing patients' priority scores and disease characteristics, the system can promptly identify cases that require the most attention. This not only improves the efficiency of patient treatment but also increases the utilization rate of hospital resources. For example, in the emergency room, emergencies often lead to resource shortages. With the help of this matching algorithm, available medical resources can be quickly identified and effectively dispatched to ensure that patients receive the necessary medical services in the shortest possible time. This efficient matching mechanism can significantly improve the success rate of treatment and provide strong support for the hospital's operational management.

[0070] Specifically, a medical resource graph structure can be constructed based on hospital departments, doctor expertise, and equipment availability information. The nodes of the medical resource graph represent departments, doctors, or equipment, and the edges of the medical resource graph represent the collaborative relationships or dependencies between resources. By introducing a real-time resource status update mechanism, the node and edge information of the graph are dynamically maintained, generating a real-time medical resource graph. In this step, the system constructs a comprehensive medical resource graph based on the hospital's different departments, individual physician expertise, and equipment availability. This graph reflects the distribution and interconnectedness of medical resources within the hospital. Nodes represent hospital departments, physicians, and medical equipment, while edges represent the collaborative relationships or dependencies between these resources. For example, an edge between a cardiologist and an electrocardiogram device indicates that the doctor can use the device for diagnosis.

[0071] To maintain the real-time nature of the medical resource map, the system incorporates a real-time resource status update mechanism. By regularly querying the hospital information system, the system promptly updates node statuses, such as whether a doctor is on duty or whether a piece of equipment is available. This real-time update function ensures the accuracy of the medical resource map, allowing medical staff to obtain the most accurate information on resource availability when applying for resources.

[0072] The significance of constructing a medical resource graph structure lies in providing an intuitive and systematic way for resource management within a hospital, making the relationships between various medical resources more explicit. This graph structure makes resource coordination more efficient, especially in emergency situations, where quick access to accurate resource allocation information can directly impact patient treatment outcomes. Meanwhile, the introduction of real-time updating mechanisms enhances the hospital's control over medical resources, improves overall operational efficiency, and ensures that resource allocation can adapt to dynamically changing medical needs.

[0073] In this process, the system first collects structural information within the hospital, including the responsibilities of each department, the specialties of doctors, and the detailed list of equipment owned. The system will automatically extract this information through the Hospital Information System (HIS) or Electronic Medical Record System (EMR), ensuring the accuracy and timeliness of the information. For example, cardiologists in the cardiology department specialize in cardiovascular diseases, while the radiology department has high-end equipment for imaging examinations. By nodalizing these data, the system forms a preliminary medical resource graph.

[0074] Next, the system defines the edges of the graph, i.e., the collaboration relationship between resources. For example, a direct edge can be established between a cardiologist and an electrocardiogram device, indicating that the doctor can use the device to perform electrocardiogram monitoring for patients. In addition, some departments may form edges through coordination practices, such as the collaboration relationship between surgeons and anesthesiologists. In practice, such edges can help the system effectively analyze which doctor has the closest collaboration relationship with a particular device or other doctors.

[0075] Finally, to ensure the real-time nature of the medical resource graph, the system introduces a resource state real-time updating mechanism. This mechanism obtains the latest resource state information from the database through a timed task, updates the state of the nodes, such as whether the doctor is on duty, the current state of the equipment (such as whether it is being repaired or has been reserved), etc. The update frequency can be adjusted according to the actual needs of the hospital to ensure that medical staff obtain the latest resource information. For example, in emergency situations, the system may update every minute to ensure the timely allocation of emergency resources.

[0076] According to the patient priority score and disease characteristics, a matching algorithm based on graph neural networks is used to associate patient needs with nodes in the medical resource graph. Through an attention mechanism, the matching degree between patients and resource nodes is calculated, generating a preliminary patient-resource matching relationship. This step mainly utilizes the Graph Neural Network (GNN) algorithm to effectively match the patient's priority score and condition characteristics with the nodes in the medical resource graph. First, the system calculates the corresponding priority score based on the patient's condition description, vital signs, imaging examination results, and other information. Next, using the matching algorithm based on the graph neural network, the system can associate these patient needs with each node in the medical resource graph (such as departments, doctors, and equipment).

[0077] During the matching process, an attention mechanism is used to improve the accuracy of the match. Through the attention mechanism, the system can focus on the resource nodes that have the greatest impact on the patient's condition, thereby better assessing the matching degree of patient needs and resources. For example, for a patient who needs heart surgery, the system will emphasize the availability of cardiac surgeons and related equipment, while ignoring the resources of other unrelated departments. After this series of calculations, the system will generate a preliminary patient-resource matching relationship, clearly identifying which resources are best suited to meet the patient's needs.

[0078] This step significantly improves the efficiency and accuracy of patient and medical resource matching. By calculating the matching degree, hospitals can quickly identify the most suitable doctors and equipment, thereby significantly reducing patient waiting times, especially in emergency situations where timely resource matching is critical. In addition, the use of graph neural networks and attention mechanisms allows for the establishment of a more flexible and intelligent resource allocation system in complex medical environments, improving the overall service capabilities and resource utilization of the hospital.

[0079] In this step, the patient's priority score and condition characteristics are first analyzed through a deep learning model. This can include the patient's vital signs, medical history, current symptoms, and other information. These inputs are pre-processed and feature-engineered to generate a condition feature vector. For example, a heart failure patient's feature vector may reflect their heart rate, blood pressure, and recent blood test results. These feature vectors are then used to calculate the patient's priority score, ensuring that critical cases are prioritized for treatment.

[0080] Next, the system uses a Graph Neural Network (GNN) to associate patient needs with nodes in the medical resource graph. GNN can effectively utilize the graph structure of medical resources to assess the matching degree of patient needs and resources in multiple dimensions. The system first embeds each node in the graph (department, doctor, equipment) as a vector and calculates the similarity between each node and the patient's feature vector. This is like a social network where the system can intelligently recommend friends or topics that best match a user's interests.

[0081] To further enhance the accuracy of the matching, the system introduces an attention mechanism. The attention mechanism assigns different weights to the associated resource nodes based on the specific characteristics of the patient's condition. For example, if the patient's condition is exceptionally severe, the system may give higher weight to cardiologists, while reducing the weight for secondary equipment. This mechanism ensures that the matching process can flexibly respond to the needs of different patients, ultimately generating a preliminary patient-resource matching relationship, i.e., which resources are best suited to the patient's needs.

[0082] For the preliminary patient-resource matching relationship, a dynamic adjustment method based on a distributed task scheduling framework is adopted. Combining real-time resource availability and task priority, it solves resource allocation conflicts, optimizes resource allocation schemes through Nash equilibrium algorithm in game theory, and generates a conflict-free medical resource scheduling plan. In this step, the system uses a distributed task scheduling framework to dynamically adjust the preliminary patient-resource matching relationship. This framework can monitor the availability of resources and the priority of tasks in real time, and adjust resource allocation in a timely manner to avoid potential conflicts. For example, two patients may need the same equipment at the same time. At this time, the system will monitor the situation in real time, evaluate the priority scores of the two patients, and dynamically adjust the use arrangement of the equipment.

[0083] At the same time, Nash equilibrium algorithm in game theory is used to optimize the resource allocation scheme. This algorithm can seek a state in the case of multi-party interest conflict, so that no one can obtain a better result by changing the strategy. In the context of medical resource allocation, this means that once the allocation scheme is determined, the allocation of all resources has reached an optimal state, ensuring that the medical needs of all patients are reasonably met, and minimizing the idle and waste of resources.

[0084] Through dynamic adjustment and optimization of resource allocation schemes, hospitals can significantly improve the utilization efficiency of resources, especially in high-pressure emergency environments, ensuring that patients can obtain the required medical services in a timely manner. The dynamic framework makes management more flexible, allowing real-time response to various emergencies and reducing unnecessary delays caused by resource conflicts. Ultimately, the generated conflict-free medical resource scheduling plan can ensure that patients receive rapid and effective treatment in emergency situations, which is of great significance to improving the quality of medical services.

[0085] In this step, the system first obtains the preliminary patient-resource matching relationship and inputs it into the distributed task scheduling framework. Distributed task scheduling allows the system to distribute computing tasks among different processing nodes, ensuring efficient resource allocation. For example, when multiple patients need the same equipment, the system will analyze all patient needs in parallel to identify potential resource conflicts in a timely manner.

[0086] Next, the system dynamically adjusts resources based on real-time availability and task priority. By monitoring the real-time status of each department and equipment, the system updates available resources in real-time and intelligently schedules them according to patient priority scores. If a patient's condition suddenly worsens, the system will immediately re-evaluate and adjust the original scheduling strategy to prioritize emergency needs. At the same time, the system can also analyze the best strategy for resource use based on historical data using intelligent algorithms, thereby reducing delays caused by repeated scheduling.

[0087] Finally, the Nash equilibrium algorithm in game theory is used to optimize the resource allocation scheme. By simulating the interaction between patients, doctors and equipment, the system can find an optimal allocation state that maximizes resource use. In this process, the algorithm continuously evaluates each participant's strategy to ensure that no one can gain more benefits by actively changing their strategy. Ultimately, the generated conflict-free medical resource scheduling plan ensures that all relevant parties can reach a consensus in resource allocation, thereby improving medical efficiency.

[0088] For the conflict-free medical resource scheduling plan, a matching result generation method based on visualization technology is used to provide real-time feedback to medical staff. Through a feedback correction mechanism, the matching algorithm parameters are dynamically adjusted to generate the final medical resource matching result.

[0089] In this step, the system first presents the conflict-free medical resource scheduling plan to medical staff through visualization technology. This process allows medical staff to intuitively understand available resources and patient demand. The visualization interface may include resource allocation graphs, timelines, and real-time status of each department, so that medical staff can quickly identify patients for priority treatment and required medical resources, thereby making more efficient decisions.

[0090] In addition to visualization, the system also designs a feedback correction mechanism. Medical staff can provide real-time feedback on problems encountered during actual operations through the interface. For example, if a patient's condition changes during treatment, medical staff can quickly update the corresponding demand, and the system will dynamically adjust the parameters of the matching algorithm to better adapt to changes in actual demand. This flexible feedback mechanism means that the system is not just static, but can be optimized and improved in real time.

[0091] Through this process, the matching results of medical resources can not only be quickly fed back to medical staff, but also be continuously updated and optimized according to actual conditions, greatly improving the flexibility and accuracy of decision-making. Visualization technology not only improves the transparency of information, but also helps medical staff better understand complex resource allocation strategies and promotes collaboration and communication among teams. This dynamic matching process enables hospitals to quickly adjust resource allocation strategies in response to unexpected situations, achieving efficient medical services and providing the best care experience for patients.

[0092] In this step, the system displays the conflict-free medical resource scheduling plan to medical staff through visualization technology. The visualization interface may include a resource scheduling calendar view, real-time status monitoring chart, and patient demand priority list, etc. This allows medical staff to easily understand the current resource usage and patient demand, thereby optimizing on-site decision-making.

[0093] Subsequently, the system introduces a feedback correction mechanism, allowing medical staff to provide real-time feedback on the matching results. For example, medical staff can record the actual use of resources and patient reactions after treatment. This data will be quickly fed back to the system as a basis for adjusting the subsequent matching process. If the matching effect between a certain type of resource and patient demand is not ideal, the system can dynamically adjust the parameters of the matching algorithm based on this feedback information to improve the accuracy of matching.

[0094] Finally, the corrected algorithm and matching results will generate the final medical resource matching results. This result not only includes the final matching relationship between patients and medical resources, but also provides suggestions for medical staff, such as which department's doctor is suitable for handling a specific emergency situation or the availability of a certain device. This dynamic and flexible feedback mechanism ensures that medical resource allocation can continuously adapt to changes in demand in actual operations, improving the service quality and response speed of the hospital.

[0095] S204, for the medical resource matching result, a data security transmission protocol based on homomorphic encryption is adopted to ensure the secure transmission of patient information in the Internet environment, and a differential privacy protection mechanism is used to add noise to sensitive data to generate the final emergency patient information data matching scheme.

[0096] In the process of emergency patient information data matching, it is crucial to ensure the secure transmission of patient information in the Internet environment. For this purpose, a data security transmission protocol based on homomorphic encryption is adopted. This protocol can perform specific computing operations on encrypted data, making it impossible for unauthorized users to access the data even if it is intercepted during transmission. First, the medical resource matching results contain a large amount of sensitive patient information, such as name, medical record, treatment process, etc. These information are converted into ciphertext by homomorphic encryption algorithm. At this time, a lightweight key management mechanism ensures that only users with appropriate permissions can decrypt, ensuring the security of the data. At the same time, in order to further protect sensitive data, the system also implements a differential privacy mechanism to add noise to the data. This means that a certain amount of random noise is added to the patient information when it is shared, avoiding the disclosure of specific personal information, and thus generating the final emergency patient information data matching scheme.

[0097] The combination of this data security transmission protocol based on homomorphic encryption and differential privacy protection mechanism plays an important role in emergency medical scenarios. First, it effectively reduces the risk of sensitive information being leaked or misused, ensuring patient privacy compliance and enhancing the trust of patients and their families in medical information management. Second, in emergency scenarios, medical personnel often need to quickly access patient information to make timely decisions, and through encryption and privacy protection mechanisms, the security of information is no longer an obstacle to efficiency. Finally, it can ensure data security without affecting data usability, enabling medical institutions to comply with laws and regulations while more efficiently using data for decision-making and resource allocation, thereby improving overall medical service quality.

[0098] Specifically, the patient information in the medical resource matching results can be encrypted as ciphertext using a data security transmission protocol based on homomorphic encryption, and a lightweight key management mechanism can be used to ensure the secure transmission of encrypted data in the Internet environment, generating an encrypted data transmission package. In this step, the system performs homomorphic encryption on the patient information in the medical resource matching results to ensure the security and privacy protection of the data in the Internet environment. First, the patient information is subjected to a sensitivity assessment, and all fields that need to be protected are marked as "sensitive". The system selects a suitable homomorphic encryption algorithm to encrypt these sensitive data, such as using Paillier or RSA encryption algorithm to convert them into ciphertext. During the encryption process, a lightweight key management mechanism is used to generate a key pair, including a public key and a private key, the public key is used to encrypt the data and the private key is used to decrypt. Once the encryption process is complete, the system forms an encrypted data transmission package ready for transmission to the target system.

[0099] By using homomorphic encryption technology on patient information, even if the data is intercepted during transmission, it cannot be illegally obtained or interpreted, greatly enhancing the security and privacy protection of patient information. At the same time, the application of lightweight key management mechanism simplifies the complexity of key usage and improves the operation efficiency of the system. This secure transmission scheme enables medical institutions to confidently exchange information while meeting legal regulations, thereby better serving patients.

[0100] In this step, the patient information in the medical resource matching result is first subjected to homomorphic encryption processing. The homomorphic encryption algorithm (such as Paillier or RSA) adopted allows calculations to be performed in an encrypted state. The specific implementation of this process includes encrypting the patient's personal information (such as name, medical record, treatment plan) to ensure that this information is not leaked during transmission. These information are marked as sensitive data and encrypted through a unified encryption protocol to ensure that each sensitive field is protected accordingly.

[0101] Next, the system implements a lightweight key management mechanism. Under this mechanism, whenever new patient information is generated, the system automatically generates a pair of public and private keys for the information. The public key is used for data encryption, while the private key is used for subsequent decryption operations. The key management mechanism ensures that only authorized users can access and manage these keys. To enhance security, the keys are stored in a dedicated secure module or hardware security unit to prevent unauthorized access. In addition, the system regularly rotates the keys to further reduce potential security risks.

[0102] Finally, through the above encryption process, the system generates a data transmission package containing encrypted patient information. During transmission, this data package is sent to the target system through a secure communication channel (such as SSL / TLS). In this way, even if the data is intercepted during transmission, attackers cannot interpret the sensitive information because the data has been encrypted, ensuring the confidentiality of the information and protecting the privacy of patients while ensuring rapid response to emergency needs.

[0103] For sensitive fields in the encrypted data transmission package, a differential privacy protection method based on the Laplace mechanism is used to dynamically calculate the noise addition amount based on the data sensitivity level. Through an adaptive noise allocation algorithm, noise is added to the key fields to generate a privacy-protected data package. In this step, the differential privacy protection method based on the Laplace mechanism is applied to the sensitive fields in the encrypted data transmission package. Differential privacy is a privacy protection technique for databases that ensures individual privacy by adding noise to the results. Specifically, the system dynamically calculates the amount of noise to add based on the sensitivity level of the data. For example, for highly sensitive information such as patient names, the system will add relatively large noise, while for less sensitive data such as disease descriptions, the noise added will be smaller. Through the adaptive noise allocation algorithm, the system can assess the sensitivity of each field in real time and adjust the allocation of noise accordingly, ensuring that differential privacy protection is appropriate and effective, ultimately generating a privacy-protected data package.

[0104] With the introduction of the differential privacy mechanism, the system can effectively protect the personal information of patients and prevent any patient's privacy from being exposed during big data analysis or information sharing. This strategy not only meets the requirements of modern data protection regulations, but also enhances the trust of patients in the medical system. The final privacy-protected data package ensures information security while still allowing medical institutions to conduct efficient data analysis and decision-making, promoting the optimization of medical services.

[0105] The key to this step is to perform differential privacy processing on the encrypted data transmission package. First, the system needs to evaluate the sensitivity level of each sensitive field. This evaluation can be based on individual patient information or general standards, such as patient medical history and diagnosis results, which are generally considered highly sensitive information, while some public information (such as admission time) can be considered less sensitive. This evaluation ensures that the amount of noise added by the differential privacy mechanism matches the data sensitivity, further protecting patient privacy.

[0106] Next, the system introduces the Laplace mechanism into the implementation of differential privacy. By applying the Laplace distribution, the specific value of the added noise is set. For sensitive fields, the system dynamically calculates the amount of noise to add to ensure that individual information cannot be inferred when the data is released. In the implementation process, the system uses an adaptive noise allocation algorithm to optimize the distribution of noise. This algorithm automatically adjusts the increment of noise based on the different characteristics of patient information, ensuring that important information does not lose its value due to excessive noise processing.

[0107] Through the above two steps, the system finally generates a privacy-protected data package that maintains data usability and analyzability while protecting privacy. For example, adding appropriate Laplace noise to patient medical history information ensures that even if the data is leaked, individual information may not be directly identifiable, thus achieving a certain degree of privacy protection. This processing method ensures that legal and regulatory requirements are met while continuing to support medical decision-making and resource optimization.

[0108] After privacy protection, the data packets are decrypted at the receiving end using homomorphic decryption algorithms. An integrity verification mechanism based on hash functions is used to verify whether the data has been tampered with during transmission, and the decrypted complete data is generated. In this step, the receiving end first decrypts the privacy-protected data packets using homomorphic decryption algorithms, thus restoring the original patient information. This step ensures that only authorized users with the corresponding key can access the data. After decryption, the system generates an integrity check value using a hash function, which is used to compare whether the data has been tampered with during transmission. The decrypted data is processed using a hash algorithm to generate a unique hash value. The receiving party compares the received integrity check value with the hash value sent by the sender. If they are consistent, it means the data has not been tampered with; if they are not consistent, it means the data may have been attacked during transmission.

[0109] This step is crucial to ensure the integrity and security of the data. On the one hand, through homomorphic decryption, the confidentiality of patient information is maintained, and only authorized personnel can access the data; on the other hand, the use of hash verification mechanism can effectively prevent data from being tampered with or lost during transmission, thus enhancing the reliability of patient information management. At the same time, this integrity verification mechanism also provides strong protection for medical institutions to comply with data protection regulations, improving the compliance and credibility of the institution.

[0110] In this step, the receiving end first needs to perform homomorphic decryption on the privacy-protected data packets. By using the private key generated in advance, the system can restore the encrypted data to readable patient information. When performing homomorphic decryption, the receiving party will use a special decryption algorithm to ensure that each piece of encrypted sensitive information is correctly restored. During this process, it is crucial to ensure the security of the decryption operation, and the decryption process should be carried out in a secure environment to prevent malicious interference.

[0111] Next, after decryption is completed, the system applies a hash function to generate a hash value as a check basis for data integrity. The hash function will process the entire decrypted data to generate a unique hash value. At this time, the receiving system will compare the generated hash value with the hash value provided by the sender during transmission. If they are completely consistent, it means that the data has not been tampered with or damaged during transmission, and the receiving end can safely use the information.

[0112] If the hash values do not match, the system immediately takes measures such as flagging the transmission as unsafe or generating an alert to ensure the integrity and security of the data are not compromised. Additionally, the system can implement a logging mechanism to track unauthorized access attempts or signs of data tampering to address potential security issues. Finally, if the data is confirmed to be complete and error-free, the decrypted complete data is integrated and delivered to the relevant medical staff for further processing.

[0113] For the decrypted complete data, a matching result generation method based on visualization technology is used to integrate the medical resource matching results with patient information. Through a dynamic permission control mechanism, only authorized personnel can access sensitive information to generate the final emergency patient information data matching plan.

[0114] In this step, the receiving end processes the decrypted complete data to effectively integrate medical resource matching results with patient information. First, the system uses a visualization-based method to display the decrypted data in the form of intuitive graphs or dashboards, making it easy for medical staff to quickly understand the overall medical resource allocation status and patient-specific conditions. For example, data can be displayed through charts showing patient priority scores, required medical resources, doctor scheduling, etc. Such visualization not only improves information transmission efficiency but also helps medical staff make quick and effective decisions in emergency situations.

[0115] Next, the system implements a dynamic permission control mechanism to ensure that only authorized personnel can access or manipulate sensitive information. This mechanism is based on the Role-Based Access Control (RBAC) model, which sets detailed permissions to determine which users can view or handle specific patient information. For example, only on-duty doctors and nurses can access emergency patient diagnosis and treatment information, while ordinary administrative staff cannot access such sensitive data. Such permission management can effectively prevent information leakage and ensure proper protection of patient privacy.

[0116] This step combines visualization technology and dynamic permission control mechanisms to improve the efficiency and security of medical resource management. With the help of visualization tools, medical staff can quickly access key information and make timely medical decisions, reducing delays caused by information asymmetry. In addition, dynamic permission control effectively maintains the security of patient information, complies with modern data protection laws and regulations, and enhances patients' trust in the medical system. This combination not only improves emergency medical efficiency but also sets a good industry standard for patient privacy protection.

[0117] In this step, the system uses a matching result generation method based on visualization technology to effectively integrate the decrypted complete data with the medical resource matching results. This process includes using visualization tools such as Tableau or D3.js to create an intuitive graphical interface that displays patient information and the configuration of corresponding medical resources. Through graphical means, medical personnel can quickly understand the status of resource allocation, patient vital signs data, and the medical resources required for emergency treatment, so as to make timely decisions in emergency situations.

[0118] At the same time, the system implements a dynamic permission control mechanism to ensure data security. This mechanism relies on a complex user identity verification system to ensure that only authorized personnel can access specific sensitive information. For example, the roles of medical personnel will be clearly defined, and only on-duty doctors and nurses can view detailed information about emergency patients, while other staff can only access non-sensitive business data. Dynamic permission control can be achieved through a role-based access control (RBAC) or policy-based access control (PBAC) system, ensuring strict access control of sensitive data.

[0119] Finally, through the integration of information through the visualization interface and dynamic permission control, the system generates the final emergency patient information data matching scheme. Medical personnel can not only quickly grasp the patient's health status through the integrated information, but also conveniently obtain the required medical resource arrangement, thereby improving efficiency in emergency situations. This integration not only maintains the privacy of data, but also effectively supports medical decision-making, improving the response speed and overall service capability of the hospital in emergency treatment.

[0120] As can be seen, according to the electronic medical records, image data and real-time vital signs data of emergency patients, real-time multi-source heterogeneous data is obtained, and a time-synchronized multi-source data set is generated through an adaptive data format conversion algorithm. For the multi-source data set, a feature extraction network is used to extract patient condition features and generate patient priority scores. Based on the patient priority scores and condition features, a medical resource matching algorithm based on graph neural networks is used to adjust the resource allocation scheme in real time, generating a preliminary medical resource matching result. For the medical resource matching result, a data security transmission protocol based on homomorphic encryption is used, and a differential privacy protection mechanism is used to generate the final emergency patient information data matching scheme, thereby meeting the demand for fast, safe and effective processing of data in emergency situations, and improving the quality and efficiency of emergency medical services.

[0121] Another embodiment of the present application provides an Internet-based emergency patient information data matching system, as shown in Figure 3 , which can include: The acquisition module 301 is configured to acquire multi-source heterogeneous data in real time according to the electronic medical record, image data and real-time vital sign data of the emergency patient by using a data acquisition framework based on edge computing, convert the data from different sources into a standardized data format by using a self-adaptive data format conversion algorithm, and generate a time-synchronized multi-source data set. The extraction module 302 is configured to extract the patient's condition characteristics by using a feature extraction network based on a multi-head attention mechanism on the time-synchronized multi-source data set, and generate a patient priority score by combining the severity of the condition and the urgency of treatment through a priority evaluation model based on fuzzy logic. The matching module 303 is configured to dynamically match optimal medical resources by using a medical resource matching algorithm based on a graph neural network according to the patient priority score and the condition characteristics, combining the hospital department, doctor expertise and equipment availability information, adjusting the resource allocation scheme in real time through a distributed task scheduling framework, and generating a preliminary medical resource matching result. The transmission module 304 is configured to ensure the secure transmission of patient information in an Internet environment by using a data security transmission protocol based on homomorphic encryption on the medical resource matching result, and generating a final emergency patient information data matching scheme by adding noise to sensitive data through a differential privacy protection mechanism.

[0122] As can be seen, multi-source heterogeneous data is acquired in real time according to the electronic medical record, image data and real-time vital sign data of the emergency patient, a time-synchronized multi-source data set is generated by using a self-adaptive data format conversion algorithm, the condition characteristics of the patient are extracted by using a feature extraction network on the multi-source data set, a patient priority score is generated, a preliminary medical resource matching result is generated by using a medical resource matching algorithm based on a graph neural network according to the patient priority score and the condition characteristics, the resource allocation scheme is adjusted in real time, and a final emergency patient information data matching scheme is generated by using a data security transmission protocol based on homomorphic encryption on the medical resource matching result and through a differential privacy protection mechanism, thereby meeting the demand for fast, safe and effective processing of data in an emergency scene and improving the quality and efficiency of emergency medical services.

[0123] The embodiment of the application also provides a storage medium in which a computer program is stored, wherein the computer program is configured to execute the steps in any of the method embodiments when running.

[0124] Specifically, in the present embodiment, the above-mentioned storage medium can be configured to store a computer program for executing the following steps: S201, based on the electronic medical record, image data and real-time vital sign data of the emergency patient, a data acquisition framework based on edge computing is used to acquire multi-source heterogeneous data in real time, different sources of data are uniformly converted into a standardized data format through an adaptive data format conversion algorithm, and a time-synchronized multi-source data set is generated; S202, for the time-synchronized multi-source data set, a feature extraction network based on a multi-head attention mechanism is used to extract the patient's condition characteristics, a priority evaluation model based on fuzzy logic is used to combine the severity of the disease and the urgency of treatment, and a patient priority score is generated; S203, according to the patient priority score and the condition characteristics, a medical resource matching algorithm based on a graph neural network is used, combined with hospital departments, doctor expertise and equipment availability information, to dynamically match the optimal medical resources, through a distributed task scheduling framework, the resource allocation scheme is adjusted in real time, and a preliminary medical resource matching result is generated. S204, for the medical resource matching result, a data security transmission protocol based on homomorphic encryption is used to ensure the secure transmission of patient information in the Internet environment, and a differential privacy protection mechanism is used to add noise to sensitive data, and a final emergency patient information data matching scheme is generated.

[0125] As can be seen, based on the electronic medical record, image data and real-time vital sign data of the emergency patient, multi-source heterogeneous data is acquired in real time, and a time-synchronized multi-source data set is generated through an adaptive data format conversion algorithm. For the multi-source data set, a feature extraction network is used to extract the patient's condition characteristics and generate a patient priority score. According to the patient priority score and the condition characteristics, a medical resource matching algorithm based on a graph neural network is used to adjust the resource allocation scheme in real time and generate a preliminary medical resource matching result. For the medical resource matching result, a data security transmission protocol based on homomorphic encryption is used, and a differential privacy protection mechanism is used to generate a final emergency patient information data matching scheme, so as to meet the demand for fast, safe and effective processing of data in the emergency scene, and improve the quality and efficiency of emergency medical services.

[0126] The embodiment of the application also provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any of the method embodiments.

[0127] Specifically, the electronic device described above can further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0128] Specifically, in the embodiment, the processor can be configured to execute the following steps through the computer program: S201, according to the electronic medical record of the emergency patient, image data and real-time vital sign data, a data acquisition framework based on edge computing is used to acquire multi-source heterogeneous data in real time, different sources of data are uniformly converted into a standardized data format through an adaptive data format conversion algorithm, and a time-synchronized multi-source data set is generated; S202, for the time-synchronized multi-source data set, a feature extraction network based on multi-head attention mechanism is used to extract the patient's condition characteristics, a priority evaluation model based on fuzzy logic is used to combine the severity of the disease and the urgency of treatment, and a patient priority score is generated; S203, according to the patient priority score and the condition characteristics, a medical resource matching algorithm based on graph neural network is used, combined with hospital departments, doctor expertise and equipment availability information, the optimal medical resource is dynamically matched, the resource allocation scheme is adjusted in real time through a distributed task scheduling framework, and a preliminary medical resource matching result is generated. S204, for the medical resource matching result, a data security transmission protocol based on homomorphic encryption is used to ensure the secure transmission of patient information in the Internet environment, and a differential privacy protection mechanism is used to add noise to sensitive data, and a final emergency patient information data matching scheme is generated.

[0129] As can be seen, according to the electronic medical record of the emergency patient, image data and real-time vital sign data, multi-source heterogeneous data is acquired in real time, a time-synchronized multi-source data set is generated through an adaptive data format conversion algorithm, the condition characteristics of the patient are extracted through a feature extraction network, and a patient priority score is generated. According to the patient priority score and the condition characteristics, a medical resource matching algorithm based on graph neural network is used to adjust the resource allocation scheme in real time, and a preliminary medical resource matching result is generated. For the medical resource matching result, a data security transmission protocol based on homomorphic encryption is used, and a differential privacy protection mechanism is used to generate a final emergency patient information data matching scheme, so as to meet the demand for fast, safe and effective processing of data in emergency scenes, and improve the quality and efficiency of emergency medical services.

[0130] The above embodiments according to the drawings illustrate the structure, features and effects of the present application. The above description is only the preferred embodiment of the present application, but the present application is not limited by the drawings. Any changes or modifications made in accordance with the concept of the present application, or equivalent embodiments with equivalent changes, are still within the scope of the present application.

Claims

1. A method for matching emergency patient information data based on the Internet, characterized in that: The method comprises: Adopting an edge computing-based data collection framework, it can acquire multi-source heterogeneous data including electronic medical records, imaging materials, and real-time vital signs data of emergency patients in real time, and generate time-synchronized multi-source data sets through an adaptive data format conversion algorithm. For time-synchronized multi-source datasets, a feature extraction network based on a multi-head attention mechanism is used to extract the text features of the patient's symptom descriptions, the visual features of the imaging data, and the temporal features of the vital signs data. Through the cross-modal attention mechanism, a multimodal disease feature representation is generated. For multimodal disease characteristics, a disease severity quantification model based on a deep neural network is used to calculate the patient's disease severity score. This disease severity score is combined with the patient's real-time vital signs data, and an urgency assessment algorithm based on time series analysis is used to predict the time window for disease deterioration. Through a dynamic weight adjustment mechanism, a treatment urgency score is generated; The severity of the disease and the urgency of treatment are scored, and a priority assessment model based on fuzzy logic is used to generate a patient priority score. Based on the patient priority score and disease characteristics, a medical resource matching algorithm based on a graph neural network is used to dynamically match the optimal medical resources and generate preliminary medical resource matching results. For the medical resource matching results, a data security transmission protocol based on homomorphic encryption is adopted, and the final emergency patient information data matching solution is generated through the differential privacy protection mechanism.

2. The method according to claim 1, characterized in that The edge computing-based data acquisition framework is used to acquire multi-source heterogeneous data including electronic medical records, imaging data, and real-time vital sign data of emergency patients in real time. Through an adaptive data format conversion algorithm, a time-synchronized multi-source data set is generated, including: Based on the electronic medical records, imaging data, and real-time vital signs data of emergency patients, an edge computing-based data collection framework is adopted to obtain multi-source heterogeneous data in real time through distributed edge nodes. Each edge node is equipped with a lightweight data caching mechanism to ensure the real-time and continuity of data collection. For multi-source heterogeneous data, a deep learning-based format recognition model is used to identify the format types of different data sources. Through an adaptive parsing algorithm, data in different formats are parsed into structured intermediate representations to generate a preliminary parsed data set. For the initially parsed dataset, a standardized conversion algorithm based on a combination of a rule engine and machine learning is used to uniformly convert data from different sources into a standardized data format. Through a context-aware mapping rule library, the conversion logic is dynamically adjusted to generate a standardized dataset. For the standardized data set, a time synchronization algorithm based on dynamic time warping is used to eliminate the timestamp differences between different data sources. The missing data is supplemented by interpolation filling method to generate a time-synchronized multi-source data set.

3. The method according to claim 2, characterized in that According to the patient priority score and condition characteristics, a medical resource matching algorithm based on a graph neural network is used to dynamically match the optimal medical resources and generate preliminary medical resource matching results, including: A medical resource graph is constructed based on hospital departments, physician expertise, and equipment availability information. Nodes in the graph represent departments, physicians, or equipment, and edges represent collaborations or dependencies between resources. By introducing a real-time resource status update mechanism, the graph's node and edge information is dynamically maintained, generating a real-time medical resource graph. Based on the patient's priority score and condition characteristics, a matching algorithm based on a graph neural network is used to associate patient needs with nodes in the medical resource graph. Through the attention mechanism, the matching degree between the patient and the resource node is calculated to generate a preliminary patient-resource matching relationship. For the initial patient-resource matching relationship, a dynamic adjustment method based on a distributed task scheduling framework is adopted. This method combines the real-time availability of resources and task priorities to resolve resource allocation conflicts. Through the Nash equilibrium algorithm in game theory, the resource allocation scheme is optimized to generate a conflict-free medical resource scheduling plan. For conflict-free medical resource scheduling plans, a matching result generation method based on visualization technology is adopted to feed back the matching results to medical staff in real time. Through the feedback correction mechanism, the matching algorithm parameters are dynamically adjusted to generate the final medical resource matching results.

4. The method according to claim 3, characterized in that The medical resource matching results are generated by using a data security transmission protocol based on homomorphic encryption and a differential privacy protection mechanism to generate the final emergency patient information data matching solution, including: For patient information in the medical resource matching results, a data security transmission protocol based on homomorphic encryption is used to encrypt sensitive data into ciphertext. A lightweight key management mechanism is used to ensure the secure transmission of encrypted data in an Internet environment and generate encrypted data transmission packages. For sensitive fields in encrypted data transmission packets, a differential privacy protection method based on the Laplace mechanism is used. The amount of noise added is dynamically calculated according to the data sensitivity level. An adaptive noise allocation algorithm is used to add noise to key fields to generate privacy-protected data packets. For privacy-protected data packets, a homomorphic decryption algorithm is used at the receiving end to restore the data. A hash function-based integrity verification mechanism is used to verify whether the data has been tampered with during transmission, generating complete decrypted data. For the complete decrypted data, a matching result generation method based on visualization technology is used to integrate the medical resource matching results with patient information. Through a dynamic permission control mechanism, it is ensured that only authorized personnel can access sensitive information, and the final emergency patient information data matching plan is generated.

5. An Internet-based emergency patient information data matching system, characterized in that: The system comprises: The acquisition module is used to acquire multi-source heterogeneous data including electronic medical records, imaging materials, and real-time vital sign data of emergency patients in real time using an edge computing-based data acquisition framework. It generates a time-synchronized multi-source data set through an adaptive data format conversion algorithm. The extraction module is used to extract text features of patient symptom descriptions, visual features of imaging data, and temporal features of vital signs data from time-synchronized multi-source datasets using a feature extraction network based on a multi-head attention mechanism. This module then generates a multimodal disease feature representation through a cross-modal attention mechanism. A generation module is used to represent multimodal disease characteristics and use a disease severity quantification model based on a deep neural network to calculate the patient's disease severity score. This disease severity score is combined with the patient's real-time vital sign data and an urgency assessment algorithm based on time series analysis to predict the time window for disease deterioration. Through a dynamic weight adjustment mechanism, a treatment urgency score is generated; The matching module is used to score the severity of the disease and the urgency of treatment. It uses a priority assessment model based on fuzzy logic to generate a patient priority score. Based on the patient priority score and disease characteristics, a medical resource matching algorithm based on a graph neural network is used to dynamically match the optimal medical resources and generate preliminary medical resource matching results. The encryption module is used to match the medical resources results. It adopts a data security transmission protocol based on homomorphic encryption and a differential privacy protection mechanism to generate the final emergency patient information data matching solution.

6. The system according to claim 5, characterized in that The acquisition module is specifically used to: Based on the electronic medical records, imaging data, and real-time vital signs data of emergency patients, an edge computing-based data collection framework is adopted to obtain multi-source heterogeneous data in real time through distributed edge nodes. Each edge node is equipped with a lightweight data caching mechanism to ensure the real-time and continuity of data collection. For multi-source heterogeneous data, a deep learning-based format recognition model is used to identify the format types of different data sources. Through an adaptive parsing algorithm, data in different formats are parsed into structured intermediate representations to generate a preliminary parsed data set. For the initially parsed dataset, a standardized conversion algorithm based on a combination of a rule engine and machine learning is used to uniformly convert data from different sources into a standardized data format. Through a context-aware mapping rule library, the conversion logic is dynamically adjusted to generate a standardized dataset. For the standardized data set, a time synchronization algorithm based on dynamic time warping is used to eliminate the timestamp differences between different data sources. The missing data is supplemented by interpolation filling method to generate a time-synchronized multi-source data set.

7. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 4 when run.

8. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 4.

Images