A medical service improvement method and system based on out-of-town medical treatment data analysis

Abstract

The application discloses a medical service improvement method and system based on out-of-town medical treatment data analysis, and belongs to the technical field of computers. Through the technical scheme, the application realizes accurate mapping of individual needs of patients and medical resources, and effectively improves the accuracy of out-of-town medical treatment recommendation. The method solves the problem of data dimension mismatch in the traditional recommendation system through deep feature fusion, improves the rationality of medical resource recommendation through a semantic coherent description information generation mechanism, and provides a hospital selection scheme that is more in line with the actual needs of patients.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method and system for improving medical services based on cross-regional medical treatment data analysis. Background Technology

[0002] With socio-economic development and increased population mobility, cross-regional medical treatment is becoming increasingly common. Patients seeking medical services outside their place of insurance often face challenges such as information asymmetry, unfamiliarity with procedures, and low resource matching rates. How to efficiently and accurately match suitable medical resources for patients seeking treatment in other locations, thereby improving their medical experience and outcomes, has become an urgent problem to be solved in the medical service field.

[0003] Existing cross-regional medical treatment service systems or platforms typically focus on simplifying processes (such as registration and settlement) or providing basic hospital directory information. They have significant limitations in handling patients' individualized and complex medical needs.

[0004] Furthermore, healthcare information systems in different regions exhibit significant differences in data processing capabilities, data models, and service focuses. Currently, there is a lack of effective mechanisms to intelligently adapt and utilize these heterogeneous computing node resources distributed across the destination based on the patient's specific situation and needs. This results in cross-regional medical recommendation services often remaining at the information display level, failing to achieve in-depth data mining and precise matching based on multi-dimensional data, and thus failing to maximize the satisfaction of patients' personalized, high-quality medical needs. Summary of the Invention

[0005] This application provides a method and system for improving medical services based on cross-regional medical treatment data analysis. This method and system can improve the accuracy of medical resource matching, optimize the efficiency of cross-regional medical treatment decisions, and enhance the personalized service experience for patients. The technical solution is as follows:

[0006] This application provides a method for improving medical services based on cross-regional medical treatment data analysis. The technical solution is as follows: In response to a target's request for cross-regional medical treatment, the target's current physiological state, medical reference information, and reason for cross-regional medical treatment are determined; based on the target's current physiological state and reason for cross-regional medical treatment, a target node matching the physiological state and reason for cross-regional medical treatment is determined from multiple candidate nodes associated with the medical destination. Different candidate nodes have different computing power and hospital recommendation modes. The medical destination is obtained from the cross-regional medical treatment request; the target's electronic medical record, medical reference information, and reason for cross-regional medical treatment are input into the target node, and the target node determines the target hospital to recommend to the target from multiple candidate hospitals at the medical destination.

[0007] Furthermore, this application proposes that the medical reference information includes medical preferences and medical needs, and the reasons for seeking medical treatment in other locations include subjective reasons for seeking medical treatment in other locations and reference reasons for seeking medical treatment in other locations. In response to the target subject's request for medical treatment in other locations, the process of determining the target subject's physiological state, medical reference information, and reasons for seeking medical treatment in other locations includes: in response to the target subject's request for medical treatment in other locations, obtaining the target subject's physiological data, subjective reasons for seeking medical treatment in other locations, historical medical data, and medical feedback data, wherein the medical feedback data is used to represent the hospital's evaluation of the target subject; determining the target subject's physiological state based on the physiological data; determining medical preferences based on historical medical data; and determining reference reasons for seeking medical treatment in other locations based on historical medical data and medical feedback data.

[0008] Furthermore, this application proposes that historical medical data includes historical hospitals and historical medical visit times. Based on historical medical data, medical preferences are determined, including: obtaining information on the primary hospital of the historical hospital, which is used to describe the historical hospital; determining the target subject's basic medical visit time pattern based on historical medical visit times; determining the target subject's reference medical visit time pattern based on the primary hospital information and historical medical visit times; determining medical preferences based on the primary hospital information, basic medical visit time pattern, and reference medical visit time pattern; and determining the reasons for considering out-of-town medical treatment based on historical medical data and medical feedback data, including: obtaining information on the secondary hospital of the historical hospital, which is used to describe the target subject's medical treatment at the historical hospital; and determining the reasons for considering out-of-town medical treatment based on the secondary hospital information and medical feedback data.

[0009] Furthermore, this application proposes determining medical preferences based on information from the first hospital, basic medical treatment time patterns, and reference medical treatment time patterns. This includes: determining the target individual's hospital and department preferences based on the first hospital information; determining the target individual's medical treatment time preferences based on the basic and reference medical treatment time patterns; and concatenating the target individual's hospital, department, and medical treatment time preferences to obtain the medical preferences. Second hospital information includes the number of medical treatment payments, the amount of medical treatment payments, the distance traveled for medical treatment, and medical treatment effect information. Based on the second hospital information and medical feedback data, the application proposes determining the reasons for seeking medical treatment in other locations. This includes: determining medical fatigue description information based on the number of medical treatment payments and the distance traveled for medical treatment; determining medical effect description information based on the amount of medical treatment payments and the medical treatment effect information; and determining the reasons for seeking medical treatment in other locations based on the medical fatigue description information, the medical effect description information, and the medical feedback data.

[0010] Furthermore, this application also proposes to determine the target node that matches the physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the destination of medical treatment, based on the physiological state of the target object and the reason for seeking medical treatment in another location. This includes: determining multiple reference nodes from multiple candidate nodes based on the physiological state, wherein the multiple reference nodes are candidate nodes that match the physiological state; and determining the target node from multiple reference nodes based on the reason for seeking medical treatment in another location.

[0011] Furthermore, this application proposes determining multiple reference nodes from multiple candidate nodes based on physiological states, wherein the multiple reference nodes are candidate nodes that match the physiological states, including: determining first node description information for each candidate node, the first node description information being used to describe the associated physiological states of the corresponding candidate node; determining a first matching degree between the physiological states and each candidate node based on the physiological states and the first node description information of each candidate node; determining the candidate nodes among the multiple candidate nodes whose first matching degree with the physiological states is greater than or equal to the first matching degree threshold as reference nodes, thus obtaining multiple reference nodes; and determining a target node from multiple reference nodes based on the reason for seeking medical treatment in a different location, including: determining second node description information for each reference node, the second node description information being used to describe the associated medical treatment reason of the reference node; determining a second matching degree between the reason for seeking medical treatment in a different location and each reference node based on the reason for seeking medical treatment in a different location and the associated medical treatment reason of each described reference node; and determining the reference node among the multiple reference nodes with the highest second matching degree with the reason for seeking medical treatment in a different location as the target node.

[0012] Furthermore, this application proposes a method for determining a target hospital to recommend to a target individual from multiple hospitals at the destination of medical treatment through a target node. This includes: determining multiple candidate hospitals from multiple hospitals based on electronic medical records, medical reference information, and hospital descriptions of each hospital through the target node; sending the target individual's electronic medical records, medical reference information, and reasons for seeking medical treatment in other locations to the hospital nodes of each candidate hospital, so that the hospital nodes of each candidate hospital can determine the degree of matching for treatment, the willingness to receive treatment, and the reasons for generating the degree of matching for treatment and the willingness to receive treatment based on the target individual's electronic medical records, medical reference information, and reasons for seeking medical treatment in other locations; obtaining the degree of matching for treatment, the willingness to receive treatment, and the reasons for generating the degree of matching for treatment and the willingness to receive treatment from each candidate hospital through the target node; and determining the target hospital from multiple candidate hospitals based on the degree of matching for treatment, the willingness to receive treatment, the reasons for generating the willingness to receive treatment, and the medical reference information of each candidate hospital through the target node.

[0013] Furthermore, this application proposes to determine multiple candidate hospitals from multiple hospitals through a target node, based on electronic medical records, medical reference information, and hospital descriptions of each hospital. This includes: determining a third degree of matching between the target object and each hospital based on electronic medical records, medical reference information, and hospital descriptions of each hospital through the target node; identifying hospitals with a third degree of matching greater than or equal to a second degree of matching threshold as candidate hospitals, thus obtaining multiple candidate hospitals; and determining a target hospital from multiple candidate hospitals through the target node based on the patient reception matching degree, patient reception intention, generation reason, and medical reference information of each candidate hospital. This includes: determining a patient reception intention score for the target object from each candidate hospital based on the patient reception matching degree, patient reception intention, and generation reason of each candidate hospital through the target node; determining a fourth degree of matching between the target object and each candidate hospital based on the patient reception intention and medical reference information of each candidate hospital; and determining the target hospital from multiple candidate hospitals based on the patient reception intention score and the fourth degree of matching.

[0014] Furthermore, this application proposes determining the third matching degree between a target object and various hospitals based on electronic medical records, medical reference information, and hospital description information of each hospital through target nodes. This includes: generating object description information of the target object based on electronic medical records and medical reference information through target nodes; determining the third matching degree between the target object and various hospitals based on the object description information of the target object and the hospital description information of each hospital; and determining the patient admission intention score of each candidate hospital for the target object based on the patient admission matching degree, patient admission intention, and generation reason of each candidate hospital through target nodes. This includes: inputting the patient admission matching degree, patient admission intention, and generation reason of each candidate hospital into the patient admission intention score determination model, and using the patient admission intention score determination model to determine the patient admission matching degree. The system processes the degree of interest, willingness to receive treatment, and reasons for the generation of each candidate hospital to obtain a score indicating the willingness of each candidate hospital to receive treatment from the target individual. Based on the willingness to receive treatment and medical reference information of each candidate hospital, the system determines the fourth degree of matching between the target individual and each candidate hospital. This includes: inputting the willingness to receive treatment and medical reference information of each candidate hospital into the matching degree determination model, processing the willingness to receive treatment and medical reference information through the matching degree determination model, and outputting the fourth degree of matching between the target individual and each candidate hospital; and determining the target hospital from multiple candidate hospitals based on the willingness to receive treatment score and the fourth degree of matching, including: fusing the willingness to receive treatment score and the fourth degree of matching to obtain the target matching score of each candidate hospital; and determining the candidate hospital with the highest target matching score among multiple candidate hospitals as the target hospital.

[0015] Furthermore, this application proposes generating object description information for a target object based on electronic medical records and medical reference information, including: extracting features from the electronic medical records and medical reference information to obtain electronic medical record features and reference information features from the medical reference information; fusing the electronic medical record features and reference information features to obtain object features of the target object; performing multi-round iterative decoding on the object features to obtain object description information for the target object; and determining a third matching degree between the target object and each hospital based on the object description information of the target object and the hospital description information of each hospital, including: extracting features from the object description information of the target object and the hospital description information of each hospital to obtain first description information features of the object description information and second description information features of each hospital; and determining the feature similarity between the first description information features and the second description information features corresponding to each hospital as the third matching degree between the target object and each hospital.

[0016] On the one hand, a medical service improvement system based on cross-regional medical treatment data analysis is provided, the system comprising:

[0017] The first determining module is used to determine the target object's current physiological state, medical reference information, and reason for seeking medical treatment in another location in response to the target object's request for medical treatment in another location.

[0018] The second determining module is used to determine a target node that matches the current physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the destination of medical treatment, based on the current physiological state of the target object and the reason for seeking medical treatment in another location. Different candidate nodes have different computing power and hospital recommendation modes, and the destination of medical treatment is obtained from the request for seeking medical treatment in another location.

[0019] The third determining module is used to input the target object's electronic medical record, the medical treatment reference information, and the reason for seeking medical treatment in another location into the target node, and to determine the target hospital to recommend to the target object from multiple candidate hospitals at the destination of medical treatment through the target node.

[0020] On one hand, a computer device is provided, the computer device including one or more processors and one or more memories, the one or more memories storing at least one computer program, the computer program being loaded and executed by the one or more processors to implement the medical service improvement method based on cross-regional medical treatment data analysis.

[0021] On the one hand, a computer-readable storage medium is provided, wherein at least one computer program is stored in the computer-readable storage medium, the computer program being loaded and executed by a processor to implement the medical service improvement method based on cross-regional medical treatment data analysis.

[0022] On the one hand, a computer program product or computer program is provided, which includes program code stored in a computer-readable storage medium. The processor of a computer device reads the program code from the computer-readable storage medium and executes the program code, causing the computer device to perform the aforementioned method for improving medical services based on cross-regional medical treatment data analysis.

[0023] As can be seen from the above, the medical service improvement method and system based on cross-regional medical treatment data analysis provided in this application integrates patients' physiological status, historical preferences and reasons for seeking medical treatment, dynamically matches node resources with different computing power and recommendation modes, and combines multi-dimensional data mining and two-way matching mechanism to effectively solve the problem of mismatch between medical resources and patient needs. It has the advantages of improving matching accuracy, optimizing decision-making efficiency and improving service experience. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of the implementation environment of a medical service improvement method based on cross-regional medical treatment data analysis provided in an embodiment of this application;

[0026] Figure 2 This is a flowchart of a method for improving medical services based on cross-regional medical treatment data analysis provided in an embodiment of this application;

[0027] Figure 3 This is a flowchart of determining the reasons for seeking medical treatment in a different location, provided in an embodiment of this application;

[0028] Figure 4 This is a flowchart of determining the target node provided in an embodiment of this application;

[0029] Figure 5 This is a flowchart of the process for determining the target hospital provided in an embodiment of this application;

[0030] Figure 6 This is a schematic diagram of the structure of a medical service improvement system based on cross-regional medical treatment data analysis provided in an embodiment of this application;

[0031] Figure 7 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0033] In this application, the terms "first," "second," etc., are used to distinguish identical or similar items with essentially the same function. It should be understood that there is no logical or temporal dependency between "first," "second," and "nth," nor are there any restrictions on quantity or execution order.

[0034] Artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain better results.

[0035] Machine learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory, among others. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills, and reorganize existing knowledge sub-models to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence; its applications span all areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and instruction-based learning.

[0036] Out-of-town medical treatment: Out-of-town medical treatment refers to the behavior of patients leaving their medical insurance coverage area (usually their registered residence or place of work) to seek medical treatment in other regions. This phenomenon is usually driven by uneven distribution of medical resources, patients' demand for high-quality medical services, or caused by population mobility (such as migrant work, elderly care, etc.).

[0037] A Medical Information System (MIS) is a comprehensive system that integrates medical data, optimizes medical processes, and improves service efficiency through information technology. Its core objectives include supporting clinical decision-making, improving patient management, and promoting inter-institutional collaboration.

[0038] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0039] Figure 1This is a schematic diagram illustrating the implementation environment of a medical service improvement method based on cross-regional medical treatment data analysis provided in this application embodiment. See also... Figure 1 This implementation environment may include node 110 and server 140.

[0040] Node 110 is connected to server 140 via a wireless or wired network. Optionally, node 110 can be a laptop, desktop computer, etc., but is not limited to these. Node 110 has applications installed and running that support improvements in healthcare services based on remote medical data analysis.

[0041] Server 140 is a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms. Server 140 can provide background services for applications running on node 110.

[0042] In related technologies, patients seeking medical treatment across regions often face the problem of inefficient matching of medical resources. Existing service platforms typically only provide basic hospital directories or simplified registration processes, lacking the ability to deeply analyze individual patient needs. When a patient needs urgent cross-provincial medical treatment due to a sudden illness, existing systems struggle to quickly identify the urgency of the patient's condition and cannot dynamically match patients based on the real-time capacity of medical institutions. For example, when a myocardial infarction patient needs to quickly find a hospital with interventional treatment capabilities in a different location, traditional systems can only display a list of hospitals and cannot comprehensively assess the hospital's current bed availability, specialty strength, and suitability for the patient's condition.

[0043] To address these issues, researchers discovered that the key to matching medical resources lies in constructing a dynamic adaptation mechanism. First, they observed differences in computing power among medical information systems in different regions; some nodes excel at real-time data processing, while others possess the capability for complex model computation. Second, they found that the reasons patients seek medical care directly influence resource selection strategies; for example, cancer patients prioritize specialist rankings, while emergency patients need to consider response speed. By establishing a mapping relationship between patient characteristics and computing node capabilities, they attempted to match patient data with node characteristics across multiple dimensions, ultimately forming a phased screening mechanism: first, filtering computing nodes based on physiological state; then, determining the optimal node based on the reason for seeking medical care; and finally, completing hospital matching through a recommendation algorithm for the selected nodes.

[0044] Therefore, this application proposes a method for improving medical services based on cross-regional medical treatment data analysis, see [link to relevant documentation]. Figure 2 Taking the server as the executing entity as an example, the methods include:

[0045] 201. In response to the target object's request for medical treatment in a different location, the server determines the target object's current physiological state, medical reference information, and the reason for seeking medical treatment in a different location;

[0046] 202. Based on the target object's current physiological state and the reason for seeking medical treatment in another location, the server determines the target node that matches the physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the medical destination. Different candidate nodes have different computing power and hospital recommendation modes. The medical destination is obtained from the medical treatment request in another location.

[0047] 203. The server inputs the target object's electronic medical record, medical reference information, and reasons for seeking medical treatment in another location into the target node. The target node then determines the target hospital to recommend to the target object from multiple candidate hospitals in the destination of medical treatment.

[0048] The current physiological state refers to real-time health indicators obtained through wearable devices or electronic medical record analysis. Specifically, this can be achieved using heart rate variability analysis combined with medical record keyword extraction, used to assess the urgency of medical needs. Medical reference information includes a preference model formed from historical medical records. This can be generated by processing historical medical data using collaborative filtering algorithms, reflecting the patient's personalized needs. Candidate nodes refer to heterogeneous computing units deployed at the medical destination. This can be implemented using a containerized microservice architecture, with each node configured with an independent resource evaluation algorithm and recommendation strategy. The hospital recommendation model refers to the medical institution evaluation system. Specifically, this can be achieved using a multi-index decision model based on the analytic hierarchy process (AHP), used to quantify the matching degree between hospital and patient needs.

[0049] Specifically, when a patient initiates a request for out-of-town medical treatment, the system first parses the diagnostic records and real-time monitoring data in the electronic medical record to generate a physiological status assessment report containing disease severity classification indicators. Simultaneously, it extracts the patient's medical records from the past three years in their insured location, analyzing characteristics such as the proportion of patients choosing tertiary hospitals and the interval between follow-up visits, constructing a reference information model that includes departmental preferences and hospital level preferences. Based on the patient's stated reasons for referral and the system's inferred potential needs, three candidate nodes are selected from multiple pre-deployed computing nodes at the destination: the first node is configured with a real-time bed availability query interface, suitable for emergency scenarios; the second node carries a disease prognosis prediction model, suitable for chronic disease follow-up visits; and the third node integrates multi-specialty ranking data to meet elective surgery needs. By comparing the time sensitivity of the patient's physiological status with the processing capacity of each node, the node with the fastest response speed is prioritized as the initial screening result. Then, based on the keyword "seeking specialist treatment" in the referral reason, the target node equipped with a specialist ranking analysis model is finally selected. This node matches the patient's tumor staging data with the radiotherapy equipment configuration and expert consultation schedule of the destination hospital, generates a recommendation list containing predicted treatment success rates, and selects the top-tier tertiary oncology hospital with the highest comprehensive score as the target hospital.

[0050] Compared to related technologies, traditional methods use fixed algorithms to process all medical requests, failing to dynamically adjust computing resources based on the urgency of the condition. This solution constructs a configurable node pool, automatically assigning cardiovascular emergency patients to nodes with real-time data processing capabilities, while rare disease patients are routed to deep computing nodes equipped with knowledge graphs, achieving precise matching of computing resources to medical needs. Existing systems typically analyze static hospital data separately; this solution innovatively correlates patients' physiological states, historical preferences, and the hospital's dynamic reception capacity in multiple dimensions. For example, during flu outbreaks, it automatically increases the weight of fever clinic bed turnover rates, making recommendations more consistent with reality.

[0051] Through the above technical solutions, this application achieves intelligent scheduling of medical computing resources, reducing the node matching time for emergency patients to one-fifth of that of traditional methods. By dynamically combining physiological state and reason for seeking medical treatment as a dual screening mechanism, the accuracy of hospital recommendations for complex cases is significantly improved, especially for cancer patients transferred across provinces, where the consistency between the recommended target hospital and the actual hospital visited reaches a high level. The on-demand call mechanism for heterogeneous computing nodes effectively balances the system load, maintaining stable service quality even when handling concentrated medical requests caused by large-scale outbreaks of epidemics.

[0052] This application further proposes that medical treatment reference information includes medical treatment preferences and needs, and that reasons for seeking medical treatment in other locations include subjective reasons and suggested reasons. In response to a target individual's request for medical treatment in another location, the application determines the target individual's physiological state, medical treatment reference information, and reasons for seeking medical treatment in another location. (See [link to application]). Figure 3 Taking the server as the executing entity as an example, the steps include the following:

[0053] 301. In response to the target object's request for medical treatment in another location, the server obtains the target object's physiological data, subjective reasons for seeking medical treatment in another location, historical medical data, and medical feedback data. The medical feedback data is used to represent the hospital's evaluation of the target object.

[0054] 302. The server determines the physiological state of the target object based on physiological data;

[0055] 303. The server determines medical preferences based on historical medical data; and determines the reasons for seeking medical treatment in other locations based on historical medical data and medical feedback data.

[0056] Among these, "medical treatment preference" refers to a patient's long-term, formed tendencies in medical treatment behavior. This can be achieved through pattern analysis using information from the primary hospital where the patient has previously sought treatment and historical treatment times. By analyzing hospital level, department type, and the distribution of treatment times, patient preference for medical institutions can be identified. "Medical treatment needs" refers to the specific demands within the current medical treatment scenario. This can be achieved through feature extraction from diagnostic records and treatment plans in electronic medical records, reflecting the patient's functional requirements for medical services. "Subjective reasons for seeking medical treatment in other locations" refers to the patient's actively stated motivations for cross-regional medical treatment. This can be achieved through keyword extraction from the patient's input text using natural language processing technology, such as referral needs and preferences for specialist resources. "Reference reasons for seeking medical treatment in other locations" refers to potential motivations derived from objective data analysis. This can be achieved through correlation analysis using information from the secondary hospital where the patient has previously sought treatment and medical feedback data, such as identifying patient dissatisfaction with the service efficiency or cost structure of previously visited hospitals.

[0057] Specifically, when a request for out-of-town medical treatment is triggered, the patient's physiological data is collected in real time to assess their current health status, and the subjective reasons for seeking medical treatment in another location are stored in a structured manner. Simultaneously, hospital attributes and appointment time series from historical medical data are extracted, and a basic appointment time pattern is generated using a time-series pattern mining algorithm. This, combined with hospital level characteristics, forms a reference appointment time pattern, ultimately deriving medical preferences that include hospital type preferences, department selection tendencies, and patterns in appointment times. Medical feedback data is jointly analyzed with data such as payment frequency and travel distance from historical medical records. By calculating a medical fatigue index and effect deviation, objective motivations for seeking medical treatment in another location that are not explicitly stated by the patient are identified, such as the transportation burden caused by frequent follow-up visits or unsatisfactory treatment results.

[0058] Compared to related technologies, traditional methods rely solely on patients' self-reported reasons for seeking medical treatment and a single-dimensional historical record for recommendations, making it impossible to identify implicit needs. This solution constructs a cross-validation mechanism between subjective statements and objective behavioral data, correlating service quality evaluations from hospital feedback with patients' actual medical journeys to uncover unspoken referral motivations. Furthermore, based on a dual analysis of temporal characteristics and hospital attributes, it accurately captures the periodic patterns in patients' medical behavior and their institutional selection preferences.

[0059] Through the above technical solutions, this application achieves a comprehensive analysis of patients' individualized medical treatment characteristics, solving the recommendation bias problem caused by the single data dimension in traditional systems. The medical treatment preference analysis module can accurately identify patients' long-term preference for specific hospital levels and department types, while the reference to the out-of-town medical treatment reason analysis module can effectively uncover hidden referral needs caused by medical fatigue or disappointment in treatment outcomes, providing multi-dimensional data support for subsequent accurate matching of medical resources.

[0060] This application further proposes technical solutions for determining the medical preferences of target subjects based on the information of the first hospital where they have received medical treatment and the time of their medical treatment, and for determining the reasons for seeking medical treatment in other places based on the information of the second hospital where they have received medical treatment and the medical feedback data.

[0061] The data is categorized into three parts: First, the primary hospital information, which describes the attributes of historical hospitals visited. This can be achieved using data on hospital level, department setup, and service type, and is used to analyze patients' basic preference for hospital type and department. The basic medical visit time pattern refers to the preliminary time distribution pattern derived from historical medical visit statistics. This can be achieved using time series clustering algorithms to periodically analyze historical medical visit times, and is used to reflect patients' initial time preferences for medical visits. The reference medical visit time pattern refers to the time distribution pattern adjusted based on hospital service characteristics. This can be achieved by cross-validating hospital operating hours data with historical medical visit times, and is used to adjust for time periods in the basic time pattern that do not match the hospital's actual service capacity. Second, the secondary hospital information describes the specific medical process of patients at historical hospitals. This can be achieved using data on the number of visits, travel distance, and payment records, and is used to quantify the degree of patient fatigue and differences in treatment effectiveness at specific hospitals. Medical feedback data refers to the hospital's evaluation of the patient's medical process. This can be achieved using satisfaction scores or doctor-patient communication records recorded by the hospital system, and is used to supplement subjective medical experiences that cannot be reflected by objective data.

[0062] Specifically, in determining medical treatment preferences, the system first extracts information on the level and department configuration of historical hospitals, identifying the types of hospitals and departments frequently chosen by patients to form hospital and department preferences. Next, it performs periodic analysis of historical medical treatment times to identify patients' habitual appointment times as a baseline time pattern. Then, by combining the operating hours and departmental duty schedules of each hospital, it corrects time periods in the baseline time pattern that fail to reflect true preferences due to hospital service limitations, generating a reference time pattern. Finally, it integrates hospital preferences, department preferences, and the corrected time preferences to form a three-dimensional medical treatment preference model that includes institution selection, departmental preference, and time patterns. When determining the reasons for seeking medical treatment in other locations, the system calculates the physical exertion index generated by medical treatment behavior by statistically analyzing the number of visits and travel distances at historical hospitals. This is combined with the ratio of payment amount to treatment effect data to analyze medical cost-effectiveness, forming objective indicators for evaluating medical fatigue and effectiveness. Simultaneously, by integrating the doctor-patient communication records and satisfaction scores recorded by the hospital system, we can identify patients' subjective evaluations of the quality of medical services, cross-validate objective indicators with subjective evaluations, and distinguish the root causes of patients choosing to seek medical treatment in other places due to objective limitations or subjective dissatisfaction.

[0063] Compared to related technologies, existing methods typically only statistically analyze visit frequency or simply classify hospital types, failing to correlate hospital service characteristics with time distribution patterns, leading to distorted time preference analysis. Traditional techniques rely on single-dimensional visit frequency or cost data to infer the reasons for seeking medical treatment in other locations, neglecting the interaction between hospital service capacity and patient subjective experience. This solution, through multi-dimensional cross-analysis of hospital attribute data and time data, accurately distinguishes between patients' true time preferences and passive selection behaviors limited by hospital services. Furthermore, by combining a dual verification mechanism of objective behavioral data and subjective evaluation data, it effectively identifies the subjective and objective driving factors in the decision-making process for seeking medical treatment in other locations.

[0064] Through the above technical solution, this application solves the problem that traditional methods neglect the impact of hospital service characteristics on time distribution in medical preference analysis, improves the accuracy of time preference extraction, and avoids misjudgment of preferences due to hospital operating hour restrictions. At the same time, it overcomes the limitations of analyzing the reasons for seeking medical treatment in other locations from a single data source. Through collaborative verification of subjective and objective data, it accurately distinguishes the needs of patients seeking medical treatment in other locations due to objective limitations or subjective dissatisfaction, providing a reliable data foundation for subsequent accurate matching of target hospitals.

[0065] This application further proposes to determine the target object's hospital preference and department preference based on the information of the first hospital, to determine the medical time preference based on the basic medical time pattern and the reference medical time pattern, and to combine the hospital preference, department preference and medical time preference to form the medical preference; the information of the second hospital includes the number of medical payment, payment amount, travel distance and medical effect information, to generate medical fatigue description information based on the number of payment and travel distance, to generate medical effect description information based on the payment amount and medical effect information, and to determine the reference reasons for medical treatment in other places by combining medical feedback data.

[0066] Among these, hospital preference refers to patients' preferred choice regarding the level, nature, or service type of medical institutions. This can be achieved using algorithms that analyze the correlation between historical hospital level classification data and the departments visited, reflecting patients' trust in medical institutions. Department preference refers to patients' preference for specialist treatment for specific disease types. This can be achieved using historical department distribution statistics and disease type matching models, identifying patients' specialist treatment habits. Medical visit time preference refers to the regularity of patients' medical visits within specific time periods. This can be achieved using time series pattern mining algorithms combined with historical registration time clustering analysis, capturing patients' time-sensitive needs. Medical fatigue description information refers to a quantitative indicator of the physical and mental burden caused by multiple hospital visits. This can be achieved using a weighted calculation model of travel distance and number of visits, assessing patients' potential demand for nearby medical resources. Medical treatment outcome description information refers to a comprehensive evaluation indicator of medical expenses and treatment effectiveness. This can be achieved using a cost-effectiveness ratio algorithm combined with treatment effectiveness grading standards, quantifying patients' satisfaction with medical quality.

[0067] Specifically, in the stage of generating medical preferences, structured data such as hospital level and department setup are extracted by analyzing the primary hospital information of historical hospitals visited. A hierarchical clustering algorithm is then used to identify patients' preferences for tertiary hospitals or specialized hospitals, forming hospital preferences. Simultaneously, the correlation between historical departments visited and disease diagnoses is analyzed to establish a department selection probability model, generating department preferences. In the time preference analysis, the basic medical treatment time pattern is obtained by statistically analyzing the distribution of historical registration times. The reference medical treatment time pattern is modified by incorporating dynamic data on hospital capacity. Finally, the optimal medical treatment time is determined using a time window matching algorithm. In the stage of analyzing the reasons for seeking medical treatment in other locations, the number of medical treatment payments and travel distance are normalized and input into a fatigue calculation model to generate quantitative indicators reflecting traffic burden. Medical treatment payment amount and treatment effect are processed by a cost-benefit analysis model to generate medical quality evaluation indicators. These two types of objective indicators are integrated with subjective evaluations from medical treatment feedback data in a multi-dimensional manner. A decision tree classification algorithm is used to identify the core factors leading to seeking medical treatment in other locations.

[0068] Compared to related technologies, traditional methods rely solely on patients' self-reported reasons for seeking medical treatment, lacking in-depth analysis of historical behavioral data and prone to information bias. Existing systems typically employ single-dimensional time statistics or simple department matching, failing to dynamically adjust time preference models. This solution integrates multi-dimensional data such as hospital level, departmental association, and dynamic time adjustments to construct a composite preference model, effectively addressing the problem of traditional recommendation systems neglecting implicit medical habits. Regarding the analysis of reasons for seeking medical treatment in other locations, related technologies largely rely on manually entered surface-level reasons, while this solution, through quantitative analysis of medical fatigue and cost-effectiveness ratios, combined with cross-validation of subjective and objective data, can accurately identify the underlying motivations for seeking medical treatment that patients have not explicitly expressed.

[0069] Through the aforementioned technical solution, this application can automatically identify patients' complex preferences regarding hospital level, specialty type, and appointment time, thus solving the matching bias problem caused by traditional recommendation systems neglecting historical behavioral patterns. By quantitatively analyzing the ratio of transportation burden to medical quality cost during the medical treatment process, combined with subjective and objective evaluation data, it accurately captures potential factors leading to cross-regional medical treatment, improving the alignment between medical resource recommendations and patients' actual needs. This technical solution effectively overcomes the limitations of single data source analysis, realizing the synergistic effect of multi-dimensional heterogeneous data, and providing accurate data support for personalized medical recommendations.

[0070] This application further proposes a method to determine target nodes that match the physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the medical destination, based on the target object's physiological state and the reason for seeking medical treatment in another location. See [link to relevant documentation]. Figure 4 The method includes the following steps:

[0071] 401. Based on physiological state, the server determines multiple reference nodes from multiple candidate nodes, and the multiple reference nodes are candidate nodes that match the physiological state.

[0072] 402. The server determines the target node from multiple reference nodes based on the reason of seeking medical treatment in a different location.

[0073] Physiological status refers to the current health indicators of the target object assessed through physiological data. Specifically, this can be achieved by combining sensor data such as heart rate, blood pressure, and blood oxygen saturation with diagnostic records from electronic medical records. Machine learning models are used for classification and processing to screen candidate nodes with corresponding medical data processing capabilities. Reasons for seeking medical treatment in other locations refer to the subjective and objective factors triggering the target object's choice to seek medical treatment in another location. Natural language processing techniques can be used to extract keywords and classify intents from the text information submitted by patients, such as insufficient medical resources, poor treatment outcomes, and the need for convenient transportation, to adapt recommendation algorithm modes for different nodes. Candidate nodes refer to heterogeneous computing resources distributed at the medical destination. A node registration mechanism can be used to manage server clusters with different computing power configurations, data processing models, and recommendation strategies. For example, GPU-accelerated nodes are suitable for image data analysis, while CPU-intensive nodes are suitable for medical record text processing. Reference nodes refer to a subset of candidate nodes initially screened based on physiological status. A matching degree threshold filtering mechanism can be used, for example, setting a matching degree threshold of 80% or higher for nodes to proceed to the next stage of screening, reducing subsequent computational overhead. The target node refers to the final selected computing node. Specifically, a weighted scoring mechanism can be adopted, which combines the matching degree between the reasons for seeking medical treatment in other places and the node recommendation mode. For example, nodes that support multi-factor decision trees can be selected to handle complex reasons for seeking medical treatment.

[0074] Specifically, the process begins by filtering candidate nodes with corresponding medical data processing capabilities based on the patient's physiological state. For example, when a patient is in the acute phase of cardiovascular disease, nodes supporting real-time vital sign monitoring are prioritized. A second matching process is then performed based on the reason for seeking medical treatment in a different location. For instance, if the reason for the patient's out-of-town medical treatment is "lack of specialized equipment in the local area," nodes supporting medical resource gap analysis are selected. This two-stage filtering mechanism avoids the waste of computational resources caused by traversing all nodes by first eliminating nodes unrelated to the physiological state and then selecting nodes with recommended patterns based on the specific reason. It also adapts to the differences in computing power distribution and data processing models among different nodes.

[0075] Compared to related technologies, traditional methods typically employ single-dimensional matching, such as filtering solely based on hospital level or directly calling fixed nodes to process all requests. This solution, through a phased screening mechanism, is the first to collaboratively analyze physiological status and reasons for seeking medical treatment in different locations, dynamically adapting to heterogeneous node resources. Related technologies cannot distinguish differences in node computing power, leading to high latency or recommendation bias. This solution, however, optimizes resource utilization through two-stage screening while simultaneously improving recommendation accuracy.

[0076] Through the above technical solution, this application solves the problem that distributed heterogeneous node resources cannot adapt to the individualized needs of patients, and realizes the matching of computing power based on physiological status and the adaptation of recommendation mode based on the reason for seeking medical treatment. For example, in the scenario where a patient chooses to seek medical treatment in another location for postoperative follow-up, the system can prioritize calling nodes that support image data comparison, and then select a recommendation mode that supports traffic condition analysis based on the reason of "convenience of follow-up examination", and finally generate hospital recommendation results that take into account both medical professionalism and geographical accessibility.

[0077] This application further proposes a method for determining multiple reference nodes from multiple candidate nodes based on physiological states, wherein the multiple reference nodes are candidate nodes that match the physiological states. This includes: determining first node description information for each candidate node, where the first node description information describes the associated physiological state of the corresponding candidate node; determining a first matching degree between the physiological state and each candidate node based on the physiological state and the first node description information of each candidate node; identifying candidate nodes among the multiple candidate nodes whose first matching degree with the physiological state is greater than or equal to a first matching degree threshold as reference nodes, thus obtaining multiple reference nodes; and determining a target node from the multiple reference nodes based on the reason for seeking medical treatment in a different location, including: determining second node description information for each reference node, where the second node description information describes the associated reason for seeking medical treatment of the reference node; determining a second matching degree between the reason for seeking medical treatment in a different location and each reference node based on the reason for seeking medical treatment in a different location and the associated reason for seeking medical treatment of each reference node; and identifying the reference node among the multiple reference nodes with the highest second matching degree with the reason for seeking medical treatment in a different location as the target node.

[0078] The first node description information describes the physiological range of the candidate node's processing capabilities. This can be generated using natural language processing (NLP) techniques to semantically analyze node service logs, used to filter computing nodes capable of handling the current patient's condition. The second node description information describes the type of medical treatment reason the candidate node is suited for. This can be achieved by constructing a knowledge graph to establish relationships between nodes and medical treatment scenarios, used to match the core needs of patients seeking medical care in different locations. The first matching degree threshold is the minimum suitability standard for node selection. This can be automatically set using a dynamic adjustment algorithm based on the number of candidate nodes and computing power distribution, used to balance computational efficiency and matching accuracy. The second highest matching degree refers to the ranking of the association strength between the node and the medical treatment reason. This can be achieved using a cosine similarity algorithm to calculate the semantic matching degree between the reason text and the node description information, used to select the node resource that best meets the patient's needs.

[0079] Specifically, when a patient submits a request for out-of-town medical treatment, the system first generates a standardized description of the patient's physiological state by parsing the electronic medical record, such as converting the diagnostic conclusions in the medical record into ICD-10 disease codes. The first node description information of candidate nodes is extracted using a pre-trained medical entity recognition model, for example, by extracting high-frequency diagnostic keywords from the node's historical case processing. The first matching degree is calculated using a vector space model, mapping the patient's physiological state and node description information to feature vectors and then calculating the cosine similarity. When the matching degree exceeds a dynamic threshold, the node enters the reference node set. Subsequently, the patient's reason for seeking out-of-town medical treatment is processed through word segmentation and semantically matched with the second node description information of the reference nodes, for example, using a BERT model to calculate a text similarity score. Finally, the node with the highest score is selected as the target node.

[0080] Compared to related technologies, traditional methods rely solely on single-dimensional matching based on hospital geography or departmental classification, failing to address the differentiated service capabilities of heterogeneous nodes. This solution, however, establishes a dual-description system for nodes, dynamically adapting to the computational needs of different medical scenarios while ensuring basic node service capabilities. The resource waste caused by static threshold settings in related technologies is optimized in this solution through a dynamic threshold adjustment strategy. Simultaneously, a maximum matching degree optimization mechanism avoids resource conflicts arising from parallel computing across multiple nodes.

[0081] Through the above technical solution, this application achieves precise and dynamic allocation of medical computing node resources. In scenarios where patients experience sudden acute illness requiring urgent transfer, nodes with experience in handling the disease and supporting green channel services can be quickly selected. In scenarios where patients choose to seek medical treatment in other locations due to insufficient medical resources, nodes specializing in the disease and supporting cross-regional collaboration can be accurately matched. This dual screening mechanism for node resources effectively reduces the probability of distributing invalid computing tasks while ensuring the priority service needs of critically ill patients.

[0082] This application further proposes a method for determining the target hospital to recommend to the target individual from multiple hospitals at the medical destination based on the target node, see [link to relevant documentation]. Figure 5 Taking the server as the executing entity as an example, the method includes the following steps:

[0083] 501. The server identifies multiple candidate hospitals from multiple hospitals based on electronic medical records, medical reference information, and hospital description information of each hospital through the target node.

[0084] 502. The server sends the target object's electronic medical record, medical reference information, and reasons for seeking medical treatment in other locations to the hospital nodes of each candidate hospital, so that the hospital nodes of each candidate hospital can determine the degree of matching for treatment, the willingness to receive treatment, and the reasons for generating the degree of matching for treatment and the willingness to receive treatment based on the target object's electronic medical record, medical reference information, and reasons for seeking medical treatment in other locations.

[0085] 503. The server obtains the matching degree of each candidate hospital, the willingness to accept patients, and the reason for the generation through the target node;

[0086] 504. The server determines the target hospital from multiple candidate hospitals based on the matching degree of each candidate hospital, the willingness to receive patients, the reason for generation, and the medical reference information of each candidate hospital.

[0087] Among them, candidate hospitals refer to medical institutions that have a potential match with the target patient's medical needs through preliminary screening. This can be achieved using a multi-dimensional matching algorithm that combines hospital descriptions with patient characteristics, narrowing the recommendation scope and improving subsequent processing efficiency. The degree of patient acceptance matches a quantifiable indicator of the hospital's suitability for the patient's condition. This can be calculated by analyzing the correlation between disease characteristics in electronic medical records and the hospital's specialty strengths, objectively assessing whether the hospital's professional capabilities meet the treatment needs. The willingness to accept patients refers to the strength of the hospital's intention to receive patients under current resource conditions. This can be generated by combining real-time resource monitoring data from hospital nodes with historical patient acceptance strategy models, reflecting the hospital's dynamic patient acceptance capacity. The reason for generation refers to the textual basis for the decision on the degree of patient acceptance matches and the willingness to accept patients. This can be achieved using natural language generation technology to transform the matching logic and resource assessment results into an interpretable explanation, providing a transparent basis for recommendation decisions.

[0088] Specifically, the target node first filters out a set of candidate hospitals that meet basic admission criteria based on static attributes such as disease type in the patient's electronic medical record, preference data in the medical reference information, and specialist settings and equipment configuration in the hospital description information. For example, when a patient has cardiovascular disease, the target node prioritizes hospitals with cardiology departments and interventional treatment equipment. Then, the patient's complete medical data package is sent to the hospital nodes of the candidate hospitals, triggering the hospitals to generate quantitative admission matching degree and admission willingness values ​​based on dynamic data such as real-time bed occupancy rate and physician scheduling status, combined with historical data such as the hospital's specialist treatment success rate and experience in handling similar cases. Simultaneously, a text explaining the decision reasons, including resource load analysis and comparison of professional advantages, is generated. After receiving feedback data from each hospital, the target node comprehensively evaluates the professional suitability reflected by the admission matching degree, the resource availability represented by the admission willingness, and the decision rationality revealed by the generated reasons. Simultaneously, it considers personalized needs such as time preferences and distance sensitivity in the patient's medical reference information, and determines the final recommended target hospital through a weighted scoring model.

[0089] Compared to related technologies, existing cross-regional medical treatment recommendation systems typically rely solely on publicly available hospital information and patients' basic needs for one-way matching, lacking real-time feedback on hospital resource status and dynamic assessment of professional suitability. This solution, however, establishes a dynamic response mechanism for hospital nodes. After initial screening, it triggers hospitals to generate quantitative feedback based on real-time operational data and professional capability models. This ensures that recommendation decisions not only incorporate static matching results but also integrate dynamic assessment data on the hospital's current capacity and professional advantages. For example, if a hospital has treatment capabilities but is currently operating at full capacity, its willingness to treat patients will automatically decrease, avoiding recommendations to overloaded medical institutions.

[0090] Through the above technical solution, this application solves the problem of the disconnect between recommended results and actual patient reception capacity caused by missing data at the hospital level. A two-way data interaction mechanism ensures that recommended results not only meet the personalized needs of patients but also accurately reflect the real-time reception status of the hospital. Simultaneously, the transparent output of the generation reasons allows patients to understand the recommendation logic, increasing their trust in the recommended results. Furthermore, through a phased processing mechanism, the complexity of data processing is effectively reduced while ensuring recommendation accuracy, avoiding the waste of system resources caused by directly processing all hospital data.

[0091] This application further proposes a technical solution that determines the third matching degree between the target object and the hospital based on the target node's electronic medical records, medical reference information, and hospital description information of each hospital; identifies hospitals with a third matching degree greater than or equal to the second matching degree threshold as candidate hospitals; determines the patient admission intention score based on the patient admission matching degree, patient admission intention, and generation reason of the candidate hospitals; determines the fourth matching degree based on patient admission intention and medical reference information; and finally integrates the patient admission intention score and the fourth matching degree to determine the target hospital.

[0092] The third matching degree refers to the static matching degree between the target object and the hospital, calculated based on electronic medical records, medical reference information, and hospital description information. This can be achieved through feature extraction and similarity calculation, used to filter out a basic set of candidate hospitals, overcoming the limitations of traditional recommendations that rely on a single indicator. The patient admission intention score is a quantitative indicator of the candidate hospital's willingness to admit the target object. This can be calculated by fusing the patient admission matching degree, admission intention, and the reasons for its generation, dynamically reflecting the hospital's admission capacity and subjective willingness. The fourth matching degree refers to the dynamic matching degree between the target object and the candidate hospital, calculated based on admission intention and medical reference information. This can be achieved by using a model to perform secondary calibration of the needs of both parties, balancing patient preferences and hospital service capabilities. The target matching score is the fusion result of the admission intention score and the fourth matching degree, implemented through weighted summation or feature concatenation, used to establish a two-way selection mechanism.

[0093] Specifically, the target node first extracts features from electronic medical records and medical reference information to generate object description information. It then combines this with hospital description information to calculate a third matching degree, filtering out candidate hospitals that meet the threshold. Subsequently, candidate hospital nodes provide feedback on their patient matching degree, willingness to receive patients, and the reasons for this information. The target node processes this feedback through a model to generate a patient intention score. Simultaneously, a fourth matching degree is calculated based on the medical reference information and patient intention, reflecting the dynamic adaptation between patient preferences and hospital service capabilities. Finally, the patient intention score and the fourth matching degree are merged into a target matching score, and the candidate hospital with the highest score is selected as the target hospital.

[0094] Compared to related technologies, which typically rely solely on static hospital information for one-way recommendations, lacking dynamic assessment of hospital capacity and failing to integrate bidirectional matching of patient preferences and hospital intentions, this solution addresses the matching bias problem caused by data heterogeneity in traditional recommendation systems by introducing a fusion mechanism of patient intention scores and a fourth matching degree. This mechanism simultaneously considers both dynamic hospital capacity and personalized patient needs when selecting candidate hospitals.

[0095] Through the above technical solutions, this application achieves the following technical effects: First, by screening candidate hospitals through the third matching degree, the fit between the static attributes of hospitals and patient characteristics is quantified, thereby improving the screening accuracy of the basic candidate set; Second, by combining the degree of matching for receiving patients, the willingness to receive patients, and the reason for generation, the intention to receive patients is calculated, dynamically reflecting the hospital's capacity to receive patients and subjective willingness, thus avoiding recommendation failure due to temporary changes in hospital resources; Third, by calibrating patient preferences and hospital service capabilities based on the fourth matching degree, the recommendation results are ensured to meet the needs of both parties simultaneously, reducing the secondary referral rate caused by information asymmetry.

[0096] This application further proposes to generate object description information of the target object based on electronic medical records and medical reference information by the target node, determine the third matching degree between the target object and each hospital based on the object description information and the hospital description information of each hospital, input the matching degree of each candidate hospital, the willingness to receive treatment and the generation reason into the model to determine the willingness to receive treatment score, and input the willingness to receive treatment and medical reference information into the matching degree determination model to output the fourth matching degree, and fuse the willingness to receive treatment and the fourth matching degree to determine the target matching score, and select the candidate hospital with the highest score as the target hospital.

[0097] The patient description information refers to a structured set of patient features generated through feature extraction and fusion of electronic medical records and medical reference information. Specifically, this can be achieved using deep neural networks for multi-dimensional feature extraction and concatenation, comprehensively representing the patient's medical history, real-time needs, and preferences. The third matching degree refers to a quantifiable indicator of the similarity between patient features and hospital service capabilities. This can be achieved by calculating the feature vector similarity between the patient description information and hospital description information using a cosine similarity algorithm, used to screen candidate hospitals with matching basic attributes. The patient acceptance intention score determination model is a computational model used to comprehensively evaluate the hospital's acceptance capacity and initiative. Specifically, it can be achieved using a multilayer perceptron to perform non-linear weighted calculations of matching degree, patient acceptance intention, and generation reasons, used to quantify the hospital's priority in accepting patients. The matching degree determination model is an analytical model used to assess the fit between patient preferences and hospital acceptance intention. Specifically, it can be achieved by using an attention mechanism to calculate the correlation between medical reference information and patient acceptance intention, used to enhance the objectivity of two-way matching.

[0098] Specifically, electronic medical records and medical reference information are input into the feature extraction module. A convolutional neural network extracts temporal features of diagnostic records and medication history, while a word embedding model analyzes textual features related to medical preferences. The feature vectors of the fused object description information and hospital description information are input into a similarity calculation layer to generate a third matching score. The patient reception data of candidate hospitals are input into a patient intention score determination model. This model uses a fully connected layer to weight the matching degree, patient intention, and reasons, outputting a 0-1 range patient intention score. Simultaneously, the matching degree determination model assigns attention weights to medical preferences and patient intention, generating a fourth matching score. Finally, the two scores are linearly weighted and fused, and the candidate hospital with the highest total score is selected as the recommendation target.

[0099] In some specific implementations, the feature extraction module can use a pre-trained BERT model to encode the electronic medical record text, the hospital description information can be used to construct a hospital service capacity vector through a knowledge graph, the model for determining the patient intention score can integrate a random forest algorithm to process discrete generation cause data, and the model for determining the matching degree can use a cross-attention mechanism to capture the potential correlation between preferences and intentions.

[0100] Compared to related technologies, traditional methods rely solely on static hospital information for one-way recommendations, neglecting the two-way matching between dynamic patient needs and hospital capacity. This solution constructs object description information to achieve a structured representation of patient characteristics and employs a dual-matching model to quantify the matching degree between hospital service capacity and patient willingness, overcoming the limitations of single-dimensional recommendations. By dynamically fusing patient willingness scores and matching degrees, a two-way selection decision-making mechanism is established, effectively addressing the dynamic adaptation problem between patient needs and medical resources.

[0101] Through the above technical solutions, this application achieves deep matching between multidimensional patient data and hospital resources, improving recommendation accuracy. The hospital's patient acceptance capacity is quantified using a patient acceptance intention score model, ensuring that the recommendation results align with the hospital's actual patient acceptance conditions. The matching degree determination model effectively integrates patient preferences and hospital intentions, reducing the matching failure rate caused by information asymmetry.

[0102] This application further proposes a method for generating object description information of a target object based on electronic medical records and medical reference information. The method includes extracting features from electronic medical records and medical reference information to obtain electronic medical record features and reference information features, fusing the two types of features and generating object description information through multiple rounds of iterative decoding. The method for determining a third matching degree based on object description information and hospital description information includes extracting features from both to obtain first description information features and second description information features, and determining the matching degree by calculating feature similarity.

[0103] Feature extraction refers to extracting representative feature vectors from raw data, which can be implemented using convolutional neural networks or recurrent neural networks to capture disease features in electronic medical records and preference features in medical reference information. Feature fusion refers to integrating feature vectors from different sources into a unified representation, which can be implemented using fully connected layers or attention mechanisms to construct a complete feature profile of the patient. Multi-round iterative decoding refers to generating semantically coherent text descriptions through multiple feature reconstructions, which can be implemented using a Transformer decoder to solve the semantic fragmentation problem caused by splicing discrete features. Feature similarity refers to measuring the degree of matching between two features through vector space distance, which can be calculated using cosine similarity or Euclidean distance to quantify the matching degree between patient needs and hospital resources.

[0104] Specifically, diagnostic records and medication histories in electronic medical records are transformed into structured feature vectors through a feature extraction network, while preference data in medical reference information is transformed into preference feature vectors through a natural language processing model. These two types of features are integrated by a fusion layer and input into a decoder, where a multi-round self-attention mechanism iteratively generates object description text containing complete semantic information. Hospital description information is transformed into hospital feature vectors through a feature extraction network with the same structure, and similarity is calculated between these vectors and patient feature vectors. Hospitals with high matching scores are added to the candidate list. This method, through deep feature fusion and semantic reconstruction, maps individualized patient data and hospital resource descriptions to the same feature space, solving the problem of insufficient matching accuracy caused by differences in feature dimensions in traditional methods.

[0105] Compared to related technologies, existing medical recommendation systems typically employ keyword matching or simple feature concatenation, failing to effectively handle the semantic differences between heterogeneous data. Traditional methods treat patient medical records and hospital information as independent features for linear matching, ignoring the deep connections between the data. This solution constructs a unified feature extraction framework to achieve deep integration of patient subjective preferences and objective medical data. It utilizes multi-round iterative decoding to generate context-sensitive descriptive information, making the matching process between patient needs and hospital resources semantically interpretable.

[0106] All of the above-mentioned optional technical solutions can be combined in any way to form the optional embodiments of this application, and will not be described in detail here.

[0107] Through the above technical solution, this application achieves a precise mapping between individualized patient needs and medical resources, effectively improving the accuracy of recommendations for cross-regional medical treatment. This method solves the data dimension mismatch problem in traditional recommendation systems through deep feature fusion, and improves the rationality of medical resource recommendations through a semantically coherent descriptive information generation mechanism, providing patients with hospital selection options that better meet their actual needs.

[0108] Figure 6 This is a schematic diagram of a medical service improvement system based on cross-regional medical treatment data analysis provided in an embodiment of this application. See also... Figure 6 The system includes:

[0109] The first determining module 601 is used to determine the target object's current physiological state, medical reference information, and reason for seeking medical treatment in another location in response to the target object's request for medical treatment in another location.

[0110] The second determining module 602 is used to determine a target node that matches the current physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the medical destination, based on the current physiological state of the target object and the reason for seeking medical treatment in another location. Different candidate nodes have different computing power and hospital recommendation modes, and the medical destination is obtained from the request for seeking medical treatment in another location.

[0111] The third determining module 603 is used to input the target object's electronic medical record, medical reference information, and reason for seeking medical treatment in another location into the target node, and to determine the target hospital to recommend to the target object from multiple candidate hospitals at the destination of medical treatment through the target node.

[0112] It should be noted that the above-described embodiments of the medical service improvement system based on cross-regional medical treatment data analysis are only illustrative examples of the functional module divisions used to improve medical services. In practical applications, the functions described above can be assigned to different functional modules as needed, that is, the internal structure of the computer equipment can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the above-described embodiments of the medical service improvement system based on cross-regional medical treatment data analysis and the embodiments of the medical service improvement method based on cross-regional medical treatment data analysis belong to the same concept, and their specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0113] Figure 7 This is a schematic diagram of a server structure provided in an embodiment of this application. The server 700 can vary significantly due to different configurations or performance. It may include one or more Central Processing Units (CPUs) 701 and one or more memories 702. The one or more memories 702 store at least one computer program, which is loaded and executed by the one or more processors 701 to implement the methods provided in the various method embodiments described above. Of course, the server 700 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The server 700 may also include other components for implementing device functions, which will not be elaborated upon here.

[0114] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including a computer program that can be executed by a processor to perform the medical service improvement method based on remote medical treatment data analysis in the above embodiments. For example, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, and optical data storage device, etc.

[0115] In an exemplary embodiment, a computer program product or computer program is also provided, which includes program code stored in a computer-readable storage medium. The processor of a computer device reads the program code from the computer-readable storage medium and executes the program code, causing the computer device to perform the above-described method for improving medical services based on cross-regional medical treatment data analysis.

[0116] In some embodiments, the computer program involved in the present application embodiments may be deployed and executed on a computer device, or executed on multiple computer devices located in one location, or executed on multiple computer devices distributed in multiple locations and interconnected through a communication network. Multiple computer devices distributed in multiple locations and interconnected through a communication network may constitute a blockchain system.

[0117] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0118] The above are merely optional embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for improving medical services based on cross-regional medical treatment data analysis, characterized in that, The method includes: In response to a target individual's request for out-of-town medical treatment, determine the target individual's current physiological state, medical reference information, and the reason for seeking out-of-town medical treatment. Based on the current physiological state of the target object and the reason for seeking medical treatment in another location, a target node matching the current physiological state and the reason for seeking medical treatment in another location is determined from multiple candidate nodes associated with the medical destination. Different candidate nodes have different computing power and hospital recommendation modes. The medical destination is obtained from the request for seeking medical treatment in another location. The hospital recommendation mode is used to quantify the matching degree between the hospital and the patient's needs. The step of determining a target node that matches the current physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the medical destination, based on the target object's physiological state and the reason for seeking medical treatment in another location, includes: Based on the current physiological state, multiple reference nodes are determined from the multiple candidate nodes, and the multiple reference nodes are candidate nodes that match the current physiological state; based on the reason for seeking medical treatment in another location, the target node is determined from the multiple reference nodes. Based on the current physiological state, determining multiple reference nodes from the multiple candidate nodes, wherein the multiple reference nodes are candidate nodes that match the current physiological state, includes: First node description information is determined for each candidate node, which describes the associated physiological state of the corresponding candidate node; based on the current physiological state and the first node description information of each candidate node, a first matching degree between the current physiological state and each candidate node is determined; candidate nodes among the plurality of candidate nodes whose first matching degree with the current physiological state is greater than or equal to the first matching degree threshold are determined as reference nodes, thus obtaining the plurality of reference nodes; The process of determining the target node from the plurality of reference nodes based on the reason for seeking medical treatment in a different location includes: Determine second node description information for each of the reference nodes, wherein the second node description information is used to describe the associated medical treatment reason of the reference node; based on the out-of-town medical treatment reason and the associated medical treatment reason of each of the described reference nodes, determine the second matching degree between the out-of-town medical treatment reason and each of the reference nodes; determine the reference node with the highest second matching degree between the multiple reference nodes and the out-of-town medical treatment reason as the target node; The target object's electronic medical record, medical reference information, and reason for seeking medical treatment in another location are input into the target node. The target node then determines the target hospital to recommend to the target object from multiple candidate hospitals at the destination of the medical treatment.

2. The method according to claim 1, characterized in that, The medical reference information includes medical preferences and medical needs; the reasons for seeking medical treatment in another location include subjective reasons and suggested reasons; and the process of responding to a target individual's request for medical treatment in another location, including determining the target individual's physiological state, medical reference information, and reasons for seeking medical treatment in another location, includes: In response to a target individual's request for out-of-town medical treatment, the system acquires the target individual's physiological data, the subjective reason for seeking out-of-town medical treatment, historical medical data, and medical feedback data. The medical feedback data is used to represent the hospital's evaluation of the target individual. Based on the physiological data, the physiological state of the target object is determined; Based on the historical medical data, the medical preference is determined; Based on the historical medical data and the medical feedback data, the reference reasons for seeking medical treatment in other locations are determined.

3. The method according to claim 2, characterized in that, The historical medical data includes the hospitals and times of historical medical visits. Determining the medical preference based on the historical medical data includes: Obtain the first hospital information of the historical medical treatment hospital, which is used to introduce the historical medical treatment hospital; determine the basic medical treatment time pattern of the target object based on the historical medical treatment time; determine the reference medical treatment time pattern of the target object based on the first hospital information and the historical medical treatment time; determine the medical treatment preference based on the first hospital information, the basic medical treatment time pattern and the reference medical treatment time pattern. The determination of the reference reasons for out-of-town medical treatment based on the historical medical data and the medical feedback data includes: Obtain the second hospital information of the historical hospital where the target object received treatment. The second hospital information is used to describe the target object's medical treatment at the historical hospital. Based on the second hospital information and the medical treatment feedback data, determine the reference reason for seeking medical treatment in another location.

4. The method according to claim 3, characterized in that, The step of determining the medical preference based on the first hospital information, the basic medical treatment time pattern, and the reference medical treatment time pattern includes: Based on the information from the first hospital, the hospital preference and department preference of the target object are determined; based on the basic medical treatment time pattern and the reference medical treatment time pattern, the medical treatment time preference of the target object is determined; the hospital preference, department preference and medical treatment time preference of the target object are concatenated to obtain the medical treatment preference. The second hospital information includes the number of medical visits and payments, the amount paid, the distance traveled for medical treatment, and the treatment effect information. The step of determining the reasons for seeking medical treatment in another location based on the second hospital information and the medical feedback data includes: Based on the number of medical visits and the distance traveled for medical treatment, medical fatigue description information is determined; based on the amount of medical expenses paid and the medical treatment effect information, medical effect description information is determined; based on the medical fatigue description information, the medical effect description information, and the medical feedback data, the reference reasons for seeking medical treatment in other locations are determined.

5. The method according to claim 1, characterized in that, The step of determining the target hospital to recommend to the target individual from multiple hospitals at the medical destination via the target node includes: Based on the electronic medical records, medical reference information, and hospital description information of each hospital, multiple candidate hospitals are determined from the multiple hospitals through the target node. The electronic medical record of the target object, the medical treatment reference information, and the reason for seeking medical treatment in another location are sent to the hospital nodes of each of the candidate hospitals, so that the hospital nodes of each of the candidate hospitals can determine the degree of matching for receiving treatment, the willingness to receive treatment, and the reasons for generating the degree of matching for receiving treatment and the willingness to receive treatment based on the electronic medical record of the target object, the medical treatment reference information, and the reason for seeking medical treatment in another location. Through the target node, obtain the patient matching degree, patient acceptance intention, and the reasons for the generation of the patient matching degree and patient acceptance intention for each candidate hospital; The target hospital is determined from the multiple candidate hospitals through the target node based on the matching degree of each candidate hospital, the willingness to receive patients, the reason for the generation, and the medical reference information.

6. The method according to claim 5, characterized in that, The step of determining multiple candidate hospitals from the multiple hospitals through the target node, based on the electronic medical records, the medical reference information, and the hospital description information of each hospital, includes: Using the target node, based on the electronic medical record, the medical reference information, and the hospital description information of each hospital, a third matching degree between the target object and each hospital is determined; hospitals among the multiple hospitals whose third matching degree is greater than or equal to the second matching degree threshold are identified as candidate hospitals, thus obtaining the multiple candidate hospitals; The step of determining the target hospital from the multiple candidate hospitals through the target node, based on the matching degree of each candidate hospital, the willingness to receive patients, the reason for generation, and the medical reference information, includes: Using the target node, based on the matching degree of each candidate hospital, the willingness to receive patients, and the reason for generation, a patient intention score for each candidate hospital is determined for the target object; based on the willingness to receive patients and the medical reference information of each candidate hospital, a fourth matching degree between the target object and each candidate hospital is determined; based on the patient intention score and the fourth matching degree, the target hospital is determined from the multiple candidate hospitals.

7. The method according to claim 6, characterized in that, The step of determining the third matching degree between the target object and each of the hospitals through the target node, based on the electronic medical record, the medical reference information, and the hospital description information of each hospital, includes: Based on the electronic medical record and the medical reference information, object description information of the target object is generated through the target node; based on the object description information of the target object and the hospital description information of each hospital, a third matching degree between the target object and each hospital is determined. The step of determining the patient admission intention score of each candidate hospital for the target patient through the target node, based on the patient admission matching degree, patient admission intention, and generation reason of each candidate hospital, includes: The matching degree of each candidate hospital, the willingness to receive patients, and the reason for generation are input into the patient intention score determination model. The patient intention score determination model processes the matching degree of patients, the willingness to receive patients, and the reason for generation to obtain the patient intention score of each candidate hospital for the target object. The determination of the fourth matching degree between the target object and each of the candidate hospitals based on the admission intentions and medical reference information includes: The willingness to receive patients and the medical reference information of each candidate hospital are input into the matching degree determination model. The matching degree determination model processes the willingness to receive patients and the medical reference information and outputs the fourth matching degree between the target object and each candidate hospital. The process of determining the target hospital from the plurality of candidate hospitals based on the patient intention score and the fourth matching degree includes: The patient intention score and the fourth matching degree are fused together to obtain the target matching score of each candidate hospital; the candidate hospital with the highest target matching score among the multiple candidate hospitals is determined as the target hospital.

8. A medical service improvement system based on cross-regional medical treatment data analysis, characterized in that, The system includes: The first determining module is used to determine the target object's current physiological state, medical reference information, and reason for seeking medical treatment in another location in response to the target object's request for medical treatment in another location. The second determining module is used to determine a target node that matches the current physiological state and the reason for seeking medical treatment in another location from multiple candidate nodes associated with the destination of medical treatment, based on the current physiological state of the target object and the reason for seeking medical treatment in another location. Different candidate nodes have different computing power and hospital recommendation modes. The destination of medical treatment is obtained from the request for seeking medical treatment in another location. The hospital recommendation mode is used to quantify the matching degree between the hospital and the patient's needs. The second determining module is used to determine multiple reference nodes from the multiple candidate nodes based on the current physiological state, wherein the multiple reference nodes are candidate nodes that match the current physiological state; and to determine the target node from the multiple reference nodes based on the reason for seeking medical treatment in another location. The second determining module is configured to: determine first node description information for each candidate node, wherein the first node description information describes the associated physiological state of the corresponding candidate node; determine a first matching degree between the current physiological state and each candidate node based on the current physiological state and the first node description information of each candidate node; determine reference nodes among the plurality of candidate nodes whose first matching degree with the current physiological state is greater than or equal to a first matching degree threshold, thereby obtaining the plurality of reference nodes; determine second node description information for each reference node, wherein the second node description information describes the associated medical treatment reason of the reference node; determine a second matching degree between the out-of-town medical treatment reason and each of the reference nodes based on the out-of-town medical treatment reason and the associated medical treatment reason of each reference node; and determine the reference node among the plurality of reference nodes with the highest second matching degree with the out-of-town medical treatment reason as the target node. The third determining module is used to input the target object's electronic medical record, the medical treatment reference information, and the reason for seeking medical treatment in another location into the target node, and to determine the target hospital to recommend to the target object from multiple candidate hospitals at the destination of medical treatment through the target node.

Images