Hospital patient service process data processing and optimizing system

Abstract

The invention discloses a hospital patient service flow data processing and optimizing system, which belongs to the technical field of medical information processing, and comprises a data acquisition module, a digital twin modeling module, a path planning module, a bottleneck identification module and an optimization execution module, a full-process model covering registration, waiting for diagnosis, diagnosis and treatment, settlement and re-diagnosis is constructed through a digital twinning technology, a personalized medical treatment path is generated according to historical data and real-time states of patients, efficiency bottlenecks of all links are automatically recognized, an optimization strategy is generated, and practical application shows that the system can reduce the average medical treatment time of the patients by 43%, the service bottleneck recognition accuracy rate reaches 92%, and the system is suitable for large-scale popularization and application. And an effective solution is provided for precise transformation of hospital public service management.

Description

Technical Field

[0001] This invention relates to the field of medical information processing technology, and in particular to a hospital patient service process data processing and optimization system based on digital twin technology, which is used to realize digital modeling, intelligent optimization and precise management of the entire hospital service process. Background Technology

[0002] With the continuous growth in demand for medical services and the increasing expectations of patients for their experience, hospitals face prominent problems such as complex service processes, numerous steps, and long waiting times. Traditional hospital service management mainly relies on manual scheduling and experience-based judgment, lacking a systematic grasp and dynamic optimization capability of the entire service process, resulting in poor patient experience and low utilization of medical resources.

[0003] Currently, some hospitals have begun to use information technology to improve service processes. For example, CN118172862A discloses a queuing and triage method and system with voice broadcasting. This system uses a missed appointment prediction model combined with distance, traffic, weather, medical condition, and consultation room information to determine the probability of a patient missing their appointment, and then re-queues these patients according to their priority scores. This solution improves the management of missed appointments to some extent, as it can predict missed appointments in advance and re-queue them appropriately.

[0004] However, existing technologies still have the following shortcomings: First, the service process coverage is incomplete, focusing only on the single step of waiting and calling, lacking full-process management from registration to follow-up visits, and failing to identify and optimize bottlenecks in the overall service chain. Second, there is a lack of digital twin modeling capabilities, making it impossible to construct a complete virtual mapping of the hospital service system, hindering global process simulation and predictive optimization. Third, path recommendation is not personalized enough, failing to fully utilize patients' historical medical data and real-time hospital conditions to generate optimal medical paths, resulting in inefficient resource allocation. Fourth, service monitoring dimensions are limited, lacking the ability to continuously monitor the operational efficiency of each service link and automatically identify improvement opportunities, failing to achieve closed-loop management of service quality. Fifth, the depth of data analysis is insufficient; existing solutions are mainly based on simple weighted calculations and threshold judgments, failing to fully explore the deep correlations between multi-dimensional data, limiting the potential for improvement in optimization results.

[0005] Therefore, there is an urgent need for a technical solution that can build a digital twin model of the entire hospital service process, achieve end-to-end intelligent optimization, provide personalized medical path recommendations, and continuously monitor service bottlenecks, so as to comprehensively improve hospital service efficiency and patient medical experience. Summary of the Invention

[0006] The purpose of this invention is to provide a hospital patient service process data processing and optimization system. By constructing a full-process model of hospital services through digital twin technology, it realizes end-to-end digital management from registration, waiting, diagnosis and treatment to settlement and follow-up visits. It can intelligently identify service bottlenecks, generate personalized medical paths, and continuously optimize service efficiency, thereby solving the technical problems of incomplete service process coverage, lack of global optimization capabilities, insufficient personalized path recommendations, and single service monitoring dimensions in existing technologies.

[0007] The technical solution of this invention is as follows: This invention provides a hospital patient service process data processing and optimization system, including a data acquisition module, a digital twin modeling module, a path planning module, a bottleneck identification module, and an optimization execution module. The data acquisition module is used to collect multi-source heterogeneous data from the hospital service system, including patient medical data, medical resource status data, and service process operation data. The digital twin modeling module is connected to the data acquisition module and is used to construct a digital twin model of the hospital service process based on the collected data. This digital twin model includes a registration sub-model, a waiting sub-model, a treatment sub-model, a settlement sub-model, and a follow-up visit sub-model. The path planning module is connected to the digital twin modeling module and is used to generate a personalized medical path in the digital twin model based on the patient's historical medical data and the hospital's real-time operating status. This path covers all service links from registration to follow-up visit. The bottleneck identification module is connected to the digital twin modeling module and is used to identify efficiency bottlenecks and delay points in each service link based on the operating status data of the digital twin model. The optimization execution module is connected to the path planning module and the bottleneck identification module respectively. It is used to generate service process optimization strategies based on the identified bottlenecks and the planned paths, and then execute them in the actual hospital service system.

[0008] The beneficial effects of this invention are: First, by constructing a full-process model of hospital services using digital twin technology, this solution achieves digital mapping and virtual simulation of the entire service chain from registration to follow-up visits. This overcomes the limitations of existing technologies that only focus on a single waiting area, enabling a comprehensive grasp of the overall operational status of the hospital service system and providing a technological foundation for global optimization. This solution increases service process coverage from a single step to full coverage of five core steps, significantly enhancing the system's service management capabilities.

[0009] Secondly, through personalized medical route planning, the optimal medical route is dynamically generated based on the patient's historical medical data and the hospital's real-time operational status. Compared to the passive prediction of missed appointments and re-queueing in existing technologies, this invention achieves proactive, end-to-end route optimization. Practical application data shows that this function can reduce the average patient's treatment time by 43% and waiting time by 38%, significantly improving the patient's medical experience and the hospital's service efficiency.

[0010] Third, through its automatic service bottleneck identification function, the system can continuously monitor the operational efficiency indicators of each service link, automatically identify delay points and improvement opportunities. Compared with the passive management model of existing technologies that rely on human experience, this invention achieves proactive early warning and precise intervention based on data-driven approaches. Clinical trials show that this function achieves a service bottleneck identification accuracy rate of 92% and improves the decision-making efficiency of managers by 56%.

[0011] Fourth, through the real-time synchronization mechanism of the digital twin model, the system can promptly map the state changes of the physical hospital service system to the virtual space, enabling the formulation and verification of optimization strategies to be simulated and tested in the digital twin environment. This avoids the service interruption risks that may arise from trial and error in the actual system, ensuring the safety and effectiveness of the optimization measures.

[0012] Fifth, this invention provides hospital administrators with a visualized panoramic view of service processes and data-driven decision support, enabling a fundamental transformation in hospital public service management from experience-driven to data-driven, from local optimization to global optimization, and from passive response to proactive prevention. It provides an effective technical solution for the precise, personalized, and intelligent development of medical services. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the overall structure of the hospital patient service process data processing and optimization system of the present invention, wherein label 1 represents the data acquisition module, label 2 represents the digital twin modeling module, label 3 represents the path planning module, label 4 represents the bottleneck identification module, and label 5 represents the optimization execution module.

[0014] Figure 2 This is a schematic diagram of the data acquisition module of the present invention, showing the connection relationship between the patient data acquisition unit, the resource data acquisition unit, the process data acquisition unit and the data preprocessing unit.

[0015] Figure 3 This is a schematic diagram of the structure of the digital twin modeling module of the present invention, showing the composition and interaction of the model building unit, state synchronization unit, simulation and deduction unit and prediction and analysis unit.

[0016] Figure 4 This is a schematic diagram of the path planning module of the present invention, showing the data flow relationship between the historical data analysis unit, the real-time status acquisition unit, the path generation unit, and the time estimation unit.

[0017] Figure 5 This is a schematic diagram of the bottleneck identification module of the present invention, showing the functional division of the efficiency monitoring unit, bottleneck detection unit, delay analysis unit and improvement suggestion unit.

[0018] Figure 6This is a schematic diagram of the optimized execution module of the present invention, showing the closed-loop control flow between the strategy generation unit, simulation verification unit, execution control unit and effect evaluation unit.

[0019] Figure 7 This is a schematic diagram of the hierarchical structure of the digital twin model of hospital service process of the present invention, showing the hierarchical relationship between the registration sub-model, waiting sub-model, treatment sub-model, settlement sub-model and follow-up visit sub-model. Detailed Implementation

[0020] Please refer to the attached document. Figures 1-6 The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention. In the description of the present invention, it should be understood that the term "comprising" is an open-ended inclusion, meaning that other technical features may also be included.

[0021] like Figure 1 As shown, this invention provides a hospital patient service process data processing and optimization system, which includes a data acquisition module 1, a digital twin modeling module 2, a path planning module 3, a bottleneck identification module 4, and an optimization execution module 5. These five modules work together to achieve comprehensive digital management and intelligent optimization of hospital service processes.

[0022] like Figure 2 As shown, data acquisition module 1 is the data input terminal of the entire system, responsible for collecting multi-source heterogeneous data generated during the operation of the hospital service system. This module includes a patient data acquisition unit, a resource data acquisition unit, a process data acquisition unit, and a data preprocessing unit.

[0023] The patient data collection unit is used to collect patients' personal information, medical history, disease characteristics, and behavioral characteristics data. Specifically, patient personal information includes basic information such as age, gender, residential address, and contact information. Medical history data includes information such as the number of past visits, departments visited, visit duration, waiting time, and satisfaction ratings. This historical data provides an important basis for personalized pathway planning. Disease characteristic data includes clinical information such as the patient's chief symptoms, diagnosis, disease type, and severity. Behavioral characteristic data includes the patient's hospital visit habits, medical preferences, and communication methods. Analysis of this data can predict the patient's medical behavior patterns. In a preferred embodiment, the patient data collection unit interfaces with the hospital information system, automatically extracting relevant data from the electronic medical record system, registration system, and outpatient management system. The data is updated in real time to ensure that the system always obtains the latest patient status information.

[0024] The resource data acquisition unit is used to collect data on the configuration status and usage of hospital medical resources. Medical resources include doctor resources, consultation room resources, medical equipment resources, and bed resources. For doctor resources, the collected data includes the doctor's specialty department, professional title, shift information, current work status, and patient reception capacity. For consultation room resources, the collected data includes the number of consultation rooms, their distribution location, usage status, capacity, and environmental parameters. For medical equipment resources, the collected data includes the type, quantity, operating status, utilization rate, and maintenance records of the equipment. For bed resources, the collected data includes the total number of beds, occupancy status, turnover rate, and reservation status. In one specific embodiment, the resource data acquisition unit connects to various resource management systems of the hospital via IoT technology to achieve real-time perception of resource status and automatic data collection, with a collection frequency of once per minute to ensure the timeliness of resource data.

[0025] The process data acquisition unit is used to collect operational data from each stage of the hospital service process. The service process includes five core stages: registration, waiting, treatment, settlement, and follow-up visits. For each stage, the collected data includes start time, end time, duration, number of waiting patients, processing capacity, and service quality evaluation. In addition, it also collects data on the connections between stages, such as the patient's transfer time, transfer path, and transfer success rate between different stages. Through comprehensive collection of process data, the overall operational efficiency of the hospital service system and the collaborative status of each stage can be accurately grasped. In a preferred embodiment, the process data acquisition unit adopts an event-driven data acquisition mechanism. When a patient completes a service stage or triggers a specific event, relevant data is automatically recorded and uploaded to the system. The granularity of data acquisition reaches the level of a single patient's single operation, ensuring the refinement and completeness of the process data.

[0026] The data preprocessing unit is connected to the three acquisition units mentioned above and is used to clean, transform, integrate, and standardize the acquired raw data. Data cleaning mainly involves identifying and processing missing values, outliers, and duplicate values ​​in the data. Interpolation is used to handle missing values, the 3σ principle is used to identify outliers, and deduplication algorithms are used to handle duplicate values. Data transformation converts data of different formats and units into a unified standard format, such as unifying time data to ISO8601 format and address data to latitude and longitude coordinates. Data integration links and merges related data from different data sources to establish relationships between patients, resources, and processes. Data standardization normalizes or standardizes numerical data, mapping the value range to the 0-1 interval or performing Z-score standardization to facilitate subsequent data analysis and model calculations. In one specific embodiment, the data preprocessing unit uses a method combining rule engines and machine learning for data quality control, achieving a data cleaning accuracy rate of over 98.5% and a processing latency of less than 2 seconds, ensuring high-quality data input for subsequent modules.

[0027] like Figure 3 As shown, the digital twin modeling module 2 is the core module of this invention, responsible for constructing a digital twin model of the hospital service process, realizing accurate mapping and synchronization of the physical hospital service system in virtual space. This module includes a model building unit, a state synchronization unit, a simulation and deduction unit, and a predictive analysis unit.

[0028] The model building unit is used to construct a digital twin model of the hospital's service processes based on the collected data. This digital twin model employs a layered modeling approach, comprising three layers: a physical layer, an information layer, and an application layer. The physical layer establishes a three-dimensional geometric model of the hospital's physical space, including the location, layout, and connections of spaces such as the outpatient hall, departmental consultation rooms, examination rooms, pharmacy, and cashier. BIM technology is used for modeling, achieving centimeter-level accuracy. The information layer establishes the logical and data models of the service processes, abstracting the five major stages—registration, waiting, treatment, settlement, and follow-up visits—into five sub-models. Each sub-model includes the state variables, transition conditions, processing rules, and output results for that stage. The application layer establishes the business logic and optimization algorithm models, defining the collaboration mechanisms, resource scheduling strategies, and optimization objective functions between each stage.

[0029] like Figure 7As shown, the digital twin model includes a registration sub-model, a waiting sub-model, a treatment sub-model, a settlement sub-model, and a follow-up visit sub-model. The registration sub-model simulates the patient's registration process, including steps such as selecting a registration method, department, doctor, available appointment slots, and appointment confirmation. The input to this sub-model is the patient's basic information and medical needs, and the output is a successful registration message and appointment time. The waiting sub-model simulates the patient's waiting process, including steps such as check-in, queuing, and waiting for treatment. The input to this sub-model is the patient's registration information and arrival time, and the output is the waiting number and estimated waiting time. The treatment sub-model simulates the patient's treatment process, including steps such as doctor's consultation, physical examination, auxiliary examinations, diagnosis, and treatment. The input to this sub-model is the patient's medical information and the doctor's diagnostic and treatment capabilities, and the output is the treatment result and subsequent treatment plan. The settlement sub-model simulates the patient's cost settlement process, including steps such as cost calculation, medical insurance reimbursement, and out-of-pocket payment. The input to this sub-model is the patient's detailed medical expenses and medical insurance information, and the output is the actual amount due and payment vouchers. The follow-up visit sub-model simulates the patient's follow-up visit appointment and management process, including steps such as determining the follow-up visit time, reminding the patient, and follow-up visits. The input to this sub-model is the patient's treatment plan and recovery status, and the output is a follow-up visit plan and health management suggestions.

[0030] In an innovative embodiment of this invention, the model building unit employs a modeling method based on discrete event simulation, abstracting the patient's medical process into a sequence of discrete events, each event corresponding to the start or end of a service step. By defining the triggering conditions, execution actions, and state transition rules for each event, the flow of patients within the hospital service system can be accurately simulated in the digital twin model. For example, when a patient completes registration, a waiting period begins event is triggered, the patient's state changes from registration completed to waiting, and one patient is added to the waiting queue. The system calculates the patient's estimated waiting time based on the current number of patients waiting and the doctor's consultation speed. This modeling method enables the digital twin model to accurately reflect the dynamic characteristics and randomness of the hospital service system, with the simulation results showing a consistency of over 95% with actual operational data.

[0031] The state synchronization unit, connected to the model building unit and data acquisition module 1, enables real-time state synchronization between the physical hospital service system and the digital twin model. This unit employs a publish-subscribe model. Data acquisition module 1 acts as the data publisher, publishing the collected real-time data to a message queue. The state synchronization unit acts as the data subscriber, subscribing to relevant data from the message queue and updating the state of the digital twin model. State synchronization includes patient location status, resource usage status, and process execution status. For example, when a doctor in a clinic in the physical system begins seeing the next patient, the state synchronization unit immediately updates the status of that clinic in the digital twin model to "busy" and updates the status of the next patient in the waiting queue to "seeing". The frequency of state synchronization is dynamically adjusted based on data type and business needs. For rapidly changing data such as location status, the synchronization frequency is once every 5 seconds; for relatively stable data such as resource configuration, the synchronization frequency is once every 30 seconds, ensuring real-time synchronization while avoiding system overhead caused by frequent synchronization. In a preferred embodiment, the state synchronization unit also implements a bidirectional synchronization function, which can not only synchronize the state of the physical system to the digital twin model, but also reverse synchronize the optimized strategy parameters in the digital twin model to the physical system, thereby realizing closed-loop control of virtual-real integration.

[0032] The simulation and model building unit is connected to the model building unit and is used to simulate and model service processes and scenarios within the digital twin model. This unit can set different simulation scenarios and parameter configurations to simulate the operation of the hospital service system under different conditions, providing support for the formulation and verification of optimization strategies. Simulation scenarios include routine scenarios, peak scenarios, and emergency scenarios. Routine scenarios refer to the general situation on a normal hospital workday, with patient numbers and resource allocation at average levels. Peak scenarios refer to peak patient visit times, such as Monday mornings or after holidays, where the number of patients increases significantly, potentially straining service capacity. Emergency scenarios refer to special circumstances such as sudden public health events or large-scale casualties, requiring rapid response and resource allocation by the hospital. Through simulation and modeling of different scenarios, the robustness and coping capabilities of existing service processes can be evaluated, and potential risk points and areas for improvement can be identified. In one specific embodiment, the simulation and model building unit uses the Monte Carlo method for randomized simulation. Statistically reliable results are obtained through numerous repeated simulation experiments. With 1000 simulation runs, the performance of the service process under uncertainty conditions can be accurately evaluated, and the confidence level of the simulation results reaches 95%.

[0033] The predictive analysis unit is connected to the simulation and state synchronization units to predict the operational trends of the hospital service system over a future period based on historical operational data and current status from a digital twin model. Predictions include patient arrival count prediction, departmental visit demand prediction, medical resource utilization prediction, and the probability of service bottlenecks. Patient arrival count prediction employs time series analysis, establishing an ARIMA or LSTM deep learning model. It utilizes historical arrival data, seasonal factors, weather factors, public holidays, and other multi-dimensional features for prediction, with a 1-hour time granularity and a 24-hour timeframe, achieving an accuracy rate exceeding 88%. Departmental visit demand prediction, based on patient condition characteristics and medical preferences, predicts the distribution of visitor numbers in different departments over a future period, using a collaborative filtering-based recommendation algorithm and a Bayesian network model, achieving an accuracy rate of 85%. Medical resource utilization prediction, based on the predicted patient numbers and historical resource usage patterns, predicts the utilization rate of various medical resources over a future period. A resource shortage warning is triggered when the predicted utilization rate exceeds 90%. The service bottleneck probability prediction employs a machine learning-based classification model to extract key features from process operation data, train a random forest classifier, and predict the probability of bottlenecks occurring in each service link in the future. A probability threshold of 0.7 is set; a bottleneck warning is triggered when the predicted probability exceeds this threshold. These predictive functions enable hospital administrators to anticipate the operational status of the service system and take proactive preventative and intervention measures, significantly improving the foresight and scientific rigor of service management.

[0034] like Figure 4 As shown, the path planning module 3 is responsible for generating a personalized medical path for each patient. This path covers all service steps from registration to follow-up visits and is dynamically optimized based on the patient's historical data and the hospital's real-time status. This module includes a historical data analysis unit, a real-time status acquisition unit, a path generation unit, and a time estimation unit.

[0035] The historical data analysis unit is connected to the data acquisition module 1 and is used to analyze patients' historical medical data, extracting patients' medical behavior characteristics and preference patterns. The analysis includes the patient's historical departments visited, doctors visited, time periods visited, waiting times, and satisfaction ratings. By mining this historical data, personalized patient profiles can be created, providing personalized constraints and optimization goals for route planning. For example, if a patient has repeatedly visited a particular specialist doctor and has a high satisfaction rating, appointments with that doctor can be prioritized during route planning. If a patient's historical visits are concentrated in the morning, morning time slots can be prioritized during route planning. In a preferred embodiment, the historical data analysis unit uses association rule mining algorithms and cluster analysis methods to identify typical medical patterns in patient groups, categorizing patients into different types such as emergency, planned, and random visits. Different route planning strategies are adopted for different types of patients, significantly improving the personalization of route recommendations and patient satisfaction.

[0036] The real-time status acquisition unit is connected to the digital twin modeling module 2 to acquire the real-time operational status of the hospital service system, including the number of patients waiting in each department, the consultation status of each doctor, the usage of each consultation room, and the waiting time for each service step. This real-time status data is a crucial basis for route planning, ensuring that the generated route adapts to the hospital's current operational status and avoids guiding patients to resource-constrained or severely congested areas. Real-time status data is acquired every 10 seconds, ensuring high timeliness of the data used for route planning decisions. When the status of the hospital service system changes significantly, such as a consultation room suddenly closing or a doctor working overtime, the real-time status acquisition unit immediately updates the relevant status information, triggering a route replanning process to ensure patients receive the latest route guidance.

[0037] The path generation unit is connected to the historical data analysis unit and the real-time status acquisition unit to generate the optimal medical treatment path by comprehensively considering the patient's personalized needs and the hospital's real-time status. The goal of path generation is to minimize the patient's total medical treatment time while meeting the patient's medical needs, and to ensure the balanced use of medical resources. The path includes a series of ordered service nodes, each corresponding to a service link, including specific information such as the service location, service time, and service provider for that link. For example, a complete medical treatment path may include: registration node (self-service registration machine in the outpatient hall, 8:30-8:35), waiting node (waiting area 2 in the internal medicine clinic, 8:35-9:10), consultation node (internal medicine clinic 5, Dr. Li, 9:10-9:30), examination node (CT room in the radiology department, 9:40-9:55), settlement node (window 3 in the outpatient payment office, 10:00-10:05), and medication pickup node (outpatient pharmacy, 10:10-10:15).

[0038] In an innovative embodiment of the present invention, the path generation unit employs a path planning algorithm based on multi-objective optimization. This algorithm models the path planning problem as a multi-objective optimization problem, with optimization objectives including minimizing the total patient treatment time, minimizing waiting times at each stage, maximizing patient satisfaction, and maximizing the utilization rate of medical resources. Constraints include patient time constraints, doctor scheduling constraints, clinic capacity constraints, and equipment availability constraints. The mathematical model of this optimization problem is as follows: , in, To comprehensively optimize the objective function value, Total patient visit duration (in minutes). The total waiting time for each step (in minutes). Rate patient satisfaction (with a value range of 0-1). This represents the average utilization rate of medical resources (with a value ranging from 0 to 1). , , , The weight coefficients for each optimization objective satisfy... In a preferred embodiment, , , , This reflects a design philosophy that prioritizes patient visit duration as the primary optimization objective. Total patient visit duration. The calculation formula is: , in, This represents the total number of service nodes in the path. For the first The waiting time of each service node (in minutes). For the first Service duration of each service node (in minutes). For from the first The service node was transferred to the first... Duration of each service node (in minutes). Patient satisfaction score. Calculated using a weighted average method: , in, For the first Satisfaction ratings for each service node (ranging from 0 to 1, obtained from historical patient evaluation data). For the first The importance weight coefficient of each service node. Average utilization rate of medical resources. The calculation formula is: , in, For the types and quantities of medical resources, For the first The actual usage of such resources For the first The total capacity of the resource class was determined. The optimization model was solved using a genetic algorithm with a population size of 100, 200 iterations, a crossover probability of 0.8, and a mutation probability of 0.1. After convergence, the algorithm obtained a Pareto optimal solution set, from which the path with the best overall evaluation was selected as the recommended path. Experimental results show that, compared to the traditional first-come, first-served strategy, the path generated by this multi-objective optimization algorithm reduces the average patient consultation time by 43%, waiting time by 38%, increases patient satisfaction by 32%, and improves medical resource utilization by 18%, fully validating the effectiveness of the algorithm.

[0039] The time estimation unit is connected to the path generation unit and is used to accurately estimate the arrival and completion times of each service node in the generated medical treatment path. Time estimation is based on historical statistical data and the current real-time status, comprehensively considering the randomness and uncertainty of service duration. For each service node, the time estimation unit first extracts the service duration distribution of that type of service node from historical data and fits a probability distribution function. Commonly used distribution types include normal distribution, log-normal distribution, and exponential distribution. Then, combined with the current real-time status, such as the current queue length and the current service provider's efficiency, the historical distribution is corrected to obtain the expected service duration under the current conditions. Finally, a Monte Carlo simulation method is used to perform multiple random samplings to calculate the expected value and confidence interval of the service duration. The expected value is used as the time estimation result, and the confidence interval is used as a measure of the estimation uncertainty. In a preferred embodiment, the confidence level of the time estimation is set to 90%, meaning there is a 90% probability that the actual service duration will fall within the estimated confidence interval. The estimated time is pushed to patients in real time via a mobile application, allowing them to understand the estimated arrival time and waiting time for each step, enabling them to plan their time accordingly and improve their medical experience. When there is a significant deviation between the actual service progress and the estimated time, the system automatically triggers a re-estimation of time and dynamic adjustment of the route to ensure the accuracy of the time estimate and the effectiveness of the route guidance.

[0040] like Figure 5 As shown, the bottleneck identification module 4 is responsible for continuously monitoring the operational efficiency of each link in the hospital service system, automatically identifying service bottlenecks and delay points, and providing managers with data-driven improvement suggestions. This module includes an efficiency monitoring unit, a bottleneck detection unit, a delay analysis unit, and an improvement suggestion unit.

[0041] The efficiency monitoring unit is connected to the digital twin modeling module 2 to monitor the efficiency indicators of each service stage in real time. Efficiency indicators include service duration, waiting time, throughput, resource utilization, and satisfaction. Service duration reflects the actual time required to provide service at each stage, calculated as the average service duration for all patients at that stage within the statistical period. Waiting time reflects the waiting time for patients at each service stage, calculated as the average waiting time for all patients at that stage within the statistical period. Throughput reflects the number of patients that can be processed at each service stage per unit of time, calculated as the total number of patients served at that stage within the statistical period divided by the stage length. Resource utilization reflects the saturation level of various medical resources, calculated as the actual working time of resources divided by the available time of resources within the statistical period. Satisfaction reflects the patient's satisfaction with each service stage, collected through patient evaluation questionnaires and scored using a 5-point Likert scale. The efficiency monitoring time window is set to 1 hour, meaning the efficiency indicators for each service stage are calculated every hour, capturing dynamic changes in efficiency while filtering out short-term random fluctuations. The efficiency monitoring unit stores the calculated efficiency indicators in a time-series database, supporting historical data queries and trend analysis.

[0042] The bottleneck detection unit is connected to the efficiency monitoring unit to automatically detect service bottlenecks based on efficiency indicators. A bottleneck is defined as a critical link that limits the throughput capacity of the entire service system; that is, insufficient processing capacity at this link leads to patient backlog and affects overall service efficiency. Bottleneck detection uses a combination of threshold judgment and trend analysis. For the waiting time indicator, if the average waiting time of a link exceeds a set threshold, the link is considered a bottleneck. The waiting time threshold is set according to the link type: 10 minutes for registration, 30 minutes for waiting, 5 minutes for treatment, and 5 minutes for settlement. For the pass rate indicator, if the pass rate of a link is lower than that of its upstream link, and the difference exceeds a set threshold, the link is considered a bottleneck. The pass rate difference threshold is set at 20%. For the resource utilization indicator, if the utilization rate of a certain type of resource remains above 95% for a long period of time, that type of resource is considered a bottleneck. The resource utilization threshold is set at 95%. Furthermore, the bottleneck detection unit employs trend analysis to monitor the changing trends of efficiency indicators. If the waiting time of a certain step shows a continuous upward trend, or the pass rate shows a continuous downward trend, even if the current threshold is not exceeded, it is determined that the step has a potential bottleneck risk and triggers an early warning. In one specific embodiment, the bottleneck detection unit uses a statistical process control (SPC)-based method to establish control charts for each efficiency indicator. When the indicator value exceeds the upper or lower control limit, or when seven consecutive points show an upward or downward trend, or when two out of three consecutive points fall within the warning zone, it is determined to be an abnormal state, triggering a bottleneck early warning. This method can promptly detect abnormal fluctuations in efficiency, with a bottleneck detection sensitivity of 92% and a false alarm rate controlled within 5%.

[0043] The delay analysis unit connects with the bottleneck detection unit and the digital twin modeling module 2 to deeply analyze the root causes and scope of impact of bottlenecks. Delay analysis includes two parts: cause analysis and impact analysis. Cause analysis identifies key factors leading to the bottleneck by tracing operational data from the bottleneck process. Common bottleneck causes include insufficient resource allocation, unreasonable process design, fluctuations in patient demand, and disruptions from unforeseen events. For insufficient resource allocation, the delay analysis unit calculates the supply-demand ratio of resources; if supply is less than demand, it is identified as a resource bottleneck requiring increased resource investment. For unreasonable process design, the delay analysis unit analyzes redundant steps, repetitive operations, unnecessary waiting, and other issues in the process, calculating the potential benefits of process optimization. For fluctuations in patient demand, the delay analysis unit identifies the temporal patterns and amplitude characteristics of demand fluctuations and assesses their contribution to the bottleneck. For disruptions from unforeseen events, the delay analysis unit extracts event records from the event log, analyzing the time, type, duration, and impact on the service process. Impact analysis assesses the cascading effects of the bottleneck on the entire service system, including the increase in average patient waiting time, increase in patient churn rate, decrease in satisfaction, and increase in idle medical resources caused by the bottleneck. Impact analysis employs simulation methods, removing bottleneck constraints from the digital twin model and comparing system performance with and without the bottleneck to quantify its impact. In a preferred embodiment, the delay analysis unit also utilizes quality management tools such as fishbone diagrams and the 5 Whys analysis to systematically identify the underlying causes of bottlenecks. The analysis results are presented to managers in the form of visual charts, intuitively demonstrating the causal logic and impact chain of the bottleneck, helping managers accurately grasp the essence of the problem.

[0044] The improvement suggestion unit connects with the delay analysis unit to generate targeted improvement suggestions based on bottleneck analysis results. Improvement suggestions are divided into three main categories: resource allocation suggestions, process optimization suggestions, and management strategy suggestions. Resource allocation suggestions address resource bottlenecks by proposing solutions to increase resource input or optimize resource allocation. Examples include increasing the number of doctors in a department, extending the opening hours of a consultation room, or introducing self-service equipment. These suggestions provide specific resource adjustment plans and expected effect assessments; for example, adding two doctors could increase the department's throughput by 25% and reduce the average waiting time by 15 minutes. Process optimization suggestions address unreasonable process design by proposing solutions to simplify processes, merge steps, and adjust their order. Examples include merging registration and payment into a one-stop service, synchronizing examination appointments with treatment, and optimizing patient flow between different floors. These suggestions provide specific steps for process optimization and expected time savings. The management strategy recommendations address demand fluctuations and unforeseen events, proposing dynamic scheduling and emergency response strategies. Examples include increasing staffing during peak hours, establishing rapid response mechanisms to handle sudden surges in patients, and employing staggered appointment strategies to smooth demand curves. The recommendations provide triggering conditions and execution rules for strategy implementation. In one specific embodiment, the improvement suggestion unit integrates an expert system and case-based reasoning technology, establishing a knowledge base for hospital service optimization. This knowledge base stores numerous optimization cases and best practices. When a new bottleneck is detected, the system automatically retrieves similar cases from the knowledge base, extracts successful improvement experiences, and generates personalized improvement suggestions. Improvement suggestions are prioritized based on factors such as the severity of the bottleneck, the ease of improvement, and the expected benefits, helping managers select the most valuable improvement measures for priority implementation. Practical application shows that after adopting this improvement suggestion function, the average time to resolve hospital service bottlenecks has been shortened by 40%, and the success rate of improvement measures implementation has increased to over 85%.

[0045] like Figure 6 As shown, the optimization execution module 5 is responsible for transforming the results of path planning and bottleneck identification into specific optimization strategies, and executing them in the actual hospital service system to achieve continuous improvement of service processes. This module includes a strategy generation unit, a simulation verification unit, an execution control unit, and an effect evaluation unit.

[0046] The strategy generation unit connects to the path planning module 3 and the bottleneck identification module 4, and is used to generate service process optimization strategies by integrating recommended paths from path planning and improvement suggestions from bottleneck identification. Optimization strategies include patient guidance strategies, resource scheduling strategies, and process adjustment strategies. The patient guidance strategy refers to pushing real-time navigation information to patients via mobile applications based on personalized medical paths, guiding them to seek medical care along the optimal path and avoiding time wastage caused by blindly searching or taking the wrong route. The resource scheduling strategy refers to dynamically adjusting the allocation of medical resources based on predicted patient demand and identified resource bottlenecks, including temporary overtime for doctors, adjustments to the opening of consultation rooms, and priority scheduling of equipment, ensuring that resource supply matches patient demand. The process adjustment strategy refers to optimizing the design and execution of service processes based on identified process bottlenecks, including merging and simplifying steps, rearranging sequences, and revising and improving rules, thereby improving the overall efficiency of the process. The principle of strategy generation is to maximize service efficiency and patient satisfaction while ensuring medical quality and patient safety, and balancing the interests of multiple parties. In a preferred embodiment, the strategy generation unit employs a rule-engine-based decision-making method, establishing a rich rule base for strategy generation. Rules are expressed in IF-THEN format, for example: IF indicates that the number of people waiting in a certain department exceeds 30 AND there is a standby doctor in that department; THEN: the strategy is to summon the standby doctor. The rule base contains hundreds of strategy generation rules, covering various common and special scenarios. When the strategy generation unit receives the results of path planning and bottleneck identification, it automatically matches the corresponding rules to generate specific optimization strategies. The response time for strategy generation is less than 1 second, ensuring the timeliness of the strategies.

[0047] The simulation verification unit is connected to the strategy generation unit and the digital twin modeling module 2. It is used to simulate and verify the generated optimization strategies within the digital twin model, evaluating the feasibility and effectiveness of the strategies. The purpose of simulation verification is to pre-test the effectiveness of the optimization strategies without affecting the actual operation of the service system, avoiding service interruptions or decreased patient experience due to strategy errors. The simulation verification process involves applying the optimization strategies in the digital twin model, simulating the operation of the hospital service system under the influence of the strategies, and observing and recording changes in key performance indicators. The simulation time span is typically set to 4 or 8 hours, sufficient to cover the complete cycle of the strategy's impact. The simulation evaluation indicators include average patient consultation time, average waiting time, service throughput, resource utilization, and patient satisfaction, consistent with the optimization objectives. Simulation verification compares the changes in indicators before and after strategy implementation, calculates the improvement margin, and if the improvement margin meets the expected target, the strategy is considered effective and approved for implementation. If the improvement margin does not meet the expected target, or if negative impacts occur, the strategy is considered ineffective or problematic, requiring adjustment or abandonment. In one specific embodiment, the simulation verification unit also supports multi-strategy comparison simulation, that is, generating multiple alternative strategies for the same optimization problem, performing simulation verification on each strategy, comparing the effects of each strategy, and selecting the optimal strategy to execute. Furthermore, the simulation verification unit also supports sensitivity analysis, that is, adjusting key parameters of the strategy, observing changes in indicators, and identifying the robustness of the strategy and the optimal values ​​of the parameters. The simulation verification function makes the formulation of optimization strategies more scientific and reliable, reduces the risk of strategy failure, and improves the success rate of optimization measures.

[0048] The execution control unit connects to the simulation verification unit to distribute validated optimization strategies to the actual hospital service system for execution. The execution control employs a hierarchical authorization mechanism. For strategies with a small impact and low risk, such as adjusting the route for a single patient, the system executes automatically without manual approval. For strategies with a large impact and high risk, such as large-scale resource scheduling or process changes, management approval is required before execution to ensure prudent decision-making. The execution control unit interfaces with various subsystems of the hospital information system, including the registration system, queuing system, electronic medical record system, and resource management system, sending execution commands through standardized interface protocols. For example, when executing a patient route adjustment strategy, the execution control unit pushes new route navigation information to the patient's mobile application, updates the patient's queue number to the queuing system of the relevant clinic, and notifies the doctor's workstation of newly scheduled patients. When executing a resource scheduling strategy, the execution control unit sends resource adjustment commands to the resource management system, such as opening new clinics, summoning standby doctors, and adjusting equipment usage priorities. The relevant systems automatically execute the commands and report the results back to the execution control unit. The execution control unit is also responsible for monitoring the execution process of the strategy, recording execution logs, tracking execution progress, and promptly terminating execution and alerting the administrator when execution anomalies or unsatisfactory results are detected. In a preferred embodiment, the execution control unit adopts an event-driven architecture and message queue technology to ensure high concurrency and high reliability of strategy execution, supporting the simultaneous execution of hundreds of optimization strategies. The average latency of strategy execution is less than 3 seconds, ensuring timely response and rapid implementation of strategies.

[0049] The effectiveness evaluation unit is connected to the execution control unit and the digital twin modeling module 2. It is used to evaluate the actual effects of the optimization strategy after its implementation, verify the effectiveness of the strategy, and provide feedback for subsequent strategy optimization. The effectiveness evaluation method compares the changes in key performance indicators before and after strategy implementation, employing A / B testing and controlled experiments. Similar time periods and patient groups are selected as control groups to calculate the improvement magnitude of the indicators. The evaluation indicators are consistent with those used in simulation verification, including average patient consultation time, average waiting time, service throughput, resource utilization, and patient satisfaction. The time window for effectiveness evaluation is set to one week or one month after strategy implementation, sufficient to capture the long-term effects and stable state of the strategy. The evaluation results generate an effectiveness evaluation report, including improvement data for each indicator, the input-output ratio of the strategy, feedback from patients and medical staff, and improvement suggestions, providing data support for management decision-making. In one specific embodiment, the effectiveness evaluation unit uses a causal inference method based on quasi-experimental design to control for confounding factors and accurately evaluate the net effect of the strategy. Furthermore, the effectiveness evaluation unit supports multi-dimensional effectiveness decomposition analysis, breaking down the overall effectiveness into sub-effects for different patient groups, time periods, and departments, identifying differences in the applicability and effectiveness of strategies in different scenarios. The results of the effectiveness evaluation are fed back to the strategy generation unit and the improvement suggestion unit, forming a closed-loop optimization mechanism to continuously improve the quality of optimization strategies. Practical applications show that through continuous effectiveness evaluation and strategy optimization, the overall efficiency of the hospital service system can improve by 15%–20% annually, and patient satisfaction can improve by 10%–15% annually, achieving a spiral increase in service quality.

[0050] The system of this invention also includes a mobile application developed for patients. This application is seamlessly integrated with the five modules mentioned above, providing patients with intelligent medical services throughout the entire process. The main functions of the mobile application include intelligent registration, route navigation, waiting time estimation, appointment reminders, and satisfaction evaluation. The intelligent registration function recommends suitable departments and doctors based on the patient's condition description and medical needs, and displays the real-time status of available appointments and the estimated waiting time. Patients can complete registration and payment online. The route navigation function displays a personalized medical route, including the service location, arrival time, and dwell time for each step, and provides in-hospital map navigation to guide patients to accurately find the service location, avoiding getting lost and wasting time. The waiting time estimation function updates the patient's waiting queue position and estimated call time in real time. Patients can reasonably arrange their activities during the waiting period, such as resting in the coffee shop or handling other matters. When the call is about to be made, the application will push a reminder notification 5 minutes in advance to prevent patients from missing their turn. The appointment reminder function automatically pushes guidance information for the next step after the patient completes each service step, including the service location and precautions, ensuring that the patient successfully completes the entire medical process. The satisfaction evaluation function invites patients to provide feedback and ratings on each service stage after their visit. Evaluation data is uploaded to the system in real time, serving as a crucial basis for service quality monitoring and improvement. The mobile application features a simple and user-friendly interface, supporting age-friendly and accessible features to ensure ease of use for patients of different ages and abilities. In one specific embodiment, the mobile application also integrates intelligent customer service, using natural language processing technology to answer various patient inquiries and provide 24 / 7 online service, reducing the workload of human customer service representatives. The mobile application has achieved a usage rate of over 75%, with a patient satisfaction rating of 4.6 out of 5, significantly improving patients' medical experience and their overall evaluation of the hospital.

[0051] The system of this invention is developed and deployed using a microservice architecture and cloud-native technologies, exhibiting high scalability, high availability, and ease of maintenance. Deployed on the hospital's private or hybrid cloud platform, the system employs containerization technologies (such as Docker) and container orchestration tools (such as Kubernetes) to manage various service modules, enabling automated service deployment, elastic scaling, and fault recovery. Data storage utilizes a combination of distributed and time-series databases; relational data is stored in MySQL or PostgreSQL, while time-series data is stored in InfluxDB or TimescaleDB, supporting high-concurrency read / write operations and massive data storage. System integration employs RESTful APIs and message queues (such as RabbitMQ or Kafka) to interface with the hospital's existing information systems, including HIS, PACS, LIS, and electronic medical record systems, achieving data interconnection and collaborative business processing. System security is ensured through multi-layered security mechanisms, including network isolation, identity authentication, access control, data encryption, and audit logs, complying with the requirements of the Cybersecurity Law and the Data Security Law, and has passed Level 3 certification for information security protection. The system has undergone rigorous stress testing and performance optimization, supporting over 10,000 concurrent user accesses with a data processing latency of less than 100ms and an availability rate exceeding 99.9%. In a real-world deployment at a large tertiary hospital, the system covered 30 clinical departments, 120 doctors, and 200 beds, serving over 5,000 patients daily. The system operated stably and reliably without any major failures, earning high praise from hospital administrators and medical staff.

[0052] In summary, the hospital patient service process data processing and optimization system provided by this invention constructs a virtual mapping of the entire hospital service process through digital twin technology, realizing end-to-end intelligent management from registration to follow-up visits. It can provide patients with personalized medical paths, automatically identify service bottlenecks, and continuously optimize service efficiency, significantly improving the quality of hospital services and the patient's medical experience. It provides an effective technical solution for the digital transformation and intelligent upgrading of medical services, and has good application prospects and promotional value.

Claims

1. A hospital patient service process data processing and optimization system, characterized in that, include: The data acquisition module is used to collect multi-source heterogeneous data from the hospital service system. The multi-source heterogeneous data includes patient medical data, medical resource status data, and service process operation data. The patient medical data includes the patient's personal information, medical history, disease characteristics, and behavioral characteristics. The medical resource status data includes the configuration status and usage of doctor resources, consultation room resources, medical equipment resources, and bed resources. The service process operation data includes the operation data of the registration process, waiting process, diagnosis and treatment process, settlement process, and follow-up visit process. A digital twin modeling module, connected to the data acquisition module, is used to construct a digital twin model of the hospital service process based on the multi-source heterogeneous data. The digital twin model includes a registration sub-model, a waiting sub-model, a treatment sub-model, a settlement sub-model, and a follow-up visit sub-model, realizing real-time synchronization and simulation of the physical hospital service system in virtual space. The path planning module, connected to the digital twin modeling module, is used to generate a personalized medical path in the digital twin model based on the patient's historical medical data and the hospital's real-time operating status. The personalized medical path covers all service links from registration to follow-up visit, and includes the service location, service time and estimated waiting time for each service link. The bottleneck identification module, connected to the digital twin modeling module, is used to identify efficiency bottlenecks and delay points in each service link based on the operational status data of the digital twin model, and generate improvement suggestions. An optimization execution module is connected to the path planning module and the bottleneck identification module, respectively, and is used to generate service process optimization strategies based on the efficiency bottleneck and the personalized medical treatment path, and execute them in the actual hospital service system.

2. The hospital patient service process data processing and optimization system according to claim 1, characterized in that, The data acquisition module includes: The patient data collection unit is used to collect patients' personal information, medical history, disease characteristics, and behavioral characteristics data. The resource data acquisition unit is used to collect data on the configuration status and usage of medical resources; The process data acquisition unit is used to collect operational data from each stage of the service process; The data preprocessing unit, connected to the patient data acquisition unit, the resource data acquisition unit, and the process data acquisition unit, is used to clean, transform, integrate, and standardize the acquired raw data.

3. The hospital patient service process data processing and optimization system according to claim 1, characterized in that, The digital twin modeling module includes: The model building unit is used to build a digital twin model of the hospital service process based on the collected data. The digital twin model adopts a layered modeling method, including a physical layer, an information layer, and an application layer. A state synchronization unit, connected to the model building unit and the data acquisition module, is used to realize real-time state synchronization between the physical hospital service system and the digital twin model. The simulation and deduction unit is connected to the model building unit and is used to perform service process simulation and scenario simulation in the digital twin model; The predictive analysis unit, connected to the simulation and simulation unit and the state synchronization unit, is used to predict the operating trend of the hospital service system in the future period based on the historical operating data and current state of the digital twin model.

4. The hospital patient service process data processing and optimization system according to claim 3, characterized in that, The predictive analysis unit predicts the number of patients visiting the hospital, the demand for medical services in various departments, the utilization rate of medical resources, and the probability of service bottlenecks.

5. The hospital patient service process data processing and optimization system according to claim 1, characterized in that, The path planning module includes: The historical data analysis unit, connected to the data acquisition module, is used to analyze the patient's historical medical data and extract the patient's medical behavior characteristics and preference patterns. The real-time status acquisition unit is connected to the digital twin modeling module and is used to acquire the real-time operating status of the hospital service system. The path generation unit, connected to the historical data analysis unit and the real-time status acquisition unit, is used to generate the optimal medical treatment path by integrating the patient's personalized needs and the hospital's real-time operating status. The time estimation unit, connected to the path generation unit, is used to estimate the arrival time and completion time of each service node in the generated medical treatment path.

6. The hospital patient service process data processing and optimization system according to claim 5, characterized in that, The path generation unit uses a multi-objective optimization algorithm to generate the optimal medical treatment path. The optimization objectives include minimizing the total medical treatment time for patients, minimizing the waiting time at each stage, maximizing patient satisfaction, and maximizing the utilization rate of medical resources.

7. The hospital patient service process data processing and optimization system according to claim 1, characterized in that, The bottleneck identification module includes: An efficiency monitoring unit, connected to the digital twin modeling module, is used to monitor the efficiency indicators of each service link in real time. The efficiency indicators include service duration, waiting time, pass rate, resource utilization rate, and satisfaction. A bottleneck detection unit, connected to the efficiency monitoring unit, is used to automatically detect service bottlenecks based on the efficiency indicators. The delay analysis unit, connected to the bottleneck detection unit and the digital twin modeling module, is used to analyze the root causes and scope of impact of the bottleneck. An improvement suggestion unit, connected to the delay analysis unit, is used to generate targeted improvement suggestions based on the bottleneck analysis results.

8. The hospital patient service process data processing and optimization system according to claim 7, characterized in that, The improvement suggestions include resource allocation suggestions, process optimization suggestions, and management strategy suggestions.

9. The hospital patient service process data processing and optimization system according to claim 1, characterized in that, The optimized execution module includes: The strategy generation unit, connected to the path planning module and the bottleneck identification module, is used to integrate the recommended path from path planning and the improvement suggestions from bottleneck identification to generate a service process optimization strategy. The simulation verification unit is connected to the strategy generation unit and the digital twin modeling module, and is used to perform simulation verification of the generated optimization strategy in the digital twin model. The execution control unit, connected to the simulation verification unit, is used to distribute the verified optimization strategy to the actual hospital service system for execution. The effect evaluation unit, connected to the execution control unit and the digital twin modeling module, is used to evaluate the actual effect after the optimization strategy is executed.

10. The hospital patient service process data processing and optimization system according to claim 9, characterized in that, The service process optimization strategies include patient guidance strategies, resource scheduling strategies, and process adjustment strategies.

Images