Outpatient service information query reservation management system and method based on big data

Abstract

The invention discloses an outpatient service information query and reservation management system and method based on big data, and the system comprises an outpatient service information collection and processing module, a medical map construction and retrieval module, a number source balanced distribution module, a flow dynamic prediction module, a number source elastic delivery module, and a reservation interaction management module. Wherein the number source balanced distribution module, the flow dynamic prediction module and the number source elastic release module respectively comprise a plurality of functional units. The outpatient information acquisition and processing module integrates multi-dimensional data and transmits the data to the medical map construction and retrieval module, and the module constructs an associated map and retrieves matching information; the number source balanced distribution module calculates a number source distribution proportion in combination with data, and the flow dynamic prediction module outputs a flow prediction result by integrating multiple factors; the number source elastic delivery module determines a number source delivery strategy according to the distribution proportion and the prediction result, and the reservation interaction management module provides a query reservation service and feeds back data. Through multi-module cooperation, the resource utilization rate and the intelligent level of outpatient service management are improved.

Description

Technical Field

[0001] This invention relates to the field of outpatient information query and management technology, and in particular to an outpatient information query and appointment management system and method based on big data. Background Technology

[0002] In the current healthcare system, outpatient clinics, as the primary link in patient visits, face challenges such as a continuous increase in patient numbers, diversified medical needs, and uneven distribution of medical resources. With the gradual application of big data technology in the medical field, outpatient information inquiry and appointment management needs to integrate multi-dimensional data, including basic patient information, historical medical records, doctor's appointment schedule, and clinic resource data, to achieve efficient coordination in appointment allocation, traffic control, and appointment services. In traditional outpatient management models, information collection and processing are often decentralized, making it difficult to form a unified data support system. This results in cumbersome patient appointment processes, low matching between doctor resources and patient needs, and a lack of accurate prediction of outpatient traffic fluctuations, making it difficult to adapt appointment allocation to changes in actual medical needs. Therefore, there is an urgent need to build an integrated outpatient information inquiry and appointment management system based on big data, which can improve the precision and intelligence of outpatient management through multi-module collaboration and the application of model algorithms.

[0003] Existing outpatient information inquiry and appointment management technologies have two significant drawbacks: First, the allocation of appointment slots lacks a dynamic balancing mechanism, relying heavily on fixed ratios or manual adjustments. This fails to fully integrate multi-dimensional data such as patient symptom type, doctor's expertise, and clinic resource status for accurate allocation, resulting in some departments having scarce slots while others have idle slots. Furthermore, appointment allocation is not linked to real-time appointment progress and historical usage, making it difficult to flexibly adjust the timing and quantity of slots based on actual demand, thus reducing appointment utilization efficiency. Second, outpatient traffic prediction methods are simplistic, failing to comprehensively consider factors such as holidays, seasonal disease peaks, and emergency-to-outpatient transfers. Relying solely on historical traffic data for simple calculations leads to significant discrepancies between predictions and actual traffic, failing to provide a reliable basis for appointment allocation and resource scheduling. This results in resource congestion during peak hours, excessively long patient wait times, and resource waste during off-peak hours. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides a big data-based outpatient information query and appointment management system and method.

[0005] The technical solution adopted in this invention is an outpatient information query and appointment management system based on big data, comprising: an outpatient information collection and processing module, a medical graph construction and retrieval module, an appointment resource balancing allocation module, a traffic dynamic prediction module, an appointment resource elastic allocation module, and an appointment interaction management module. The outpatient information collection and processing module collects and structures basic information of outpatients, historical medical records, doctor's outpatient information, and clinic resource data in real time, and transmits the processed data to the medical graph construction and retrieval module. The medical graph construction and retrieval module constructs a medical graph including patient-doctor-disease-clinic relationships based on the received structured data, and retrieves patient needs and doctor resource matching information through a triple-link retrieval logic. The retrieval results are transmitted to the appointment resource balancing allocation module and the appointment interaction management module, respectively. The appointment resource balancing allocation module calls the outpatient appointment resource dynamic equalization module. The balanced allocation model, combining medical atlas retrieval results and outpatient resource data, calculates the allocation ratio of appointment slots for each department and transmits the allocation results to the appointment slot flexible allocation module. The traffic dynamic prediction module constructs an outpatient traffic dynamic prediction and analysis platform based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease incidence patterns, outputting the predicted outpatient traffic for each department within a preset time period to the appointment slot flexible allocation module. The appointment slot flexible allocation module calls the dynamic appointment slot flexible allocation scheduling algorithm to determine the number of appointment slots and the timing of allocation for each time period based on the appointment slot allocation ratio and traffic prediction results, and transmits the allocation instructions to the appointment interaction management module. The appointment interaction management module receives the medical atlas retrieval results and appointment slot allocation instructions, providing patients with functions such as appointment slot query, appointment application submission, and appointment status feedback, while also feeding back patient appointment operation data to the outpatient information collection and processing module.

[0006] Furthermore, the expression for the dynamic balanced allocation model of outpatient appointment slots in the appointment slot allocation module is as follows: ,in, Let t be the number of appointment slots allocated to the j-th doctor in the i-th department at time t. Assign weight coefficients to the number sources. Let t be the historical average number of patients needed for the i-th department. Let j be the consultation time of the jth doctor at time t. The matching degree between the j-th doctor in the i-th department and the common diseases in the department. This is the departmental resource adjustment coefficient. This represents the total number of departments in the hospital. Let be the number of remaining appointment slots for the i-th department at time t. This represents the historical average number of patients needed for the k-th department at time t. This represents the adjustment coefficient for the remaining signal sources.

[0007] Furthermore, the expression for the medical atlas triple-link retrieval model in the medical atlas construction and retrieval module is as follows: ,in, The degree of triple link matching between patient p, doctor d, and clinic s. The weights for the historical medical visits of patient p and doctor d are given. Let be the correlation coefficient between the geographical locations of patient p and clinic s. Let be the matching coefficient between doctor d and clinic s in terms of outpatient schedules. The total number of medical records for patient p. Let be the total number of patients seen by doctor d. The historical number of times the clinic s has been used. For correlation compensation coefficient, This represents the total number of disease types. The degree of matching between patient p and doctor d for the treatment of disease type m. Let be the frequency coefficient of doctor d seeing patients with type m disease in clinic s.

[0008] Furthermore, the expression for the dynamic flexible allocation scheduling algorithm in the flexible allocation module is as follows: ,in, Let t represent the number of appointment slots available for the i-th department during time period t. This is the elasticity coefficient for deployment. The predicted number of patients visiting the i-th department during time period t is output by the traffic dynamic prediction module. for The number of remaining appointment slots in the i-th department during the time period. The utilization rate of clinic resources in the i-th department. This is the time-period adjustment coefficient. The rate of sudden outpatient visits in the i-th department during time period t. The outpatient rate of doctors in the i-th department. This represents the historical impact coefficient. For historical period numbers, The number of appointment slots available for the i-th department during time period h. The utilization rate of appointment slots for the i-th department during time period h.

[0009] Furthermore, the prediction model expression of the outpatient traffic dynamic prediction and analysis platform in the traffic dynamic prediction module is as follows: ,in, For the predicted outpatient flow of the i-th department during time period t, Historical traffic weighting coefficient for The actual outpatient flow of the i-th department during the time period The number of appointments made in the i-th department during time period t. Let be the appointment fulfillment rate of the i-th department during time period t. The estimated number of emergency room visits for the i-th department during time period t. The emergency-to-outpatient referral rate of the i-th department during time period t. Let t represent the number of doctors from the i-th department who made outpatient visits during time period t. The number of influencing factor types, Let be the impact value of factor q on the traffic volume of department i during time period t. Let be the weight coefficient of the q-th factor. This represents the adjustment coefficient for influencing factors, used to adjust the strength of the effect of various influencing factors on outpatient traffic prediction and balance the weights of various influencing factors in the model.

[0010] Furthermore, the expression for the appointment request processing model in the appointment interaction management module is as follows: ,in, Assign a priority score to patient p for the appointment request of department i during time period t. Let be the matching coefficient between patient p and the common diseases in department i. The remaining availability of appointment slots for the i-th department during time period t. The matching degree of patient p's visit time during time period t. Let be the total number of appointment requests for the i-th department during time period t. For patient p, the number of appointments made in history. This represents the historical influence coefficient. The number of departments the patient has visited in their history. For the history department serial number, For patient p, the historical satisfaction level during their visits to department r. Let represent the resource correlation between department r and department i during time period t.

[0011] Furthermore, the appointment allocation module includes an appointment demand analysis unit, a resource data integration unit, an allocation ratio calculation unit, and an allocation result verification unit. The appointment demand analysis unit receives basic patient information and historical medical records transmitted from the outpatient information collection and processing module, extracts data on patient symptom types, frequency of visits, and appointment preferences for each department, classifies patient groups according to urgency of symptoms and time requirements, and statistically analyzes the number and time distribution of appointment slots required by different categories of patients for each department. The resource data integration unit collects doctor's outpatient information, clinic resource data, and equipment usage status data to determine each doctor's capacity, clinic capacity, and equipment usage time restrictions, and matches doctor information with corresponding clinic information. The equipment information is linked and bound to form a resource association data table; the allocation ratio calculation unit calls the outpatient appointment dynamic balance allocation model, inputs the demand data output by the appointment demand analysis unit and the resource data output by the resource data integration unit, calculates the basic allocation ratio of appointments for each department at different times, and adjusts the basic allocation ratio by combining the doctor-patient matching information output by the medical atlas construction retrieval module; the allocation result verification unit compares the appointment allocation ratio of each department with the historical allocation data and the resource carrying capacity limit, and checks whether the allocation ratio complies with the outpatient operation specifications. If there is a situation that exceeds the resource carrying capacity limit or deviates too much from the historical reasonable allocation range, it is fed back to the allocation ratio calculation unit for recalculation.

[0012] Furthermore, the dynamic traffic prediction module includes a historical data preprocessing unit, an influencing factor extraction unit, a prediction model training unit, and a prediction result output unit. The historical data preprocessing unit receives historical outpatient traffic data and registration data transmitted by the outpatient information collection and processing module, removes outliers and missing values ​​from the data, and classifies and organizes the data by department and time period to form a standardized historical traffic dataset. The influencing factor extraction unit obtains holiday arrangements and seasonal climate change data from external data sources, extracts doctor outpatient schedule changes and disease prevalence trends from internal outpatient data, analyzes the correlation between various factors and outpatient traffic, and screens out factors that have a significant impact on traffic and quantifies them as influencing factors. The prediction model training unit uses the historical traffic dataset and the extracted influencing factors to train the prediction model of the outpatient traffic dynamic prediction and analysis platform, optimizes the prediction accuracy by adjusting the model parameters, and generates personalized prediction models suitable for each department. The prediction result output unit calls the trained personalized prediction model, inputs the influencing factor data for the current time period, calculates the predicted outpatient traffic value for each department in the future preset time period, and transmits the predicted value to the appointment resource elastic allocation module after classifying it by department and time period.

[0013] Furthermore, the flexible appointment allocation module includes an allocation time period division unit, an allocation quantity calculation unit, an allocation timing judgment unit, and an allocation instruction generation unit. The allocation time period division unit divides the day into multiple consecutive allocation time periods based on outpatient operating hours and peak patient visit patterns, determining the duration and start time of each time period to ensure that each time period covers patients' main appointment needs. The allocation quantity calculation unit receives the appointment allocation ratio output by the appointment balanced allocation module and the traffic prediction results output by the traffic dynamic prediction module, calls the dynamic flexible appointment allocation scheduling algorithm, calculates the number of appointments allocated to each department within each allocation time period, and simultaneously references... The system is corrected based on historical data on the number of appointments and their utilization rates. The appointment timing judgment unit monitors the appointment progress and remaining number of appointment slots in each department in real time, analyzes the appointment speed within the current appointment time, and determines whether the appointments for the next time period need to be released earlier or later. If the appointment speed is too fast, resulting in the early exhaustion of appointment slots, or too slow, resulting in a backlog of appointment slots, the appointment timing adjustment mechanism is triggered. The appointment instruction generation unit generates appointment release instructions, including department identifier, appointment time, and appointment quantity, based on the appointment quantity calculation results and the appointment timing judgment results. The instructions are then transmitted to the appointment interaction management module, and the generation time and parameter information of the appointment instructions are recorded.

[0014] A big data-based outpatient information query and appointment management method is proposed. This method, applied to a big data-based outpatient information query and appointment management system, includes the following steps: First, the outpatient information collection and processing module collects basic patient information, historical medical records, doctor's appointment information, and clinic resource data in real time. Unstructured data is converted into a unified format using structured processing. The processed data is then sent to the medical graph construction and retrieval module and the traffic dynamic prediction module, respectively. Second, the medical graph construction and retrieval module constructs a medical graph based on the received structured data, including patient-doctor-disease-clinic relationships. A triple-link retrieval logic is used to match patient demand information with doctor resource information. The matching results are then transmitted to the appointment allocation module and the appointment interaction management module, respectively. Third, the appointment allocation module calls the outpatient appointment dynamic allocation model, inputs the medical graph retrieval results and outpatient resource data, and calculates the allocation of appointment slots for each department at different times. The first step involves the allocation of appointment slots. The allocation ratio data is transmitted to the flexible appointment slot allocation module. The second step involves the dynamic traffic prediction module, based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease peak patterns, calculating the predicted outpatient traffic for each department within a preset time period using the outpatient traffic dynamic prediction analysis platform. This prediction result is then sent to the flexible appointment slot allocation module. The third step involves the flexible appointment slot allocation module calling a dynamic flexible appointment slot allocation scheduling algorithm. Combining the appointment allocation ratio and traffic prediction results, it determines the number of appointment slots allocated to each department within each allocation period and the timing of allocation, generating appointment slot allocation instructions and transmitting them to the appointment interaction management module. The fourth step involves the appointment interaction management module receiving medical atlas retrieval results and appointment slot allocation instructions. It provides patients with an appointment slot query interface for searching appointment information for each department, receives appointment requests submitted by patients, prioritizes them according to the appointment request processing logic, provides patients with appointment status information, and simultaneously feeds back patient appointment operation data to the outpatient information collection and processing module for data updates.

[0015] Beneficial Effects: This invention proposes a big data-based outpatient information query and appointment management system and method. The outpatient information collection and processing module integrates multi-dimensional data to form a unified data support system. The medical atlas construction and retrieval module enables precise matching of patient and doctor resources, solving the problems of scattered information and low matching degree in the traditional model. The appointment resource balancing allocation module combines a dynamic balancing allocation model to calculate the appointment resource allocation ratio based on data such as patient symptoms, doctor's expertise, and clinic status. The appointment resource elastic delivery module adjusts the delivery time and quantity through a scheduling algorithm that links real-time appointment progress and historical usage, replacing fixed ratios and manual adjustments, avoiding the coexistence of appointment shortages and idleness, and improving appointment resource utilization efficiency. The traffic dynamic prediction module builds a prediction platform by integrating factors such as holidays, seasonal disease patterns, and emergency-to-outpatient transfers, replacing single historical data estimations, improving the accuracy of traffic prediction, providing a reliable basis for appointment resource allocation and scheduling, and alleviating resource congestion during peak hours and resource waste during off-peak hours. The appointment interaction management module optimizes the appointment process, improves appointment convenience by combining patient needs and appointment availability, and achieves refined and intelligent outpatient management as a whole, reducing patient waiting time and improving the rationality of medical resource allocation and outpatient service efficiency. Attached Figure Description

[0016] Figure 1 This is a diagram showing the system module composition of the present invention; Figure 2 This is a flowchart of the method steps of the present invention. Detailed Implementation

[0017] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1 As shown, an outpatient information query and appointment management system based on big data includes: an outpatient information collection and processing module, a medical atlas construction and retrieval module, a balanced allocation module for appointment slots, a dynamic traffic prediction module, a flexible allocation module for appointment slots, and an appointment interaction management module. The outpatient information collection and processing module collects and structures basic information of outpatients, historical medical records, doctor's outpatient information, and clinic resource data in real time, and transmits the processed data to the medical atlas construction and retrieval module. Specifically, the outpatient information collection and processing module collects outpatient basic information (including age, gender, chief complaint, past medical history, etc.), historical medical records (departments visited, doctors attending, treatment plans, follow-up visits, etc. within the past 1-3 years), doctor's outpatient information (outpatient dates, outpatient hours, patient capacity limits, types of diseases specializing, etc.), and clinic resource data (clinic number, number of patients accommodated, type of medical equipment equipped, daily usage time limits, etc.) in real time. The collection frequency is set to once every 5 minutes to ensure data real-time performance. After collection, unstructured text data (such as patient complaints) is converted into standardized field formats through structured processing. The patient age field ranges from 0 to 120 years old. Doctors' consultation limits are set according to their professional titles (25-30 patients per day for chief physicians, 20-25 for associate chief physicians, and 15-20 for attending physicians). The number of patients that can be accommodated in a consultation room is set as 1 person for a single consultation room and 3-5 people for a multi-person consultation room. Equipment usage time restrictions are divided according to equipment type (e.g., electrocardiographs can be used from 8:00 to 17:00 daily, with a maximum of 8 patients served per hour). The processed structured data is stored in a database with a capacity of no less than 10TB. At the same time, abnormal data (such as records of people over 120 years old or overlapping outpatient hours) are removed through a data verification mechanism (the verification accuracy rate must reach more than 99.5%). Finally, qualified data is transmitted to the medical atlas construction and retrieval module according to the module requirements, providing accurate and real-time data support for subsequent atlas construction and retrieval. The implementation of this module can ensure the integrity and standardization of various core data in outpatient clinics and avoid the failure of subsequent module functions due to data chaos.

[0019] The medical atlas construction and retrieval module constructs a medical atlas based on the received structured data, including the relationships between patients, doctors, symptoms, and clinics. It retrieves information on matching patient needs with doctor resources through a triple-link retrieval logic, and transmits the retrieval results to the appointment allocation module and the appointment interaction management module, respectively. Specifically, the medical atlas construction and retrieval module first receives structured data transmitted from the outpatient information collection and processing module. It then constructs an association atlas with patients, doctors, symptoms, and clinics as core nodes. Patient nodes include attributes such as patient ID, symptom type, and frequency of visits; doctor nodes include attributes such as doctor ID, areas of expertise, and consultation hours; symptom nodes include attributes such as symptom code, symptom level (divided into 1-5 levels based on urgency, with level 1 being the most urgent), and common treatment departments; and clinic nodes include attributes such as clinic ID, corresponding doctor, and equipment configuration. The relationships between nodes are generated through data matching (e.g., the "having" relationship between patients and symptoms, and the "using" relationship between doctors and clinics). The data association accuracy during the atlas construction process must reach over 99%. After the system is built, a three-tiered search logic is initiated. The first tier matches the patient's symptoms with the doctor's areas of expertise (with a matching threshold of 80% or higher). The second tier matches the patient's appointment time requirements with the doctor's appointment hours (with a time overlap of 60% or higher). The third tier matches the patient's preferred appointment location with the clinic's geographical location (distance deviation not exceeding 3 kilometers). The search response time is controlled within 1-3 seconds. The search results include a list of matched doctors (sorted from highest to lowest matching score, displaying a maximum of 10 doctors), corresponding clinic information, and remaining appointment slots. The search results are then transmitted to the appointment allocation module (for optimizing appointment allocation) and the appointment interaction management module (for patient appointment selection). This module, through multi-dimensional correlation retrieval, can significantly improve the accuracy of matching patients with medical resources and reduce invalid appointments.

[0020] The appointment allocation module calls the outpatient appointment dynamic allocation model, combines the medical atlas search results with outpatient resource data, calculates the appointment allocation ratio for each department, and transmits the allocation results to the appointment flexible delivery module. Specifically, when the appointment allocation module is implemented, it first receives doctor-patient matching information transmitted by the medical atlas construction and retrieval module. At the same time, it obtains patient demand data (statistically divided into time periods, each time period is 2 hours, such as 8:00-10:00, 10:00-12:00, etc.), doctor resource data (number of doctors in each department and their capacity to receive patients) and clinic resource data (number of clinics in each department and the number of patients they can accommodate) from the outpatient information collection and processing module. Among them, the patient demand data is classified according to the urgency of the disease (30%-40% of appointments should be allocated to level 1 diseases, 25%-30% to level 2 diseases, and 30%-45% to level 3-5 diseases). The doctor's capacity to receive patients is distinguished by professional title (chief physicians see 5-6 patients per time period, associate chief physicians see 4-5 patients, and attending physicians see 3-4 patients). The number of patients that can be accommodated in a clinic determines the maximum number of patients that can be received per time period (5-6 patients per time period for a single clinic and 12-15 patients per time period for a multi-person clinic). When the module calls the outpatient appointment dynamic equilibrium allocation model, it inputs the above parameters to calculate the basic allocation ratio of appointments for each department in each time period. Then, it adjusts the ratio based on the matching degree between doctors and patients in the matching information (for every 10% increase in matching degree, the corresponding doctor's appointment allocation increases by 5%-8%). The allocation result must meet the following requirements: the total number of appointments in each department does not exceed the minimum of the department's total doctor capacity and clinic capacity, and the fluctuation range of appointment allocation in each time period does not exceed 20%. After the calculation is completed, the allocation result is transmitted to the appointment elastic allocation module. This module, through scientific allocation of appointments, can avoid the situation of some departments having an oversupply of appointments and others having a shortage of appointments, thereby improving the overall utilization efficiency of appointments.

[0021] The outpatient traffic dynamic prediction module builds an outpatient traffic dynamic prediction and analysis platform based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease high incidence patterns, and outputs the outpatient traffic prediction results of each department within the future preset time period to the appointment resource elastic allocation module. Specifically, the dynamic traffic prediction module integrates multiple types of data during implementation, including historical outpatient traffic data (the number of patients visiting each department at each time period every day for the past 1-2 years, with a data sample size of no less than 100,000 records), real-time registration data (the number of patients registered at each department at each time period on the current day, updated every 30 minutes), holiday factors (statutory holidays are assigned an impact coefficient according to type, such as 0.6-0.7 for Spring Festival, 0.7-0.8 for National Day, and 1.0 for non-holidays), seasonal disease incidence patterns (such as respiratory diseases being prevalent in spring, with a corresponding departmental traffic coefficient of 1.2-1.4; cardiovascular diseases being prevalent in winter, with a coefficient of 1.3-1.5), and also includes emergency-to-outpatient transfer data (the proportion of emergency-to-outpatient transfers to the total number of outpatients at each time period every day, usually 5%-10%). The modular outpatient traffic dynamic prediction and analysis platform adopts a hierarchical prediction logic. First, prediction units are divided by department, and then the prediction dimensions are refined by hourly time periods. During the prediction process, a basic traffic model is first established based on historical data, and then corrected by incorporating real-time registration data, holiday coefficients, seasonal coefficients, and the emergency-to-outpatient transfer ratio. The prediction error must be controlled within 15%. The prediction results include the predicted number of patients in each department for each time period within the next 24 hours (accurate to the single digit), peak periods (times when the number of patients exceeds 120% of the department's average traffic), and off-peak periods. After the prediction is completed, the results are transmitted to the flexible appointment allocation module, providing data for adjusting the timing and quantity of appointment allocation. This module enables accurate prediction of outpatient traffic, helping to allocate resources in advance to cope with traffic fluctuations.

[0022] The flexible allocation module calls the dynamic flexible allocation scheduling algorithm to determine the number of available slots and the timing of allocation for each time period based on the allocation ratio and traffic prediction results. The allocation instructions are then transmitted to the reservation interaction management module. Specifically, when the flexible appointment allocation module is implemented, it first receives the appointment allocation ratios for each department and time period from the balanced appointment allocation module, and the traffic prediction results for each department (including peak and off-peak time period identifiers) output by the dynamic traffic prediction module. When the module calls the dynamic flexible appointment allocation scheduling algorithm, it needs to set the allocation parameters (the allocation interval is once every 30 minutes, and the allocation amount each time does not exceed 30% of the total allocation for that time period; during peak hours, the allocation interval is shortened to once every 20 minutes, and the allocation amount is increased to 40%-50% of the total allocation for that time period; during off-peak hours, the allocation interval is extended to once every 60 minutes, and the allocation amount is reduced to 20%-25%). Before allocation, it is necessary to monitor the number of people who have made appointments in each department in real time (monitoring frequency is once every 10 minutes). If the number of people who have made appointments in a certain time period reaches 70% of the total allocation for that time period, the additional allocation mechanism is triggered (the additional allocation amount is 50% of the remaining allocation for that time period); if the number of people who have made appointments is less than 30% of the total allocation for that time period, the next allocation is delayed (the delay time is 30 minutes). The allocation instruction includes information such as department identifier, allocation time period, allocation quantity, and allocation start time. Once generated, it is immediately transmitted to the appointment interaction management module, and an allocation log is recorded (including allocation time, allocation quantity, subsequent appointment volume, etc., with logs retained for at least one year). This module, through a flexible allocation strategy, ensures sufficient supply of appointment slots during peak demand periods and avoids waste during off-peak periods, thereby improving the efficiency of appointment slot utilization.

[0023] The appointment interaction management module receives medical atlas search results and appointment allocation instructions, providing patients with functions such as appointment query, appointment application submission, and appointment status feedback. At the same time, it feeds back the patient's appointment operation data to the outpatient information collection and processing module.

[0024] Specifically, during the implementation of the appointment interaction management module, it simultaneously receives matching results (including doctor list, clinic information, and remaining appointment slots) transmitted by the medical atlas construction and retrieval module and placement instructions (including placement time slots and placement quantities) transmitted by the flexible placement module. The module first integrates this information into standardized appointment interface data (the interface response time must be controlled within 2 seconds, and the data update frequency must be consistent with the placement interval). When patients access the system through terminals (mobile APP, web page, self-service machine, etc.), they can query the doctors' outpatient information for each department, the number of remaining appointment slots (accurate to the single digit), and select the consultation time slot (divided into 15-minute time units, such as 8:00-8:15, 8:15-8:30, etc.). When a patient submits an appointment request, the module needs to verify the patient's identity information (match rate must reach 100%, supporting verification methods such as ID card, medical insurance card, and mobile phone number), and check for appointment time conflicts (if the patient has already booked another department at the same time, a conflict notification should be displayed and adjustments suggested). After successful verification, an appointment number is assigned to the patient (sorted by application time, the number includes department code, date, and time slot code), and the appointment status is provided in real time (successful, failed, pending confirmation). Simultaneously, the module needs to feed back the patient's appointment operation data (including the booked department, doctor, time slot, and cancellation records) to the outpatient information collection and processing module every 15 minutes for subsequent data updates and analysis. By standardizing the appointment process and providing real-time feedback, this module can improve the patient's appointment experience and reduce appointment operation errors and conflicts.

[0025] Preferably, the expression for the dynamic balanced allocation model of outpatient appointment slots in the appointment slot allocation module is: ,in, Let t be the number of appointment slots allocated to the j-th doctor in the i-th department at time t. Assign weight coefficients to the number sources. Let t be the historical average number of patients needed for the i-th department. Let j be the consultation time of the jth doctor at time t. The matching degree between the j-th doctor in the i-th department and the common diseases in the department. This is the departmental resource adjustment coefficient. This represents the total number of departments in the hospital. Let be the number of remaining appointment slots for the i-th department at time t. This represents the historical average number of patients needed for the k-th department at time t. This represents the adjustment coefficient for the remaining signal sources.

[0026] Specifically, when implementing the dynamic balanced allocation model for outpatient appointment slots, the specific numerical ranges and basis for the values ​​of various technical parameters are first determined. The weighting coefficient for appointment allocation is set between 0.6 and 0.9, adjusted according to the overall strain on outpatient resources in the hospital. The more strained the resources, the closer the coefficient is to 0.9, prioritizing the allocation of appointment slots to core departments. The departmental resource adjustment coefficient is set between 0.3 and 0.5, mainly determined based on resource differences between departments. Departments with sufficient resources have a lower coefficient, while departments with scarce resources have a higher coefficient. The historical average number of patients required requires statistical analysis of the average number of patients in the corresponding department at the same time each day over the past three months, ensuring a sufficient data sample size that covers weekdays, weekends, and holidays. Doctors' consultation hours are calculated based on actual consultation schedules, in hours, accurate to 0.5 hours. For example, if a doctor's morning consultation hours are 3.5 hours, this is directly substituted into the calculation. The matching degree between doctors and common diseases in departments is determined by the percentage of patients with the corresponding diseases seen by the doctor in the past six months, with a value ranging from 0 to 1. The higher the percentage, the closer the matching degree is to 1. The remaining number of appointment slots is based on real-time statistical data, updated every 10 minutes to ensure that the latest data is used in the calculation. During implementation, the above parameters are first collected and their accuracy is verified, and then they are substituted into the model to calculate the number of appointment slots allocated to each doctor in each department. The calculation results must not exceed the doctor's consultation limit and the clinic's capacity, and the difference in the number of appointment slots allocated to different doctors in the same department should not exceed 30%. This model can realize the dynamic allocation of appointment slots according to demand and actual resource conditions, avoid the imbalance of appointment slot allocation, and improve the rationality of appointment slot utilization.

[0027] Preferably, the expression for the medical atlas triple-link retrieval model in the medical atlas construction and retrieval module is as follows: ,in, The degree of triple link matching between patient p, doctor d, and clinic s. The weights for the historical medical visits of patient p and doctor d are given. Let be the correlation coefficient between the geographical locations of patient p and clinic s. Let be the matching coefficient between doctor d and clinic s in terms of outpatient schedules. The total number of medical records for patient p. Let be the total number of patients seen by doctor d. The historical number of times the clinic s has been used. For correlation compensation coefficient, This represents the total number of disease types. The degree of matching between patient p and doctor d for the treatment of disease type m. Let be the frequency coefficient of doctor d seeing patients with type m disease in clinic s.

[0028] Specifically, in implementing the medical atlas triple-link retrieval model, various parameters are precisely set based on actual outpatient data. The historical visit association weight between patients and doctors is determined based on the number of visits they have made in the past; the more visits, the higher the weight, with a maximum value not exceeding 1.0. The geographical location association coefficient between patients and clinics is calculated based on the straight-line distance between the patient's residential address and the clinic's location; the closer the distance, the closer the coefficient is to 1.0, and the coefficient drops below 0.3 when the distance exceeds 10 kilometers. The doctor's outpatient schedule matching coefficient is determined based on the frequency and duration of the doctor's outpatient visits to that clinic; the more outpatient visits per week and the longer the duration of each outpatient visit, the higher the coefficient, with a maximum of 1.0. The total number of patient visit records, the total number of patients seen by doctors, and the historical usage of clinics are all based on cumulative data from the past year to ensure that the data reflects long-term associations. The association compensation coefficient is set between 0.1 and 0.2 to compensate for matching biases in some special cases (such as when a patient's first visit has no historical association). The total number of disease types is determined according to the International Classification of Diseases, including more than 200 common outpatient disease types. The matching degree between doctors and patients for a certain type of disease is calculated based on the doctor's cure rate for that disease and patient satisfaction, with a value ranging from 0 to 1. The frequency coefficient of a doctor treating a certain disease in a certain clinic is calculated as the percentage of such treatments over the past 3 months; the higher the frequency, the closer the coefficient is to 1.0. During implementation, the above parameters are first integrated to build a complete associated data system, and then a triple-link retrieval logic is initiated to ensure accurate calculation of the retrieval matching degree and a response time controlled within 1-3 seconds. This model enables efficient association and matching among patients, doctors, and clinics, reducing invalid search results.

[0029] Preferably, the expression for the dynamic flexible allocation scheduling algorithm in the flexible allocation module is: ,in, Let t represent the number of appointment slots available for the i-th department during time period t. This is the elasticity coefficient for deployment. The predicted number of patients visiting the i-th department during time period t is output by the traffic dynamic prediction module. for The number of remaining appointment slots in the i-th department during the time period. The utilization rate of clinic resources in the i-th department. This is the time-period adjustment coefficient. The rate of sudden outpatient visits in the i-th department during time period t. The outpatient rate of doctors in the i-th department. This represents the historical impact coefficient. For historical period numbers, The number of appointment slots available for the i-th department during time period h. The utilization rate of appointment slots for the i-th department during time period h.

[0030] Specifically, when implementing the dynamic appointment scheduling algorithm, the elasticity coefficient is set between 0.7 and 0.95, adjusted according to the overall appointment popularity of the outpatient department. When appointment popularity is high (such as peak seasons or peak months), the coefficient is close to 0.95 to increase the number of appointments; when popularity is low, the coefficient is lowered. The time-period adjustment coefficient is set between 0.4 and 0.6, determined based on the differences in patient demand at different times. The coefficient is close to 0.6 for the morning peak (8:00-10:00) and afternoon peak (14:00-16:00), and close to 0.4 for other times. The predicted number of patients is provided by the dynamic traffic prediction module, requiring accuracy to the single digit, with a prediction error controlled within 15%. The remaining appointment slots for the previous time period are based on real-time statistics and are updated 10 minutes before calculating the current time period's allocation. The department / clinic resource utilization rate is calculated as the ratio of the actual number of clinics in use to the total number of clinics in the current time period, with a value between 0 and 1; a higher utilization rate indicates more strained resources. The emergency visit rate is set at 0-0.3, adjusted based on recent public health emergencies and seasonal disease outbreaks; in the absence of emergencies, it is set at 0.05-0.1. The doctor's outpatient rate is calculated as the ratio of actual outpatient visits to planned outpatient visits, with a value of 0-1. If the outpatient rate is below 80%, the allocation should be appropriately reduced. The historical allocation impact coefficient is set at 0.05-0.15, used to reference the allocation effect of the past three identical time periods; the coefficient is appropriately increased when the allocation effect is good (utilization rate exceeds 80%). The historical time period number covers the same time period within the past seven days to ensure the historical data has reference value. The historical time period appointment utilization rate is calculated as the ratio of the number of appointments used in that time period to the total allocation, with a value of 0-1. During implementation, all parameters are collected and updated in real time, then substituted into the algorithm to calculate the number of appointments allocated to each department in each time period. The calculation result must be rounded to an integer, and the single allocation amount should not exceed 50% of the total allocated amount for that time period. This algorithm enables flexible allocation of appointments according to actual demand, avoiding appointment backlogs or shortages.

[0031] Preferably, the prediction model expression of the outpatient traffic dynamic prediction and analysis platform in the traffic dynamic prediction module is as follows: ,in, For the predicted outpatient flow of the i-th department during time period t, Historical traffic weighting coefficient for The actual outpatient flow of the i-th department during the time period The number of appointments made in the i-th department during time period t. Let be the appointment fulfillment rate of the i-th department during time period t. The estimated number of emergency room visits for the i-th department during time period t. The emergency-to-outpatient referral rate of the i-th department during time period t. Let t represent the number of doctors from the i-th department who made outpatient visits during time period t. The number of influencing factor types, Let be the impact value of factor q on the traffic volume of department i during time period t. Let be the weight coefficient of the q-th factor. This represents the adjustment coefficient for influencing factors, used to adjust the strength of the effect of various influencing factors on outpatient traffic prediction and balance the weights of various influencing factors in the model.

[0032] Specifically, when implementing the outpatient traffic dynamic prediction and analysis platform's prediction model, the historical traffic weight coefficient is set between 0.5 and 0.8. This coefficient is adjusted based on the correlation between historical and current traffic; a strong correlation (e.g., no special events) results in a coefficient close to 0.8, while a weak correlation (e.g., policy adjustments, new hospital departments) results in a lower coefficient. The actual outpatient traffic for the previous period is the precise data collected within 5 minutes after the end of that period, ensuring data accuracy. The number of appointments made in the current period is real-time statistical data, updated every 30 minutes. The appointment fulfillment rate is the percentage of people who actually received treatment after making an appointment in the past month, ranging from 0 to 1. Prediction results need to be adjusted if the fulfillment rate is below 70%. The estimated number of emergency room visits is predicted based on recent emergency room visit trends, accurate to the single digit. The emergency-to-outpatient conversion rate is the percentage of emergency patients who were converted to outpatient visits in the past month, ranging from 0 to 1, typically maintained between 0.05 and 0.1. The number of doctors on duty during the current period is the actual number of doctors present, and this number is immediately substituted into the calculation after the statistics are completed. The number of influencing factors includes 8-10 common factors such as holidays, seasons, weather, and policy adjustments. The impact value of a certain factor on the department's traffic flow is determined based on the actual impact of the factor on the traffic flow when it has occurred multiple times in the past, with a value of -0.5 to 1.5. Positive impacts (such as high incidence of seasonal diseases) are represented by positive values, and negative impacts (such as severe weather) are represented by negative values. The factor weight coefficient is determined based on the significance of the impact of each factor on the traffic flow, with a value of 0 to 1. The more significant the impact, the higher the coefficient. During implementation, all parameters are collected and verified first, and then the predicted traffic flow for each department and each time period is calculated layer by layer according to the model logic. The prediction results need to be accurate to the single digit, and the error with the subsequent actual traffic flow should be controlled within 15%. This model can achieve accurate prediction of outpatient traffic flow and provide a reliable basis for resource allocation.

[0033] Preferably, the appointment request processing model expression in the appointment interaction management module is: ,in, Assign a priority score to patient p for the appointment request of department i during time period t. Let be the matching coefficient between patient p and the common diseases in department i. The remaining availability of appointment slots for the i-th department during time period t. The matching degree of patient p's visit time during time period t. Let be the total number of appointment requests for the i-th department during time period t. For patient p, the number of appointments made in history. This represents the historical influence coefficient. The number of departments the patient has visited in their history. For the history department serial number, For patient p, the historical satisfaction level during their visits to department r. Let represent the resource correlation between department r and department i during time period t.

[0034] Specifically, when implementing the appointment request processing model, the matching coefficient between patients and common diseases in departments is calculated based on the degree of fit between the patient's chief complaint and the department's scope of treatment, ranging from 0 to 1. A higher degree of fit results in a coefficient closer to 1.0; for example, the matching coefficient between a patient's chief complaint of respiratory symptoms and the respiratory department can reach over 0.9. The sufficiency of remaining appointment slots in the department during the current time period is calculated as the ratio of the number of remaining slots to the average demand during that time period, ranging from 0 to 1. A larger number of remaining slots indicates higher sufficiency, with a coefficient of 1.0 when demand exceeds the average. The patient's appointment time matching degree is determined based on the overlap between the patient's selected appointment time and the doctor's appointment time. A coefficient of 1.0 is obtained when there is complete overlap, calculated as a percentage of overlap duration when there is partial overlap, and 0 when there is no overlap. The total number of appointment requests in the department during the current time period is real-time statistical data, updated every 5 minutes to ensure it reflects the latest appointment popularity. The number of historical appointments for patients is calculated based on the cumulative number of appointments made within the past year; a higher number indicates greater familiarity with the appointment service. The historical record influence coefficient is set between 0.1 and 0.2 to balance the appointment priorities of new and old patients and avoid excessive bias. The number of different departments a patient has visited in the past two years is counted; a higher number indicates more complex patient needs. Patient satisfaction with past departments is calculated using post-visit evaluation scores, ranging from 0 to 1; higher satisfaction has a greater positive impact on current appointment priority. The resource correlation between past and current departments in the current time period is determined by whether there is a collaborative relationship between the two departments (e.g., referral collaboration between internal medicine and surgery). When there is a collaborative relationship, the coefficient is close to 0.8; when there is no collaborative relationship, the coefficient is close to 0.2. During implementation, relevant parameters of patients and departments are first collected, then substituted into the model to calculate the appointment request priority score. A higher score indicates a higher appointment success rate, and for patients with the same score, they are ranked according to the application time. This model can achieve reasonable ranking of appointment requests, improving the fairness and efficiency of appointment services.

[0035] Preferably, the appointment allocation module includes an appointment demand analysis unit, a resource data integration unit, an allocation ratio calculation unit, and an allocation result verification unit. The appointment demand analysis unit receives basic patient information and historical medical records transmitted by the outpatient information collection and processing module, extracts data on patient symptom types, frequency of visits, and appointment preferences for each department, classifies patient groups according to the urgency of symptoms and time requirements, and statistically analyzes the demand for appointments in each department and the distribution of demand time periods for different categories of patients. The resource data integration unit collects doctor's outpatient information, clinic resource data, and equipment usage status data to determine each doctor's capacity, clinic capacity, and equipment usage time limits, and matches doctor information with corresponding clinic information. The equipment information is linked and bound to form a resource association data table; the allocation ratio calculation unit calls the outpatient appointment dynamic balance allocation model, inputs the demand data output by the appointment demand analysis unit and the resource data output by the resource data integration unit, calculates the basic allocation ratio of appointments for each department at different times, and adjusts the basic allocation ratio by combining the doctor-patient matching information output by the medical atlas construction retrieval module; the allocation result verification unit compares the appointment allocation ratio of each department with the historical allocation data and the resource carrying capacity limit, and checks whether the allocation ratio complies with the outpatient operation specifications. If there is a situation that exceeds the resource carrying capacity limit or deviates too much from the historical reasonable allocation range, it is fed back to the allocation ratio calculation unit for recalculation.

[0036] Specifically, the appointment allocation module includes four units. The appointment demand analysis unit receives basic patient information and historical medical records transmitted by the outpatient information collection and processing module. It extracts patient disease types (classified according to the International Classification of Diseases, including more than 200 common diseases), frequency of visits (statistics of the average interval between visits in the past 3 months, accurate to the day), and appointment preference data (statistics by time period, such as the percentage of appointments for time periods like 8:00-10:00 and 10:00-12:00). It also classifies patient groups according to the urgency of their conditions (levels 1-5, with level 1 being the most urgent) and their time requirements (weekdays / weekends, mornings / afternoons). It then calculates the number of appointments needed by different categories of patients for each department (accurate to the single digit) and the distribution of demand time periods (every 2 hours is a statistical period). After classification, the data error must be controlled within 5%. The resource data integration unit collects doctor outpatient information (outpatient date, time slot, and patient capacity; 25-30 patients per day for chief physicians, 20-25 patients per day for associate chief physicians), clinic resource data (number, capacity; 1 person per single-patient clinic, 3-5 people per multi-patient clinic), and equipment usage status data (time slot restrictions, e.g., ECG machine available from 8:00 AM to 5:00 PM). This determines each doctor's patient capacity, clinic capacity, and equipment usage restrictions. Doctors are then linked to their corresponding clinics and equipment to form a resource association data table, with a 100% accuracy rate. The allocation ratio calculation unit calls the outpatient appointment dynamic equilibrium allocation model, inputs demand and resource data, and calculates the basic allocation ratio for each department at different times (accurate to two decimal places). This ratio is then adjusted based on doctor-patient matching information retrieved from the medical atlas (priority is given to matching ratios above 80%), with the adjustment not exceeding 20% ​​of the basic ratio. The allocation result verification unit compares the allocation ratio with historical allocation data (data from the same period in the past month) and the resource carrying capacity limit (the minimum of the total capacity of doctors to receive patients and the total capacity of clinics). If the deviation exceeds 15% or exceeds the carrying capacity limit, it is fed back to the calculation unit for recalculation. The verification pass rate must reach 100%. Through the collaboration of each unit, the allocation of appointment slots is ensured to be scientific and reasonable.

[0037] Preferably, the dynamic traffic prediction module includes a historical data preprocessing unit, an influencing factor extraction unit, a prediction model training unit, and a prediction result output unit. The historical data preprocessing unit receives historical outpatient traffic data and registration data transmitted by the outpatient information collection and processing module, removes outliers and missing values ​​from the data, and classifies and organizes the data by department and time period to form a standardized historical traffic dataset. The influencing factor extraction unit obtains holiday arrangements and seasonal climate change data from external data sources, extracts doctor outpatient schedule changes and disease prevalence trends from internal outpatient data, analyzes the correlation between various factors and outpatient traffic, and screens out factors that have a significant impact on traffic and quantifies them as influencing factors. The prediction model training unit uses the historical traffic dataset and the extracted influencing factors to train the prediction model of the outpatient traffic dynamic prediction and analysis platform, optimizes the prediction accuracy by adjusting the model parameters, and generates personalized prediction models suitable for each department. The prediction result output unit calls the trained personalized prediction model, inputs the influencing factor data for the current time period, calculates the predicted outpatient traffic value for each department in the future preset time period, and transmits the predicted value to the appointment resource elastic allocation module after classifying it by department and time period.

[0038] Specifically, the traffic flow dynamic prediction module includes four units. The historical data preprocessing unit receives historical outpatient traffic data (the number of patients visiting each department at each time period for the past 1-2 years, with a sample size of no less than 100,000 records) and registration data transmitted from the outpatient information collection and processing module. The data is processed through an outlier removal algorithm (removing data that exceeds three times the standard deviation of the average) and a missing value completion algorithm (using the mean of adjacent time periods to complete the missing value, with a completion error of no more than 10%). The data is then classified and organized by department (such as internal medicine, surgery, pediatrics, etc.) and time period (each hour is a time period) to form a standardized historical traffic dataset. The data integrity must reach more than 99%. The influencing factor extraction unit obtains holiday schedules (statutory holidays labeled by type, such as Spring Festival and National Day) and seasonal climate change data (temperature and humidity, accurate to ℃ and %) from external data sources. It also extracts doctor outpatient schedule changes (temporary closures and additional outpatient records) and disease prevalence trend data (weekly change rate of patients seeking treatment for a specific disease) from internal outpatient data. Influencing factors are screened using a correlation analysis algorithm (Pearson correlation coefficient, |r|≥0.3 is considered significant) and quantified into influencing factors (values ​​0-2, 1-2 for positive influence and 0-1 for negative influence). The factor quantification accuracy must reach over 90%. The prediction model training unit uses historical traffic datasets and influencing factors, employing a gradient descent algorithm to train a dynamic prediction model for outpatient traffic. The learning rate (0.01-0.1) and number of iterations (1000-5000 times) are adjusted to optimize prediction accuracy. After model training, the test set error must be controlled within 15%, generating personalized prediction models for each department (such as internal medicine models, pediatric models, etc.). The prediction result output unit calls the personalized prediction model, inputs the influencing factor data of the current time period, calculates the predicted number of patients in each department for each time period in the next 24 hours (accurate to the single digit), and transmits it to the flexible appointment allocation module by department and time period. The data transmission delay does not exceed 10 seconds. Through the collaboration of each unit, accurate prediction of outpatient traffic is achieved, providing a basis for resource scheduling.

[0039] Preferably, the flexible appointment allocation module includes an allocation time period division unit, an allocation quantity calculation unit, an allocation timing judgment unit, and an allocation instruction generation unit. The allocation time period division unit divides a day into multiple consecutive allocation time periods based on outpatient operating hours and peak patient visit patterns, determining the duration and start time of each time period to ensure that each time period covers patients' main appointment needs. The allocation quantity calculation unit receives the appointment allocation ratio output by the appointment equalization allocation module and the traffic prediction result output by the traffic dynamic prediction module, calls the dynamic flexible appointment allocation scheduling algorithm, calculates the number of appointments allocated to each department within each allocation time period, and simultaneously references... The system is corrected based on historical data on the number of appointments and their utilization rates. The appointment timing judgment unit monitors the appointment progress and remaining number of appointment slots in each department in real time, analyzes the appointment speed within the current appointment time, and determines whether the appointments for the next time period need to be released earlier or later. If the appointment speed is too fast, resulting in the early exhaustion of appointment slots, or too slow, resulting in a backlog of appointment slots, the appointment timing adjustment mechanism is triggered. The appointment instruction generation unit generates appointment release instructions, including department identifier, appointment time, and appointment quantity, based on the appointment quantity calculation results and the appointment timing judgment results. The instructions are then transmitted to the appointment interaction management module, and the generation time and parameter information of the appointment instructions are recorded.

[0040] Specifically, the flexible appointment allocation module includes four units. The allocation time slot division unit divides the day into multiple consecutive allocation time slots based on the outpatient operating hours (e.g., 8:00-17:00) and the peak patient visit patterns (statistics on the number of patients seen per hour over the past 3 months, with peak hours being defined as peak periods). During peak periods, each time slot is divided into 20-minute intervals, and during off-peak periods, each time slot is divided into 60-minute intervals. The duration of each time slot (20 minutes, 30 minutes, or 60 minutes) and the start time (e.g., 8:00, 8:20, 8:40, etc.) are determined to ensure coverage of patients' main appointment needs (covering more than 95% of patients' appointment time slots). After the time slots are divided, they need to be confirmed by the outpatient operations department, and the confirmation pass rate must reach 100%. The allocation calculation unit receives the allocation ratio of appointment slots (the allocation percentage of each department and time period, accurate to two decimal places) output by the appointment slot equalization module and the traffic prediction result (the predicted number of patients in each department and time period) output by the traffic dynamic prediction module. It calls the dynamic appointment slot elastic allocation scheduling algorithm to calculate the number of appointment slots allocated to each department in each allocation period (accurate to the single digit). At the same time, it refers to the historical allocation quantity (the allocation quantity of the same period on the same date in the past 3 days) and the utilization rate data (the utilization rate of the same period in the past 3 days, utilization rate = used appointment slots / total allocation) for correction. The correction range shall not exceed 20% of the initial calculated value, and the allocation quantity shall not exceed 50% of the total allocation quantity in that period. The timing of appointment allocation is determined by the real-time monitoring of appointment progress in each department (statistics of the number of people who have made appointments are compiled every 10 minutes) and the number of remaining appointment slots (updated in real time with a delay of no more than 5 seconds). The unit analyzes the appointment allocation speed for the current time slot (number of people who have made appointments / time slot duration, unit: people / minute). If the appointment allocation speed exceeds 1.5 times the historical average speed (determined as too fast), the early allocation mechanism is triggered (appointments for the next time slot are allocated 10-20 minutes earlier); if the appointment allocation speed is less than 0.5 times the historical average speed (determined as too slow), the delayed allocation mechanism is triggered (appointments for the next time slot are allocated 30-60 minutes later). The accuracy rate of the determination must be above 90%. The appointment allocation instruction generation unit generates an appointment allocation instruction based on the calculation results of the allocation quantity and the judgment results of the allocation timing. The instruction includes the department identifier (e.g., internal medicine 01, surgery 02), the allocation time period (e.g., 8:00-8:20), the allocation quantity (e.g., 20), and the allocation start time (e.g., 8:00). The instruction is then transmitted to the appointment interaction management module (the transmission success rate must reach 100%). At the same time, the allocation log is recorded (including allocation time, allocation quantity, subsequent appointment quantity, etc.). The log is kept for no less than 1 year. Through the collaboration of various units, the flexible allocation of appointment resources is realized, thereby improving the efficiency of appointment resource utilization.

[0041] The outpatient appointment dynamic equilibrium allocation model is a core tool for the scientific allocation of outpatient appointments. Specifically, it's a mathematical model that combines multi-dimensional data such as patient demand, doctor resources, and clinic conditions to calculate the number of appointments allocated to each department and doctor at different times. In implementation, the appointment demand analysis unit first extracts patient disease types (classified according to the International Classification of Diseases, over 200 types), visit frequency (average interval over the past 3 months, accurate to the day), and appointment preferences (statistical percentage by time period) for each department, and categorizes demand by disease urgency (levels 1-5). Next, the resource data integration unit collects doctor's outpatient information (chief physician sees 25-30 patients per day), clinic capacity (1 person per clinic), and equipment usage restrictions, forming a resource association table. Then, the allocation ratio calculation unit inputs the demand and resource data, calls the model to calculate the basic allocation ratio (accurate to two decimal places), and combines it with doctor-patient matching information from the medical atlas (adjustments are prioritized for matching rates above 80%), with adjustments not exceeding 20%. Finally, the allocation result verification unit compares the historical data (same period over the past month) with the resource limit; if the deviation exceeds 15%, it recalculates. Its purpose is to avoid the coexistence of shortages and underutilization of appointment slots, ensuring that slots are allocated according to demand and actual resource availability. Its significance lies in improving the rational use of appointment slots, laying the foundation for subsequent flexible allocation, and reducing the waste of medical resources.

[0042] The Medical Graph Triple Link Retrieval Model is a tool for constructing a patient-doctor-disease-clinic association graph and achieving accurate retrieval. Specifically, it calculates the matching degree among patients, doctors, and clinics through triple association logic, outputting the optimal matching result. In implementation, it first constructs a graph based on structured data (patient basic information, doctor's outpatient records, etc.) from the outpatient information collection and processing module, using patients, doctors, diseases, and clinics as core nodes. Node attributes include patient ID and disease, doctor's specialty diseases and outpatient hours, etc. Associations are generated through data matching (e.g., the patient "has" a disease). The graph construction accuracy is over 99%. Then, a triple retrieval is initiated: the first matching is between the patient's disease and the doctor's specialty diseases (threshold over 80%); the second matching is between the patient's time needs and the doctor's outpatient hours (overlap over 60%); and the third matching is between the patient's location preferences and the clinic location (distance deviation ≤ 3 km). The retrieval response time is 1-3 seconds. Finally, it outputs a list of doctors (maximum 10) sorted by matching degree, along with clinic and appointment information, which are transmitted to the appointment allocation and appointment interaction management modules, respectively. Its function is to achieve efficient matching between patients and medical resources, filtering out invalid matches. Its significance lies in improving the accuracy of resource matching, reducing invalid patient appointments, and providing a scientific basis for appointment services.

[0043] The dynamic appointment scheduling algorithm is a tool that dynamically adjusts the allocation of appointment slots based on outpatient traffic and appointment progress. Specifically, it combines the allocation ratio of appointment slots, traffic prediction results, and historical allocation data to calculate the number and timing of appointment slots allocated to each department at each time period. In implementation, the system first divides the appointment slots into peak hours (20 minutes / segment) and off-peak hours (60 minutes / segment) based on outpatient operating hours (e.g., 8:00-17:00) and peak patterns (number of patients per hour in the past 3 months), ensuring coverage of over 95% of appointment demand. Then, the appointment quantity calculation unit receives the allocation ratio of available appointment slots (accurate to two decimal places) and traffic prediction results (predicted number of patients in each time slot), and uses an algorithm to calculate the appointment quantity (accurate to a single digit). This is adjusted by referencing historical appointment quantities for the same period (the past 3 identical dates) and utilization rates (used appointment slots / total appointments), with a correction margin of ≤20%, and a single appointment quantity not exceeding 50% of the total allocated slot. Subsequently, the appointment timing judgment unit monitors appointment progress and remaining appointment slots every 10 minutes. If the appointment speed exceeds 1.5 times the historical rate, appointments are made 10-20 minutes earlier; if it is less than 0.5 times, appointments are made 30-60 minutes later, with an accuracy rate ≥90%. Finally, an appointment instruction is generated and transmitted to the appointment interaction management module, and logs are stored for ≥1 year. Its purpose is to adjust the allocation of appointment slots according to actual demand, avoiding a backlog or shortage of slots. Its significance lies in improving the efficiency of appointment utilization, responding to fluctuations in outpatient traffic, and ensuring the stability of appointment services.

[0044] The outpatient traffic dynamic prediction and analysis platform is a tool for accurately predicting outpatient traffic. Specifically, it is an analysis system that integrates historical data, real-time data, and external influencing factors to predict the number of patients visiting each department and at each time period in the next 24 hours. In implementation, the historical data preprocessing unit first receives outpatient traffic data (sample size ≥ 100,000 records) and registration data from the past 1-2 years. Through outlier removal (removing data with a mean of 3 times the standard deviation) and missing value completion (completing with the mean of adjacent time periods, error ≤ 10%), a standardized dataset is formed, categorized by department and 1-hour time period, with completeness ≥ 99%. Then, the influencing factor extraction unit obtains data on holidays (such as Spring Festival and National Day), seasonal climate (temperature accurate to ℃), changes in doctor's outpatient visits, and disease prevalence trends (week-on-week change rate), and uses the Pearson correlation coefficient (|r| ≥ 0). 3) Significant factors are screened and quantified as influence factors of 0-2 (positive 1-2, negative 0-1), with a quantification accuracy ≥90%. Subsequently, the prediction model training unit uses historical datasets and influence factors to train the model using the gradient descent algorithm, adjusting the learning rate (0.01-0.1) and iteration count (1000-5000 times), with a test set error ≤15%, generating a personalized model for the department. Finally, the output unit calls the model, inputs the current influence factors, calculates the predicted number of patients for each time period in the next 24 hours (accurate to single digits), and transmits it to the flexible appointment allocation module with a delay ≤10 seconds. Its function is to accurately predict traffic fluctuations, providing data support for appointment allocation and resource scheduling. Its significance lies in avoiding peak-hour resource congestion and off-peak waste, improving the level of refined outpatient management.

[0045] like Figure 2As shown, a big data-based outpatient information query and appointment management method is applied to an outpatient information query and appointment management system, including the following steps: First, the outpatient information collection and processing module collects basic patient information, historical medical records, doctor's outpatient information, and clinic resource data in real time. Unstructured data is converted into a unified format using structured processing, and the processed data is sent to the medical graph construction and retrieval module and the traffic dynamic prediction module, respectively. Second, the medical graph construction and retrieval module constructs a medical graph based on the received structured data, including the patient-doctor-disease-clinic relationship. It uses a triple-link retrieval logic to match patient demand information with doctor resource information, and transmits the matching results to the appointment allocation module and the appointment interaction management module, respectively. Third, the appointment allocation module calls the outpatient appointment dynamic allocation model, inputs the medical graph retrieval results and outpatient resource data, and calculates the allocation of appointments for each department at different times. The process involves six steps: First, the appointment allocation ratio is determined and transmitted to the appointment allocation module. Second, the dynamic traffic prediction module calculates the predicted outpatient traffic for each department within a preset time period based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease incidence patterns. This prediction result is then sent to the appointment allocation module. Third, the dynamic appointment allocation module uses a dynamic appointment allocation scheduling algorithm, combining the appointment allocation ratio and traffic prediction results, to determine the number and timing of appointments for each department within each allocation period. This generates an appointment allocation instruction and transmits it to the appointment interaction management module. Fourth, the appointment interaction management module receives the medical atlas retrieval results and the appointment allocation instruction. It provides patients with an appointment query interface to search for appointment information for each department, receives appointment requests submitted by patients, prioritizes them according to the appointment request processing logic, provides patients with appointment status information, and simultaneously sends the patient's appointment operation data to the outpatient information collection and processing module for data updates.

[0046] A big data-based outpatient information query and appointment management system and method addresses the problems of fragmented information and low matching accuracy in traditional systems. The outpatient information collection and processing module integrates patient information, medical records, and doctor's outpatient data to form a unified data system. The medical graph construction and retrieval module achieves accurate matching of patient and doctor resources based on multi-dimensional relationships, avoiding matching biases caused by fragmented information. To address the issues of rigid and inefficient allocation of appointment slots, the appointment slot balancing module calculates scientific allocation ratios using multi-dimensional data, while the appointment slot flexible allocation module dynamically adjusts allocation strategies based on real-time appointment progress and historical usage, replacing fixed ratios and manual adjustments to resolve the contradiction between appointment slot shortages and idle slots. Finally, to address the problems of simplistic traffic prediction and blind resource scheduling, the dynamic traffic prediction module constructs a prediction platform integrating multiple factors such as holidays, seasonal disease patterns, and emergency-to-outpatient transfers, replacing single historical data extrapolations and providing a reliable basis for resource scheduling, alleviating peak-hour congestion and off-peak waste.

[0047] At the resource allocation level, this invention achieves efficient utilization of resources such as doctors, consultation rooms, and appointment slots through dynamic and balanced allocation and flexible deployment of appointment slots, reducing resource idleness and waste, and improving the overall rationality of medical resource allocation. At the service efficiency level, the appointment interaction management module optimizes the appointment process, providing convenient query, appointment, and status feedback services based on patient needs and appointment availability, shortening patient appointment operation time. At the same time, accurate traffic prediction and resource scheduling reduce patient waiting time. At the management precision level, multi-module collaboration and model algorithm support enable the transformation of outpatient management from traditional manual dominance to data-driven approaches. Through real-time data collection, dynamic analysis, and precise control, the invention improves the refinement and intelligence of outpatient management, providing strong technical support for improving the quality of outpatient services.

[0048] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0049] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A big data-based outpatient information query and appointment management system, characterized in that, include: The module includes: outpatient information collection and processing module, medical atlas construction and retrieval module, balanced allocation of appointment slots module, dynamic traffic prediction module, flexible allocation of appointment slots module, and appointment interaction management module. The outpatient information collection and processing module collects and structures basic information of outpatients, historical medical records, doctor's outpatient information, and clinic resource data in real time, and transmits the processed data to the medical atlas construction and retrieval module. The medical atlas construction and retrieval module constructs a medical atlas based on the received structured data, including the relationships between patients, doctors, symptoms, and clinics. It retrieves information matching patient needs with doctor resources through a triple-link retrieval logic, and the retrieval results are transmitted to the appointment allocation module and the appointment interaction management module, respectively. The appointment allocation module calls the outpatient appointment dynamic allocation model, combines the medical atlas retrieval results with outpatient resource data, calculates the allocation ratio of appointments for each department, and transmits the allocation results to the appointment flexible allocation module. The traffic dynamic prediction module constructs an outpatient traffic dynamic prediction and analysis platform based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease peak patterns, and outputs the predicted outpatient traffic for each department within a preset future time period to the appointment flexible allocation module. The flexible appointment allocation module calls the dynamic flexible appointment allocation scheduling algorithm to determine the number of appointments allocated in each time period and the timing of allocation based on the appointment allocation ratio and traffic prediction results. The allocation instructions are transmitted to the appointment interaction management module. The appointment interaction management module receives the medical atlas search results and appointment allocation instructions, and provides patients with appointment query, appointment application submission and appointment status feedback functions. At the same time, it feeds back the patient appointment operation data to the outpatient information collection and processing module.

2. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The expression for the dynamic balanced allocation model of outpatient appointment slots in the appointment slot allocation module is as follows: ,in, Let t be the number of appointment slots allocated to the j-th doctor in the i-th department at time t. Assign weight coefficients to the number sources. Let t be the historical average number of patients needed for the i-th department. Let j be the consultation time of the jth doctor at time t. The matching degree between the j-th doctor in the i-th department and the common diseases in the department. This is the departmental resource adjustment coefficient. This represents the total number of departments in the hospital. Let be the number of remaining appointment slots for the i-th department at time t. This represents the historical average number of patients needed for the k-th department at time t. This represents the adjustment coefficient for the remaining signal sources.

3. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The expression for the medical atlas triple-link retrieval model in the medical atlas construction and retrieval module is as follows: ,in, The degree of triple link matching between patient p, doctor d, and clinic s. The weights for the historical medical visits of patient p and doctor d are given. Let be the correlation coefficient between the geographical locations of patient p and clinic s. Let be the matching coefficient between doctor d and clinic s in terms of outpatient schedules. The total number of medical records for patient p. Let be the total number of patients seen by doctor d. The historical number of times the clinic s has been used. For correlation compensation coefficient, This represents the total number of disease types. The degree of matching between patient p and doctor d for the treatment of disease type m. Let be the frequency coefficient of doctor d seeing patients with type m disease in clinic s.

4. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The expression for the dynamic flexible allocation scheduling algorithm in the flexible allocation module is as follows: ,in, Let t represent the number of appointment slots available for the i-th department during time period t. This is the elasticity coefficient for deployment. The predicted number of patients visiting the i-th department during time period t is output by the traffic dynamic prediction module. for The number of remaining appointment slots in the i-th department during the time period. The utilization rate of clinic resources in the i-th department. This is the time-period adjustment coefficient. The rate of sudden outpatient visits in the i-th department during time period t. The outpatient rate of doctors in the i-th department. This represents the historical impact coefficient. For historical period numbers, The number of appointment slots available for the i-th department during time period h. The utilization rate of appointment slots for the i-th department during time period h.

5. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The prediction model expression of the outpatient traffic dynamic prediction and analysis platform in the traffic dynamic prediction module is as follows: ,in, For the predicted outpatient flow of the i-th department during time period t, Historical traffic weighting coefficient for The actual outpatient flow of the i-th department during the time period The number of appointments made in the i-th department during time period t. Let be the appointment fulfillment rate of the i-th department during time period t. The estimated number of emergency room visits for the i-th department during time period t. The emergency-to-outpatient referral rate of the i-th department during time period t. Let t represent the number of doctors from the i-th department who made outpatient visits during time period t. The number of influencing factor types, Let be the impact value of factor q on the traffic volume of department i during time period t. Let be the weight coefficient of the q-th factor. This represents the adjustment coefficient for influencing factors, used to adjust the strength of the effect of various influencing factors on outpatient traffic prediction and balance the weights of various influencing factors in the model.

6. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The expression for the appointment request processing model in the appointment interaction management module is: ,in, Assign a priority score to patient p for the appointment request of department i during time period t. Let be the matching coefficient between patient p and the common diseases in department i. The remaining availability of appointment slots for the i-th department during time period t. The matching degree of patient p's visit time during time period t. Let be the total number of appointment requests for the i-th department during time period t. For patient p, the number of appointments made in history. This represents the historical influence coefficient. The number of departments the patient has visited in their history. For the history department serial number, For patient p, the historical satisfaction level during their visits to department r. Let represent the resource correlation between department r and department i during time period t.

7. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The appointment allocation module includes an appointment demand analysis unit, a resource data integration unit, an allocation ratio calculation unit, and an allocation result verification unit. The appointment demand analysis unit receives basic patient information and historical medical records from the outpatient information collection and processing module, extracts data on patient ailment types, frequency of visits, and appointment preferences for each department, classifies patient groups according to urgency of ailments and time requirements, and statistically analyzes the demand for appointments in each department and the distribution of demand time periods for different categories of patients. The resource data integration unit collects doctor's outpatient information, clinic resource data, and equipment usage status data to determine each doctor's capacity, clinic capacity, and equipment usage time limits, and integrates doctor information with corresponding clinics and equipment... The information is linked and bound to form a resource association data table; the allocation ratio calculation unit calls the outpatient appointment dynamic balance allocation model, inputs the demand data output by the appointment demand analysis unit and the resource data output by the resource data integration unit, calculates the basic allocation ratio of appointments for each department at different times, and adjusts the basic allocation ratio by combining the doctor-patient matching information output by the medical atlas construction retrieval module; the allocation result verification unit compares the appointment allocation ratio of each department with the historical allocation data and the resource carrying capacity limit, and checks whether the allocation ratio complies with the outpatient operation specifications. If there is a situation that exceeds the resource carrying capacity limit or deviates too much from the historical reasonable allocation range, it is fed back to the allocation ratio calculation unit for recalculation.

8. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The dynamic traffic prediction module includes a historical data preprocessing unit, an influencing factor extraction unit, a prediction model training unit, and a prediction result output unit. The historical data preprocessing unit receives historical outpatient traffic data and registration data transmitted from the outpatient information collection and processing module, removes outliers and missing values, and classifies and organizes the data by department and time period to form a standardized historical traffic dataset. The influencing factor extraction unit obtains holiday schedule and seasonal climate change data from external data sources, extracts doctor's outpatient schedule changes and disease prevalence trends from internal outpatient data, analyzes the correlation between various factors and outpatient traffic, screens out factors that have a significant impact on traffic and quantifies them as influencing factors. The prediction model training unit uses the historical traffic dataset and the extracted influencing factors to train the prediction model of the outpatient traffic dynamic prediction and analysis platform, optimizes the prediction accuracy by adjusting the model parameters, and generates personalized prediction models suitable for each department. The prediction result output unit calls the trained personalized prediction model, inputs the influencing factor data of the current time period, calculates the predicted outpatient traffic of each department in the future preset time period, and transmits the predicted values ​​to the flexible allocation module of appointment slots after classifying them by department and time period.

9. The outpatient information query and appointment management system based on big data according to claim 1, characterized in that, The flexible allocation module for appointment slots includes an allocation time period division unit, an allocation quantity calculation unit, an allocation timing judgment unit, and an allocation instruction generation unit. The allocation time period division unit divides a day into multiple consecutive allocation time periods based on outpatient operating hours and peak patient visit patterns, and determines the duration and start time of each time period to ensure that each time period covers the main appointment needs of patients. The number of appointments to be allocated is calculated by receiving the number of appointments allocated by the number of appointments balanced allocation module and the traffic prediction result from the traffic dynamic prediction module. The module calls the dynamic appointments elastic allocation scheduling algorithm to calculate the number of appointments allocated to each department in each allocation period, and makes corrections by referring to the historical allocation and utilization data of the same period. The timing of appointments is determined by the real-time monitoring of appointment progress and remaining appointment slots in each department. It analyzes the appointment speed of appointment slots in the current time slot and determines whether the appointment slots for the next time slot need to be released earlier or later. If the appointment speed is too fast, resulting in the early exhaustion of appointment slots, or too slow, resulting in a backlog of appointment slots, the timing of appointments is adjusted. Based on the calculation results of the number of appointments to be made and the judgment results of the timing of the appointments, the appointment instruction generation unit generates an appointment allocation instruction that includes the department identifier, the time period for allocation, and the number of appointments to be made. The instruction is then transmitted to the appointment interaction management module, and the generation time and parameter information of the appointment instruction are recorded at the same time.

10. A method for outpatient information query and appointment management based on big data, characterized in that, This method is applied to the outpatient information query and appointment management system based on big data as described in claim 1, and includes the following steps: First, the outpatient information collection and processing module collects basic information of outpatients, historical medical records, doctor's outpatient information, and clinic resource data in real time. Unstructured data is converted into a unified format using structured processing, and the processed data is sent to the medical graph construction and retrieval module and the traffic dynamic prediction module, respectively. Second, the medical graph construction and retrieval module constructs a medical graph including patient-doctor-disease-clinic relationships based on the received structured data. It uses a triple-link retrieval logic to match patient demand information with doctor resource information, and transmits the matching results to the appointment allocation module and the appointment interaction management module, respectively. Third, the appointment allocation module calls the outpatient appointment dynamic allocation model, inputs the medical graph retrieval results and outpatient resource data, and calculates the appointment allocation ratio for each department at different times. The allocation ratio data is transmitted to the flexible appointment allocation module. Fourth, the dynamic traffic prediction module, based on historical outpatient traffic data, real-time registration data, holiday factors, and seasonal disease peak patterns, calculates the predicted outpatient traffic for each department within a preset time period using the outpatient traffic dynamic prediction analysis platform, and sends the prediction results to the flexible appointment allocation module. Fifth, the flexible appointment allocation module calls the dynamic flexible appointment allocation scheduling algorithm, combining the appointment allocation ratio and traffic prediction results to determine the number and timing of appointments for each department within each allocation period, generates appointment allocation instructions, and transmits them to the appointment interaction management module. Sixth, the appointment interaction management module receives the medical atlas retrieval results and appointment allocation instructions, provides patients with an appointment query interface for querying appointment information for each department, receives appointment applications submitted by patients, prioritizes them according to the appointment request processing logic, provides patients with appointment status information, and simultaneously feeds back patient appointment operation data to the outpatient information collection and processing module to update the data.

Images