A hospital patient visit plan recommendation method
By dynamically modeling the hospital's integrated management system, personalized medical plans are generated, which solves the problem of information asymmetry in the hospital's examination queuing system and improves patients' medical efficiency and experience.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING PUBLIC HEALTH MEDICAL TREATMENT CENT
- Filing Date
- 2026-03-17
- Publication Date
- 2026-06-16
Smart Images
Figure CN122224441A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of smart healthcare auxiliary service technology, and more specifically, to a method for recommending in-hospital patient treatment plans. Background Technology
[0002] The current hospital examination queuing system suffers from severe information asymmetry. Patients who haven't yet arrived at the examination window cannot obtain real-time information such as the number of people queuing in each department and the waiting time, leading to difficulties in decision-making. The core pain points are concentrated in the following aspects: First, patients face an information black box, often having to repeatedly run around to check the queuing situation of each department, resulting in significant waste of time; second, when undergoing multiple examinations, intelligent route planning is not available, forcing patients to passively become human dispatchers.
[0003] Hospitals can collect a wealth of data using existing equipment (e.g., doctor terminals record patient information and ordered tests, service terminals provide queue numbers, and doctors can record test results; service terminals also have built-in hospital map navigation). However, hospitals are not currently utilizing this data effectively. A key challenge for hospitals is how to dynamically model and provide personalized treatment plans (e.g., recommended procedures, routes, and estimated treatment time) based on this information. Summary of the Invention
[0004] The purpose of this application is to provide a method for recommending in-hospital patient treatment plans. By collecting information from each node and performing real-time dynamic planning, the method can intelligently recommend treatment processes, routes, and time estimates for each patient's required examinations. This provides patients with more intelligent services, effectively improves their medical efficiency, minimizes queuing time at various locations, enhances their medical experience, and also improves the hospital's patient capacity.
[0005] To achieve the above objectives, the embodiments of this application are implemented in the following manner: This application provides a method for recommending in-hospital patient treatment plans. The in-hospital integrated management system includes a server, a doctor terminal, a task terminal, and a service terminal. The method is applied to the server and includes: obtaining patient information entered by the attending physician through the doctor terminal and adding it to the patient information set for the day, wherein each patient information includes the patient's basic information, unique number, task information, and attending physician information; obtaining the patient's query request through the service terminal, wherein the query request includes the patient's unique number and the location information of the service terminal; determining the patient's target task information based on the patient's unique number, and obtaining the target location and target task terminal corresponding to each task in the target task information, and obtaining the target task scheduling data through the target task terminal; generating a recommended treatment plan based on the target location and target task scheduling data corresponding to each task, wherein the recommended treatment plan includes a recommended treatment process, a recommended treatment route, and an estimated treatment time; and feeding back the recommended treatment plan to the service terminal that sent the query request, so that the service terminal can display the recommended treatment plan to the patient.
[0006] Furthermore, obtaining the patient's query request through the service terminal includes: obtaining the patient's unique number entered through the service terminal and generating a query request; or, scanning the QR code in the patient's document through the service terminal and generating a query request.
[0007] Furthermore, based on the patient's unique ID, the patient's target task information is determined, and the target location and target task terminal corresponding to each task in the target task information are obtained. Target task scheduling data is then obtained through the target task terminal, including: determining the patient's target task information based on the patient's unique ID, wherein the target task information contains at least one task, each task corresponds to a target location, and each target location has at least one target task terminal; obtaining the target location and target task terminal corresponding to each task in the target task information; and obtaining the target task scheduling data for each target task terminal, wherein the target task scheduling data includes completed scheduling data and real-time scheduling data for the target task on that day.
[0008] Furthermore, based on the target location and target task scheduling data corresponding to each task, a recommended medical treatment plan is generated, including: performing task problem modeling, path problem modeling, and dynamic time consumption modeling based on task information and each target location; generating an objective function to be optimized; generating constraints based on completed scheduling data and real-time scheduling data; finding the optimal solution for the objective function, and determining the optimal solution as the recommended medical treatment plan.
[0009] Furthermore, based on task information and each target location, task problem modeling, path problem modeling, and dynamic time consumption modeling are performed, including: constructing a patient task set based on task information. , This refers to the number of tasks the patient needs to complete. For the first Task ; Based on location information, target location, and attending physician information, a location set is constructed. ,in, For the patient's location information, For the first Task target location The location of the attending physician; Based on patient task set and location set Construct path sequence ,in, , , For the first in the path sequence One of the points along the way is one of the target locations; Define arrival time series ,in, At the current time point, For the patient's route to the destination The point in time, For the patient's route to the destination The point in time, This refers to the time at which the patient arrives at the location of the attending physician.
[0010] Furthermore, the objective function to be optimized is generated, including: based on the path sequence. and arrival time series Generate the objective function to be optimized. : , in, Let be the objective function to be optimized. This is the equivalent optimization coefficient for path distance per unit time. For the first Item weight coefficient, , Indicates the route from the points along the way to the places along the way The shortest path.
[0011] Furthermore, weighting coefficients satisfy: .
[0012] Furthermore, based on completed and real-time order data, constraints are generated, including those based on location sets. First, generate the first constraint; based on the patient information set, completed scheduling data, and real-time scheduling data, determine potential scheduling data; based on the completed scheduling data, real-time scheduling data, and potential scheduling data, generate the second constraint; based on the path sequence... This generates the third constraint.
[0013] Furthermore, the first constraint is: path sequence In For the set of locations All permutations, For the first in the path sequence Each point along the route is one of the target locations, indicating that the execution order of each task is variable; The second constraint is: , , in, For the patient's route to the destination The point in time, Patient arrives at the location along the way The point in time, Indicates the patient's route through locations to the places along the way The duration of movement, Indicates the patient's location along the route The required length of stay, Indicates the places passed through The initial queue length at the location, Locations along the route The number of new employees per unit time at the location Locations along the route The number of tasks processed per unit time at the location. Locations along the route Execution time of a single task at the location; The third constraint is: ,in, This indicates a group of tasks with dependencies. For a set of dependencies, Indicates task The target location In path sequence The index number in the middle, Indicates task The target location In path sequence The index number in the database.
[0014] Furthermore, the optimal solution to the objective function is determined, and the optimal solution is identified as the recommended medical treatment plan, including: obtaining the patient's location information. The location of the attending physician Dependency set various locations along the route Initial queue length at New number of people per unit time Number of tasks processed per unit time Execution time of a single task The system generates a travel time matrix between each pre-stored route location and a candidate task sequence that satisfies all constraints. For each candidate task sequence, a corresponding path sequence is constructed, and the corresponding arrival time sequence is calculated. The system calculates the objective function value for each candidate task sequence and selects the candidate task sequence with the smallest objective function value as the optimal solution. Based on the optimal solution, a recommended medical treatment plan is generated, which includes a recommended medical treatment process, a recommended medical treatment route, and an estimated medical treatment time.
[0015] Beneficial Effects: This solution utilizes the hospital's integrated management system (including servers, doctor terminals, task terminals, and service terminals) to collect necessary information: Doctor terminals retrieve patient information entered by attending physicians and add it to the daily patient information set. Each patient's information includes basic patient information, a unique ID, task information, and attending physician information. Service terminals retrieve patient query requests (generated through the patient's unique ID or by scanning a QR code on the patient's document). Query requests include the patient's unique ID and the location information of the service terminal (this location information is the patient's current location). Based on the patient's unique ID, the solution determines the patient's target task information and retrieves the target location and target task terminal for each task. The target task terminal retrieves target task scheduling data (including completed and real-time scheduling data for the day). Based on the target location and scheduling data for each task, a recommended treatment plan is generated (including a recommended treatment process, recommended route, and estimated treatment time). The recommended treatment plan is then fed back to the service terminal that sent the query request, allowing the service terminal to display the recommended treatment plan to the patient. This solution utilizes a collaborative framework comprised of a doctor's terminal (task generation), a service terminal (location and interaction), a task terminal (real-time queue collection, including completed and real-time scheduling data), and a server (dynamic programming) to intelligently provide patients with recommended treatment plans (including recommended procedures, routes, and estimated time) for each patient's required examinations. This provides patients with more intelligent services, effectively improving their medical efficiency, minimizing queuing time at various locations, enhancing their overall experience, and simultaneously increasing the hospital's patient capacity.
[0016] Based on task information and each target location, perform task problem modeling (construct a patient task set). , Therefore, the number of tasks to be completed by the patient) and path problem modeling (path sequence) are used. , , , For the first in the path sequence (locations along the route) and dynamic time consumption modeling (arrival time series) , At the current time point, For the patient's route to the destination The point in time, (Given the time point at which the patient arrives at the location of the attending physician), generate the objective function to be optimized. Based on completed and real-time scheduling data, constraints are generated (including three constraints, comprehensively considering task path constraints, dynamic time consumption constraints, and task dependency constraints). Considering that the number of examinations for a single patient is not too high (usually no more than 5), an exhaustive approach can be adopted to strictly obtain the optimal solution, thus determining the optimal solution as the recommended treatment plan. Overall, this approach comprehensively considers various factors and sets reasonable and effective constraints during dynamic modeling. In the dynamic time consumption constraint, not only the static queue situation is considered, but also the new queues that may be generated during the time it takes for patients to reach the target location and obtain a queue number. These new queues depend on potential task scheduling data (depending on patients who have already been prescribed by doctors but have not completed them and are not in queue, as well as patients who may be added later to complete the task), which makes the prediction more accurate. In designing the objective function, the focus is not simply on the shortest time or the shortest path, but on a comprehensive approach. When prioritizing tasks, timeliness is paramount because the earlier the timeliness prediction is made, the more accurate it is. As the prediction time progresses, the error in the timeliness prediction gradually accumulates. Therefore, in planning tasks that are scheduled to be completed later, the weight of the path component is gradually increased to reduce the patient's travel burden. Ultimately, a suitable recommended medical plan is generated, providing patients with a recommended medical process, path, and estimated time consumption, minimizing queuing time, providing relevant information, and effectively improving the patient's medical experience.
[0017] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a schematic diagram of the hospital's integrated management system architecture.
[0020] Figure 2 This is a diagram of the patient's receipt.
[0021] Figure 3 This is a schematic diagram of a wall-mounted service terminal.
[0022] Figure 4 A flowchart for recommending methods for in-hospital patient visit plans.
[0023] Figure 5 A diagram illustrating how a patient's list of examination items is displayed on a service terminal.
[0024] Figure 6 A diagram illustrating the recommended medical treatment process for service terminals. Detailed Implementation
[0025] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.
[0026] Please see Figure 1 , Figure 1 This is a schematic diagram of the architecture of the hospital's integrated management system. In this embodiment, the integrated management system includes a server, doctor terminals, task terminals, and service terminals.
[0027] Each attending physician corresponds to one physician terminal (or multiple physician terminals; this embodiment uses one physician terminal as an example). The attending physician can prescribe necessary tests, oral or topical medications, and payment instructions for the patient; these are all considered tasks. The attending physician can enter patient information (including basic patient information, unique ID, task information, and attending physician information) into the system and upload it to the server. At this point, the doctor can issue a receipt to the patient, which includes patient information and a QR code, such as... Figure 2 As shown ( Figure 2 It was obtained by merging two documents from the same patient, with patient privacy information redacted. Currently, each task within the hospital requires a separate document, but merging may be possible in the future.
[0028] Each service location open to patients within the hospital is equipped with at least one task terminal, such as registration, payment, pharmacy, electrocardiogram testing, color Doppler ultrasound testing, and blood routine testing. Medical staff at these locations can enter service information into the system and upload it to the server after providing services to patients.
[0029] The number of service terminals: Each location should generally be equipped with at least one service terminal, providing ticket dispensing and inquiry services. The location (or area) of each service terminal is relatively fixed and dedicated to that specific location. To facilitate patient inquiries, some wall-mounted service terminals for route searching can also be deployed throughout the hospital (these do not provide ticket dispensing functions but can be used for patient inquiries), such as... Figure 3 As shown.
[0030] The server is used to run the in-hospital patient visit plan recommendation method provided in this embodiment. Please refer to [link / reference]. Figure 4 , Figure 4 A flowchart for recommending methods for in-hospital patient visit plans.
[0031] In this embodiment, the method for recommending in-hospital patient treatment plans may include steps S10, S20, S30, S40, and S50.
[0032] First, the server can run step S10.
[0033] Step S10: Obtain the patient information entered by the attending physician through the doctor's terminal and add it to the patient information set for the day. Each patient information includes the patient's basic information, unique number, task information and attending physician information.
[0034] In this embodiment, the server can obtain patient information entered by the attending physician through the doctor's terminal. On a daily basis, each patient information is added to the patient information set for that day. Each patient information includes the patient's basic information, unique number (such as registration number), task information (such as 1. 12-channel routine electrocardiogram + electrocardiogram event record, 2. abdominal color Doppler ultrasound) and attending physician information (mainly used to associate the location of the attending physician).
[0035] It should be noted that in some cases, patients need to return to the attending physician after completing all procedures, while in others, they do not. This embodiment assumes the patient needs to return to the doctor. During modeling, the starting point is fixed as the patient's location, and the ending point is the attending physician's location. Of course, for cases where the patient does not need to return to the doctor, the ending point can be set to a fixed location (such as the hospital exit or lobby), or it can be set to the target location of the last task, making the path distance between the target location of the last task and the ending point zero. For ease of explanation, this embodiment uses the attending physician's location as the ending point, as other cases are not significantly different and have similar mechanisms, differing only in the final path distance. Therefore, using the attending physician's location as the ending point should not be considered a limitation of this application.
[0036] After that, the server can receive query requests initiated by any patient in real time. At this time, the server can run step S20.
[0037] Step S20: Obtain the patient's query request through the service terminal, wherein the query request includes the patient's unique number and the location information of the service terminal.
[0038] In this embodiment, any patient can enter their unique number through the service terminal (e.g., manually, by swiping a card). Once the service terminal obtains the entered unique number, it can generate a query request. Alternatively, any patient can also generate a query request by scanning the QR code on their medical record through the service terminal.
[0039] After generating a query request, the service terminal can send the query request (carrying the service terminal's location information, or carrying the service terminal's ID, which is associated with the service terminal's location information) to the server. At this point, the server can obtain the patient's query request through the service terminal.
[0040] Upon receiving a query request, the server can proceed to step S30.
[0041] Step S30: Based on the patient's unique number, determine the patient's target task information, and obtain the target location and target task terminal corresponding to each task in the target task information, and obtain the target task scheduling data through the target task terminal.
[0042] In this embodiment, the server can determine the patient's target task information based on the patient's unique ID. This target task information includes at least one task, each task corresponding to a target location, and each target location having at least one target task terminal. The server can then obtain the target location and target task terminal corresponding to each task in the target task information, and retrieve the target task scheduling data (including completed scheduling data and real-time scheduling data for the current day) through the target task terminal. Of course, for hospitals, the number of task terminals is not particularly large; therefore, in this embodiment, the server can obtain task scheduling data from all task terminals to facilitate accurate incremental prediction during subsequent modeling.
[0043] Accordingly, the server can proceed to step S40.
[0044] Step S40: Based on the target location and target task scheduling data corresponding to each task, generate a recommended medical treatment plan, which includes a recommended medical treatment process, a recommended medical treatment route, and an estimated medical treatment time.
[0045] In this embodiment, the server can perform task problem modeling, path problem modeling, and dynamic time consumption modeling based on task information and each target location.
[0046] For example, based on task information, the server can construct a set of patient tasks: (1) in, This refers to the number of tasks the patient needs to complete. For the first Task .
[0047] Furthermore, based on location information, target location, and attending physician information, a location set is constructed: (2) in, For the patient's location information, For the first Task target location The location of the attending physician (this embodiment uses the location of the attending physician as an example, but it is not limited to this. In other embodiments, for cases where it is not necessary to return to the location of the attending physician, a fixed location can be used as the destination, such as a certain place in the lobby on the first floor of the hospital).
[0048] And, based on patient task sets and location set Construct the path sequence: (3) in, , , For the first in the path sequence One of the places along the way is one of the destination locations.
[0049] And, define the arrival time series: (4) in, At the current time point, For the patient's route to the destination The point in time, For the patient's route to the destination The point in time, The location for the patient to arrive at the attending physician's location. The point in time.
[0050] Accordingly, the server can generate the objective function to be optimized. Considering that the estimation of newly added queues will gradually accumulate and amplify the error over time, this embodiment considers both timeliness and distance to construct a comprehensive objective function. : (5) in, Let be the objective function to be optimized. The equivalent optimization coefficient for unit time and path distance (in this embodiment, time is in minutes and path distance is in meters. A value between 15 and 30 is more appropriate for this equivalent optimization coefficient. In this embodiment, 20 is used as an example). For the first Item weight coefficient, , Indicates the route from the points along the way to the places along the way The shortest path.
[0051] Weighting coefficient satisfy: (6) Since the earlier the time-delivery forecast is made, the more accurate it is. As the forecast time is pushed forward, the error in the time-delivery forecast will gradually accumulate. Therefore, in the planning of tasks that need to be completed later in the order, the weight of the path part should be gradually increased, which will help reduce the travel burden on patients.
[0052] Afterwards, the server can generate constraints based on the completed order data and the real-time order data.
[0053] For example, the server can be based on a set of locations. The first constraint condition is generated. The first constraint condition is: Path sequence In For the set of locations All permutations, For the first in the path sequence Each point along the route is one of the target locations, indicating that the execution order of each task is variable.
[0054] Furthermore, the server can determine potential scheduling data based on the patient information set, completed scheduling data, and real-time scheduling data: by subtracting the completed scheduling data and real-time scheduling data for each task from the number of patients in the patient information set for each task, the number of patients who are certain to perform the task but have not yet entered the queue stage is obtained. This is one part of the potential scheduling data. The other part is the number of patients newly treated by the attending physician during this period, and the number of patients who need to complete the task at the prescription counter. This part can be determined by setting an empirical value (based on the analysis of historical data, how many new patients for the task are added per unit time) to ultimately obtain the potential scheduling data.
[0055] Therefore, the server can generate a second constraint based on completed order data, real-time order data, and potential order data: (7) (8) in, For the patient's route to the destination The point in time, Patient arrives at the location along the way The point in time, Indicates the patient's route through locations to the places along the way The duration of movement, Indicates the patient's location along the route The required length of stay, Indicates the places passed through The initial queue length at the location, Locations along the route The number of new employees per unit time at the location Locations along the route The number of tasks processed per unit time at the location. Locations along the route Execution time of a single task at the location; Furthermore, the server can be based on path sequences. Generate the third constraint: (9) in, This indicates a group of tasks with dependencies. For a set of dependencies, Indicates task The target location In path sequence The index number in the middle, Indicates task The target location In path sequence The index number in the sequence. The third constraint indicates that task groups with dependencies must be forced to conform to the order of dependencies in the path sequence planning.
[0056] Overall, the first constraint condition constrains the path sequence. The required rules are that each transit point must originate from the target location of the objective task, i.e., from the set of locations. However, they can be arranged freely. The second constraint constrains the dynamic time consumption, and the third constraint constrains the dependencies between tasks. This can be seen as a supplement to the first constraint. When arranging the task order, it is necessary to exclude arrangements that do not conform to the dependencies.
[0057] After modeling is completed, the server can find the optimal solution for the objective function and determine the optimal solution as the recommended medical treatment plan.
[0058] Since the number of tasks for each patient is relatively small, generally not exceeding 5, an exhaustive search approach is suitable for finding the global optimal solution.
[0059] In this embodiment, the server can obtain the patient's location information. (Obtained through the location information of the service terminal, as the patient is making the inquiry at this time and place), and the location of the attending physician. (Determined through information associated with the attending physician), Dependency set (System presets), various locations along the route Initial queue length at (Real-time acquisition) Number of new users per unit time (Determined through potential scheduling data), number of tasks processed per unit time (Calculated from real-time acquired completed order data, or preset experience values, or calculated by combining completed order data with historical values), single task execution time. (Acquired in real time, or based on preset experience values, or calculated by combining both) and a pre-stored matrix of travel time between each route location (this is preset in the system, because the target locations of each task within the hospital are known and can be set through multiple test values).
[0060] Then, the server can generate candidate task sequences that satisfy all constraints, construct a corresponding path sequence for each candidate task sequence, calculate the corresponding arrival time sequence, calculate the objective function value for each candidate task sequence, and select the candidate task sequence with the smallest objective function value as the optimal solution.
[0061] Based on this, the server can generate a recommended medical treatment process (i.e., the task sequence A corresponding to the optimal solution) and a recommended medical treatment route (i.e., the path sequence corresponding to the optimal solution) based on the optimal solution. ) and estimated time to medical visit (based on arrival time series) (Confirmed) Recommended medical treatment plan.
[0062] At this point, the server can proceed to step S50.
[0063] Step S50: Feedback the recommended medical treatment plan to the service terminal that sent the query request, so that the service terminal can display the recommended medical treatment plan to the patient.
[0064] In this embodiment, the server can provide a recommended medical treatment plan to the service terminal that sent the query request, so that the service terminal can display the recommended medical treatment plan to the patient, such as... Figure 5 As shown. Of course, to make things easier for patients, printing services can also be provided, or the recommended medical plan can be sent to the patient's associated terminal device (such as the patient's mobile phone, which can scan a QR code through a mobile app to establish a connection between the patient and their own terminal device, thereby sending the recommended medical plan to the patient's terminal device).
[0065] In addition, regarding the recommended route display for the medical visit plan, since the example provided had relatively few examination tasks, a route with more examination tasks was used for a more comprehensive display, such as... Figure 6 As shown: The patient's examination tasks include blood tests, color Doppler ultrasound, electrocardiogram, and CT scan. The path of each examination task and the estimated total time (including the travel time of the planned route) are displayed.
[0066] It should be noted that in this embodiment, the reason for using a service terminal (and even potentially an additional wall-mounted service terminal) is primarily to obtain the patient's accurate location information. Currently, directly using the patient's mobile phone for location tracking makes it difficult to identify floors within a building, resulting in poor effectiveness in generating recommended treatment routes to guide the patient. However, if better mobile phone location technology becomes available in the future to solve this floor location problem, this solution can also replace it by using the patient's terminal device to generate query requests and directly provide the generated recommended treatment process to the patient's terminal.
[0067] In summary, this application provides a method for recommending in-hospital patient treatment plans. It utilizes an in-hospital integrated management system (including a server, doctor terminals, task terminals, and service terminals) to collect necessary information: The doctor terminal obtains patient information entered by the attending physician and adds it to the daily patient information set. Each patient information includes basic patient information, a unique ID, task information, and attending physician information. The service terminal obtains the patient's query request (generated by scanning the patient's unique ID or a QR code on the patient's document). The query request includes the patient's unique ID and the location information of the service terminal (this location information is the patient's current location). Based on the patient's unique ID, the target task information is determined, and the target location and target task terminal corresponding to each task are obtained. The target task scheduling data (including completed and real-time scheduling data for the target task on that day) is obtained through the target task terminal. Based on the target location and target task scheduling data for each task, a recommended treatment plan (including a recommended treatment process, a recommended treatment route, and an estimated treatment time) is generated. The recommended treatment plan is then fed back to the service terminal that sent the query request, allowing the service terminal to display the recommended treatment plan to the patient. This solution utilizes a collaborative framework comprised of a doctor's terminal (task generation), a service terminal (location and interaction), a task terminal (real-time queue collection, including completed and real-time scheduling data), and a server (dynamic programming) to intelligently provide patients with recommended treatment plans (including recommended procedures, routes, and estimated time) for each patient's required examinations. This provides patients with more intelligent services, effectively improving their medical efficiency, minimizing queuing time at various locations, enhancing their overall experience, and simultaneously increasing the hospital's patient capacity.
[0068] Based on task information and each target location, perform task problem modeling (construct a patient task set). , Therefore, the number of tasks to be completed by the patient) and path problem modeling (path sequence) are used. , , , For the first in the path sequence (locations along the route) and dynamic time consumption modeling (arrival time series) , At the current time point, For the patient's route to the destination The point in time, (Given the time point at which the patient arrives at the location of the attending physician), generate the objective function to be optimized. Based on completed and real-time scheduling data, constraints are generated (including three constraints, comprehensively considering task path constraints, dynamic time consumption constraints, and task dependency constraints). Considering that the number of examinations for a single patient is not too high (usually no more than 5), an exhaustive approach can be adopted to strictly obtain the optimal solution, thus determining the optimal solution as the recommended treatment plan. Overall, this approach comprehensively considers various factors and sets reasonable and effective constraints during dynamic modeling. In the dynamic time consumption constraint, not only the static queue situation is considered, but also the new queues that may be generated during the time it takes for patients to reach the target location and obtain a queue number. These new queues depend on potential task scheduling data (depending on patients who have already been prescribed by doctors but have not completed them and are not in queue, as well as patients who may be added later to complete the task), which makes the prediction more accurate. In designing the objective function, the focus is not simply on the shortest time or the shortest path, but on a comprehensive approach. When prioritizing tasks, timeliness is paramount because the earlier the timeliness prediction is made, the more accurate it is. As the prediction time progresses, the error in the timeliness prediction gradually accumulates. Therefore, in planning tasks that are scheduled to be completed later, the weight of the path component is gradually increased to reduce the patient's travel burden. Ultimately, a suitable recommended medical plan is generated, providing patients with a recommended medical process, path, and estimated time consumption, minimizing queuing time, providing relevant information, and effectively improving the patient's medical experience.
[0069] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.
[0070] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.
Claims
1. A method for recommending in-hospital patient treatment plans, characterized in that, The hospital's integrated management system includes a server, doctor terminals, task terminals, and service terminals. The method is applied to the server and includes: The system retrieves patient information entered by the attending physician through the doctor's terminal and adds it to the patient information set for the day. Each patient information set includes the patient's basic information, unique ID, task information, and attending physician information. The system obtains the patient's query request through the service terminal. The query request includes the patient's unique number and the location information of the service terminal. Based on the patient's unique number, the patient's target task information is determined, and the target location and target task terminal corresponding to each task in the target task information are obtained. The target task scheduling data is obtained through the target task terminal. Based on the target location and target task scheduling data for each task, a recommended medical treatment plan is generated. The recommended medical treatment plan includes a recommended medical treatment process, a recommended medical treatment route, and an estimated medical treatment time. The system sends a recommended medical treatment plan back to the service terminal that sent the query request, so that the service terminal can display the recommended medical treatment plan to the patient.
2. The method for recommending in-hospital patient treatment plans according to claim 1, characterized in that, The system obtains the patient's query request through the service terminal, including: The unique number entered by the patient is obtained through the service terminal, and a query request is generated. Alternatively, a query request can be generated by scanning the QR code on the patient's receipt using a service terminal.
3. The method for recommending in-hospital patient treatment plans according to claim 1, characterized in that, Based on the patient's unique identification number, the patient's target task information is determined, and the target location and target task terminal corresponding to each task in the target task information are obtained. The target task scheduling data is then retrieved through the target task terminal, including: Based on the patient's unique number, the patient's target task information is determined. The target task information includes at least one task, each task corresponds to a target location, and each target location is equipped with at least one target task terminal. Obtain the target location and target task terminal for each task in the target task information; Obtain the target task scheduling data for each target task terminal. The target task scheduling data includes the completed scheduling data and real-time scheduling data of the target task on the current day.
4. The method for recommending in-hospital patient treatment plans according to claim 3, characterized in that, Based on the target location and task scheduling data for each task, a recommended medical visit plan is generated, including: Based on task information and each target location, we perform task problem modeling, path problem modeling, and dynamic time consumption modeling. Generate the objective function to be optimized; Generate constraints based on completed and real-time order data; Find the optimal solution for the objective function, and determine the optimal solution as the recommended medical treatment plan.
5. The method for recommending in-hospital patient treatment plans according to claim 4, characterized in that, Based on task information and each target location, task problem modeling, path problem modeling, and dynamic time consumption modeling are performed, including: Construct a patient task set based on task information. , This refers to the number of tasks the patient needs to complete. For the first Task ; Based on location information, target location, and attending physician information, a location set is constructed. ,in, For the patient's location information, For the first Task target location The location of the attending physician; Based on patient task set and location set Construct path sequence ,in, , , For the first in the path sequence One of the points along the way is one of the target locations; Define arrival time series ,in, At the current time point, For the patient's route to the destination The point in time, For the patient's route to the destination The point in time, This refers to the time at which the patient arrives at the location of the attending physician.
6. The in-hospital patient visit plan recommendation method according to claim 5, characterized in that, Generate the objective function to be optimized, including: Based on path sequence and arrival time series Generate the objective function to be optimized. : ， in, Let be the objective function to be optimized. This is the equivalent optimization coefficient for path distance per unit time. For the first Item weight coefficient, , Indicates the route from the points along the way to the places along the way The shortest path.
7. The method for recommending in-hospital patient treatment plans according to claim 6, characterized in that, Weighting coefficient satisfy: 。 8. The method for recommending in-hospital patient treatment plans according to claim 6, characterized in that, Based on completed order scheduling data and real-time order scheduling data, generate constraints, including: Based on location set Generate the first constraint condition; Based on the patient information set, completed scheduling data, and real-time scheduling data, potential scheduling data was identified. Based on completed order data, real-time order data, and potential order data, generate the second constraint. Based on path sequence Generate the third constraint condition.
9. The method for recommending in-hospital patient treatment plans according to claim 8, characterized in that, The first constraint is: path sequence In For the set of locations All permutations, For the first in the path sequence Each point along the route is one of the target locations, indicating that the execution order of each task is variable; The second constraint is: ， ， in, For the patient's route to the destination The point in time, Patient arrives at the location along the way The point in time, Indicates the patient's route through locations to the places along the way The duration of movement, Indicates the patient's location along the route The required length of stay, Indicates the places along the way The initial queue length at the location, Locations along the route The number of new employees per unit time at the location Locations along the route The number of tasks processed per unit time at the location. Locations along the route Execution time of a single task at the location; The third constraint is: ,in, This indicates a group of tasks with dependencies. For a set of dependencies, Indicates task The target location In path sequence The index number in the middle, Indicates task The target location In path sequence The index number in the database.
10. The method for recommending in-hospital patient treatment plans according to claim 5, characterized in that, Find the optimal solution for the objective function, and determine the optimal solution as the recommended medical treatment plan, including: Obtain patient location information The location of the attending physician Dependency set Each location along the route Initial queue length at New number of people per unit time Number of tasks processed per unit time Execution time of a single task The movement time matrix between each route and the pre-stored location; Generate candidate task sequences that satisfy all constraints, construct a corresponding path sequence for each candidate task sequence, and calculate the corresponding arrival time sequence. Calculate the objective function value for each candidate task sequence, and select the candidate task sequence with the smallest objective function value as the optimal solution; Based on the optimal solution, a recommended medical treatment plan is generated, which includes a recommended medical treatment process, a recommended medical treatment route, and an estimated medical treatment time.