A medical treatment route generation method, device, equipment and storage medium

Abstract

This application provides a method, apparatus, device, and storage medium for generating a patient visit route, relating to the field of smart healthcare technology. The solution includes: predicting the required testing time for each patient in each department based on patient information, the test information for each patient in that department, and the doctor's information; obtaining the total waiting time for each department based on the testing time; generating candidate routes based on the total waiting time for each department, the location of each department, and the testing constraints between the target patient's test items; predicting the total waiting time for the target patient to undergo testing according to each candidate route in each department based on the total waiting time for each department; and determining the target patient's designated visit route from the candidate routes based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's visit preferences. This solution can improve the patient's medical experience.

Description

Technical Field

[0001] This application relates to the field of smart healthcare technology, and in particular to a method, apparatus, device, and storage medium for generating a patient visit route. Background Technology

[0002] When patients visit a hospital, doctors may order various types of tests, such as complete blood count, CT scans, and ultrasounds. Different types of tests need to be performed in different laboratories, requiring patients to visit each department to complete the required tests.

[0003] Currently, when doctors prescribe multiple types of tests for patients, patients can only rely on a rough estimate of the waiting times at each laboratory department obtained from the hospital and their own experience to go to each department one by one for the tests. However, since most patients are not very familiar with the locations of the various laboratories and lack sufficient understanding of the characteristics of different tests, these factors combined can lead to long waiting times for multiple types of tests, resulting in even longer completion times for all tests and severely impacting the patient's medical experience. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, device, and storage medium for generating patient visit routes, so as to reduce idle waiting caused by blind queuing or improper ordering, and improve the patient's medical experience. The specific technical solution is as follows:

[0005] In a first aspect, embodiments of this application provide a method for generating a medical visit route, the method comprising:

[0006] For each target department, based on the patient information of each queued patient in that target department, the test information of each queued patient in that target department, and the doctor information of the department, the required testing time for each queued patient in that target department is predicted, and the total waiting time for that target department is obtained based on the testing time of each test; each target department is: the department that performs the tests for each category of test items for the target patients;

[0007] Based on the total queuing time of each target department, the location of each target department, and the testing constraints between the various categories of tests to be tested for the target patient, candidate routes for the target patient to undergo each category of tests are generated.

[0008] For each candidate route, based on the total queuing time in each target department, predict the total waiting time of the target patient in each target department when undergoing examination according to the candidate route;

[0009] Based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the patient's medical preferences, the patient's medical route is determined from the candidate routes.

[0010] Secondly, embodiments of this application provide a medical appointment route generation device, the device comprising:

[0011] The first prediction module is used to predict the required testing time for each queued patient in each target department based on the patient information, the test information of each queued patient in the target department, and the doctor information of the department. Based on the testing time, the total queuing time of the target department is obtained. Each target department is the department that conducts tests on each category of test items for the target patients.

[0012] The generation module is used to generate candidate routes for the target patient to perform each category of tests based on the total queuing time of each target department, the location of each target department, and the test constraints between each category of tests to be performed by the target patient.

[0013] The second prediction module is used to predict the total waiting time of the target patient in each target department according to the candidate route, based on the total queuing time of each target department.

[0014] The first determining module is used to determine the target patient's medical route from among the candidate routes based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's medical preferences.

[0015] Thirdly, embodiments of this application provide an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0016] Memory, used to store computer programs;

[0017] When a processor executes a program stored in memory, it implements the method described in the first aspect above.

[0018] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0019] Beneficial effects of the embodiments in this application:

[0020] The solution provided in this application, for each target department, utilizes patient information, test information for each patient in the target department, and physician information to predict the required testing time for each patient in that department. This method integrates multiple factors to predict the required testing time, thereby improving the accuracy of the predicted testing time. This results in a more accurate total waiting time for each target department. Based on the total waiting time for each target department, the location of each target department, and the testing constraints between different categories of test items for each patient, candidate routes that meet the testing constraints can be selected from various combinations of testing sequences for each target department. Then, based on the accurately predicted total waiting time for each target department, the waiting time for patients to arrive at each target department according to each candidate route can be accurately predicted, thus obtaining the total waiting time corresponding to each candidate route. By combining the total waiting time of each candidate route, the distance of each candidate route, and the patient's preferences, a suitable route can be accurately determined from among the candidate routes, thereby improving the user's medical experience. Furthermore, because the total waiting time for each candidate route can be predicted more accurately, determining the route by comprehensively considering the time cost, distance cost, and patient preferences can reduce idle waiting caused by blind queuing or improper ordering, ultimately achieving a shorter completion time for the medical visit than decision-making based on experience.

[0021] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0023] Figure 1 This is a timeline diagram of a hospital visit in the prior art;

[0024] Figure 2 A flowchart of a candidate route generation method provided in an embodiment of this application;

[0025] Figure 3 Another flowchart of the candidate route generation method provided in the embodiments of this application;

[0026] Figure 4 A flowchart for calculating the optimal check-in time is provided as an embodiment of this application;

[0027] Figure 5 A flowchart for generating candidate routes is provided as an embodiment of this application;

[0028] Figure 6 A flowchart for determining a medical treatment route is provided as an embodiment of this application;

[0029] Figure 7A A flowchart for determining the medical route of a target patient is provided in an embodiment of this application;

[0030] Figure 7B Another flowchart of the candidate route generation method provided in the embodiments of this application;

[0031] Figure 8 A timeline diagram of a specific example of the candidate route generation method provided in an embodiment of this application;

[0032] Figure 9 A system architecture diagram of an IMCS provided for an embodiment of this application;

[0033] Figure 10 A schematic diagram of a system use case for IMCS provided in an embodiment of this application;

[0034] Figure 11 A schematic diagram of the functional modules of an IMCS provided in an embodiment of this application;

[0035] Figure 12 A schematic diagram of a time prediction model provided in an embodiment of this application;

[0036] Figure 13 A schematic diagram of a diagnostic and treatment optimization model provided in an embodiment of this application;

[0037] Figure 14 A flowchart of the operation of an IMCS system monitoring module provided in an embodiment of this application;

[0038] Figure 15 A flowchart of the inspection status management module of IMCS provided for an embodiment of this application;

[0039] Figure 16 A flowchart of the operation of an intelligent planning module of IMCS provided for an embodiment of this application;

[0040] Figure 17 A flowchart of the intelligent check-in module of IMCS provided for embodiments of this application;

[0041] Figure 18 This is a schematic diagram of the structure of a candidate route generation device provided in an embodiment of this application;

[0042] Figure 19A block diagram of an electronic device for implementing the candidate route generation method provided in the embodiments of this application. Detailed Implementation

[0043] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0044] Currently, the process of patients seeking medical treatment at the hospital is as follows: Figure 1 As shown in the diagram, the doctor first orders tests for the patient, such as blood tests, ultrasound, and CT scans. These tests are stored in a list of pending tests. When the list is not empty, the patient selects the desired test and proceeds to the corresponding laboratory. Upon arrival at the laboratory, the patient checks in using the self-service check-in machine. This check-in information is then synchronized to the outpatient system and displayed on a large screen at the laboratory entrance. The patient waits for their number to be called by the laboratory's queuing system. Once the system calls, the patient enters the laboratory to complete the test. After the test is performed, the test is removed from the list. The patient can then view the remaining tests in the list and return to the previous steps to select the desired test, repeating this process until all tests are completed, at which point the loop ends. Once all tests are completed, the patient returns to the doctor for a follow-up visit.

[0045] It is evident that when patients need to complete multiple tests, they can only rely on personal experience to decide which laboratory to visit first. Since most patients are not very familiar with the locations of various laboratories and lack sufficient understanding of the characteristics of different tests, improper planning can lead to long waiting times for multiple tests or unnecessary back-and-forth movement within the hospital, thus increasing the overall consultation time.

[0046] To address the aforementioned problems, embodiments of this application provide a method for generating a medical appointment route, such as... Figure 2As shown, the process includes: S201, for each target department, based on the patient information of each queued patient in that target department, the item information of each queued patient's test items in that target department, and the department doctor information, predicting the required testing time for each queued patient in that target department, and obtaining the total queuing time for that target department based on each testing time; each target department is: the department that performs tests on each category of test items for the target patient; S202, based on the total queuing time of each target department, the location of each target department, and the testing constraints between each category of test items for the target patient, generating candidate routes for each category of test items for the target patient; S203, for each candidate route, based on the total queuing time of each target department, predicting the total waiting time for the target patient to undergo testing according to the candidate route in each target department; S204, based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's medical preferences, determining the target patient's medical route from each candidate route.

[0047] The solution provided in this application, for each target department, utilizes patient information, test information for each patient in the target department, and physician information to predict the required testing time for each patient in that department. This method integrates multiple factors to predict the required testing time, thereby improving the accuracy of the predicted testing time. This results in a more accurate total waiting time for each target department. Based on the total waiting time for each target department, the location of each target department, and the testing constraints between different categories of test items for each patient, candidate routes that meet the testing constraints can be selected from various combinations of testing sequences for each target department. Then, based on the accurately predicted total waiting time for each target department, the waiting time for patients to arrive at each target department according to each candidate route can be accurately predicted, thus obtaining the total waiting time corresponding to each candidate route. By combining the total waiting time of each candidate route, the distance of each candidate route, and the patient's preferences, a suitable route can be accurately determined from among the candidate routes, thereby improving the user's medical experience. Furthermore, because the total waiting time for each candidate route can be predicted more accurately, determining the route by comprehensively considering the time cost, distance cost, and patient preferences can reduce idle waiting caused by blind queuing or improper ordering, ultimately achieving a shorter completion time for the medical visit than decision-making based on experience.

[0048] The method for generating medical routes provided in this application can be applied to various electronic devices, such as personal computers, servers, and other devices with data processing capabilities. Furthermore, it is understood that the method for generating medical routes provided in this application can be implemented through software, hardware, or a combination of both.

[0049] In step S201, the categories of tests to be performed on the target patient are ordered by the doctor for the target patient. These categories may include complete blood count, ultrasound, and CT scans, etc. The target patient can be any patient generated from the proposed medical route.

[0050] For example, if the target patient's tests include complete blood count, ultrasound, and CT scan, then the target departments are the blood collection room, ultrasound room, and CT scan room. If the target patient's tests include electrocardiogram (ECG) and ultrasound scan, then the target departments are the ECG room and ultrasound room.

[0051] In this embodiment, patient information, test information for each patient in each target department, and physician information can be obtained from the hospital system. All operations involving the acquisition, storage, use, processing, transmission, provision, and disclosure of patient information, test information, and physician information in this embodiment are performed with the appropriate authorization obtained.

[0052] For example, patient information may include age and / or past medical history. The test details for each queued patient in each target department may include the specific test items. For instance, if the test is an ultrasound, the specific test items may be abdominal color Doppler ultrasound or 4D ultrasound; if the test is a CT scan, the specific test items may be head CT, chest CT, etc. The department doctor information may include the doctor's seniority and / or operating habits.

[0053] Understandably, factors such as a patient's age and medical history can influence the duration of their testing. For example, older patients often require longer testing times due to mobility issues, and patients with pre-existing medical conditions typically require more detailed examinations, also necessitating longer testing times. Therefore, patient information such as age and medical history can be used as factors in predicting the duration of a patient's testing.

[0054] Because the testing time varies for different categories of tests, for example, ultrasound usually takes longer than blood tests; and the testing time for different sub-items within the same category also varies, for example, 4D ultrasound usually takes longer than regular ultrasound, and enhanced CT scan takes longer than plain CT scan. Therefore, the specific sub-items of the tests can be used as influencing factors to predict the patient's testing time.

[0055] Since senior doctors in a department tend to have more experience and can perform tests on patients more quickly, and different doctors have different operating habits, this can also affect the testing speed. Therefore, information such as the seniority and operating habits of doctors in a department can be used as factors in predicting the testing time of patients. For example, in practical applications, the average testing time required by different doctors for each patient over the years can be statistically analyzed to represent the doctors' operating habits.

[0056] Since patient information, test item information, and department doctor information all affect the required testing time for a patient, for each target department, by using the patient information of each queued patient in that target department, the test item information of each queued patient in that target department, and the department doctor information, the required testing time for each queued patient in that target department can be predicted. This can comprehensively consider multiple dimensions of factors to predict the required testing time for patients, thereby improving the accuracy of the predicted testing time.

[0057] For each target department, after predicting the required testing time for each patient in the queue, the predicted testing times are summed to obtain the total waiting time for that target department. For example, if the target department is the ultrasound room, and there are currently 10 patients in the queue, then based on the patient information, the test information for each of the 10 patients in the ultrasound room, and the information of the doctor in the ultrasound room, the required testing time for each of the 10 patients in the target department is predicted. Then, the predicted testing times are summed to obtain the total waiting time for the ultrasound room.

[0058] For example, in one implementation, a cluster analysis-based method can be used to predict test duration. Specifically, the patient information, the test items to be tested for each historical patient in the target department, and the doctor's information can be clustered to divide patients into different groups. The actual test duration of historical patients in the target department within each group will form a distribution. Then, for new patients, the existing cluster can be determined based on the patient information, the test items to be tested for the patient in the target department, and the doctor's information. The median, mean, or specific quantile (such as P80) of the test duration within that cluster can then be used as the predicted test duration.

[0059] For example, in another implementation, the patient information of the queued patient, the item information of the items to be tested for the queued patient in the target department, and the information of the doctor in the target department can be input into a pre-trained time prediction model to obtain the model output the required testing time for the queued patient in the target department. Specifically, training samples can be constructed based on the patient information of each historical queued patient in each department, the item information of the items to be tested for each historical queued patient in that department, and the information of the doctor in that department. The initial time prediction model is trained until the model converges or reaches a preset number of iterations, resulting in a trained time prediction model. During training, the label of each training sample is the actual testing time of each historical queued patient in that department. For example, the initial time prediction model can use CNN (Convolutional Neural Network) or DNN (Convolutional Neural Network), etc.

[0060] To ensure clarity of the solution, the specific methods for predicting the test duration will be described in the following embodiments, and will not be repeated here.

[0061] For step S202, for example, the location of each target department includes latitude and longitude coordinates and floor level. In practical applications, a hospital map can be obtained, and the location of each target department can be determined based on the hospital map. The testing constraints between different categories of items to be tested can include dependencies between items and / or the available testing time periods for each item. For example, imaging examinations such as CT scans usually require blood tests first, so the dependencies between items can include CT scans depending on blood tests. Due to the hospital's scheduling of doctors and the availability of equipment in each department, the service hours of different departments in the hospital vary. Therefore, when planning a patient's route, the available testing time periods for each item also need to be considered.

[0062] For example, the testing constraints between different categories of items to be tested can be set by relevant technical personnel based on their experience. In practical applications, basic medical knowledge related to the items to be tested can be incorporated into the testing constraints. For example, blood tests need to be performed before enhanced CT scans, or abdominal ultrasounds need to be performed before gastrointestinal endoscopy, etc., so that the generated candidate routes conform to basic medical knowledge.

[0063] For example, in one implementation, the required travel time for a patient to travel between target departments can be calculated based on the location of each target department and the preset patient movement speed. Correspondingly, the required travel time for the patient to reach each target department can be calculated based on the patient's current location and the location of each target department. Thus, the target departments can be arranged and combined in different orders to obtain multiple initial routes. Then, based on the travel time of the target patient to each target department according to each initial route and the total queuing time of each target department, the predicted consultation time of the target patient in each target department is determined. Then, based on the predicted consultation time of the target patient in each target department, initial routes that satisfy the testing constraints between each category of tests are determined as candidate routes for the target patient to perform each category of tests. For clarity, this implementation is described in detail in the following embodiments and will not be repeated here. The current location of the target patient may include the latitude and longitude coordinates of the target patient obtained using automatic positioning technology and the current floor actively reported by the patient.

[0064] For example, in another implementation, the order of some departments can be fixed based on the testing constraints between different categories of items to be tested (e.g., blood must be drawn before urine test). When constructing candidate routes, starting from the current position, only departments that "must be ahead of it in the queue have been assigned" can be selected each time. Then, from these selectable departments, they are added to the route in ascending order of "the current queuing time of the department plus the travel time from the current position to the department," and the current position and cumulative time are updated. This process is repeated until all departments are assigned, thus obtaining a candidate route that satisfies all fixed order constraints. To obtain multiple candidate routes, the above process can be repeated multiple times, each time randomly sorting the selectable departments (or randomly swapping adjacent departments without violating constraints) to generate different candidate routes.

[0065] Regarding step S203, it is understood that since the waiting time of the target patient in the previous target department affects the arrival time of the patient in the next target department, thus affecting the waiting time of the patient in the next target department, and the waiting time of the target patient in the previous target department is related to the total queuing time of the previous target department, the total waiting time of the target patient in each target department can be predicted according to the total queuing time of each target department when performing the test according to the candidate route.

[0066] For example, in one implementation, it can be assumed that when the patient arrives at each department, the remaining queuing time of that department remains unchanged. Then, the total queuing time of each department in each candidate route can be directly added together to obtain the total waiting time of the target patient in each target department when the test is performed according to the candidate route.

[0067] For example, in another implementation, the time taken for a target patient to reach each target department can be predicted based on the total queuing time for each target department. The time taken for a patient to reach each target department is the time between the current moment and the arrival time of the target patient in that target department. Then, based on the total queuing time corresponding to each target department and the time taken for the patient to reach that target department, the waiting time for the target patient in each target department is determined. Finally, the sum of the waiting times for the target patient in each target department is calculated to obtain the total waiting time for the target patient in all target departments. For clarity of the solution layout, this implementation method is described in detail in the following embodiments, and will not be repeated here.

[0068] For step S204, for example, the location of each target department on each candidate route can be determined based on the obtained hospital map, and then the shortest route between each department can be calculated. The distance of each candidate route can be calculated.

[0069] For example, patient preferences may include time priority, distance priority, or custom preferences. For example, a target patient's custom preferences may include "taking the elevator," or the order in which two or more tests are scheduled, etc.

[0070] For example, in one implementation, for each candidate route, the total waiting time and distance corresponding to that route can be normalized, and then weighted according to the target patient's medical preferences. For instance, if the medical preference is time-priority, the weight of the total waiting time is set to 0.8–0.9, and the weight of the distance is set to 0.2–0.1; the opposite is true if the preference is distance-priority. Next, the weighted sum of the total waiting time and distance corresponding to each candidate route is calculated as the total cost, and the route with the minimum total cost is selected as the target patient's medical route.

[0071] If the patient's medical preferences are customized, then the weights for total waiting time and distance can both be set to 0.5. The candidate route that best matches the customized preferences and has the lowest total cost is then selected as the target patient's route. For example, if the candidate routes include three routes (A, B, and C), and the customized medical preference is "to take the vertical elevator," then if routes B and C match the customized preferences, and route B has the lowest total cost, then route B is selected as the target patient's route.

[0072] For example, in another implementation, if the patient's preference is time-priority, then all routes with the shortest total waiting time are first selected. If there are multiple candidate routes with the shortest total waiting time, then the route with the shortest distance is selected as the patient's route. If the patient's preference is distance-priority, then all routes with the shortest distance are first selected. If there are multiple routes, then the route with the shortest total waiting time is selected as the patient's route. This optimizes the secondary objective while satisfying the primary objective.

[0073] Understandably, selecting the final medical route based on the total waiting time, distance, and the patient's preferences for each candidate route allows for personalized decision-making, taking into account the patient's priorities regarding time or distance (e.g., elderly people may prefer shorter distances and less walking, while working people value shorter total waiting times), thus improving the medical experience. By comprehensively considering both total waiting time and distance, it avoids the situation where simply pursuing the shortest distance might lead to long queues in popular departments, while simply pursuing the shortest waiting time might cause patients to run back and forth between floors. Combining the two can find a more balanced and efficient route.

[0074] Alternatively, in another embodiment of this application, in Figure 2 Based on the illustrated embodiments, as Figure 3 As shown, the above-mentioned method for generating a medical route may further include: S301, after the first test item for the target patient begins, returning to each target department, based on the patient information of each queued patient in that target department, the test item information of each queued patient in that target department, and the doctor information of the department, predicting the test time required for each queued patient in that target department, and obtaining the total queuing time of the target department based on each test time, and redetermining the medical route to obtain the current medical route for the target patient; S302, based on the current medical route of the target patient, determining the second test item for the target patient to sign in; S303, obtaining the remaining test time of the first test item; S304, based on the target patient's current location, remaining test time, and the total queuing time of the department to be signed in, calculating the optimal sign-in time for the second test item, executing the sign-in task based on the optimal sign-in time to sign in for the target patient in the department to be signed in, wherein the sign-in task is triggered by the target patient's prompt information based on the optimal sign-in time; the department to be signed in is: the target department for the test of the second test item.

[0075] Understandably, after the first test begins, the optimal check-in time for the second test is calculated based on the remaining test time, the target patient's current location, and the total waiting time in the departments to be checked in. The check-in process is then triggered by a prompt message from the target patient based on this optimal check-in time. This guides the target patient to check in at the optimal time, eliminating the need for on-site check-in at each department. This allows patients to proceed more conveniently to the next department, reducing time costs and improving the overall patient experience.

[0076] For step S301, the first item is the item that the target patient is currently to be tested. After the first item is tested, the patient can return to step S201 to redetermine the medical route according to the above steps S201-S204, and use the newly determined medical route as the current medical route, so as to dynamically update the medical route according to the real-time queuing situation of each department.

[0077] In step S302, based on the current patient route, the second item to be checked in for the target patient is determined. The second item is the next item to be tested after the first item, which is the first item to be tested on the current patient route.

[0078] Regarding step S303, the required testing time for the first item can be predicted based on the target patient's patient information, the item information of the first item, and the physician information of the target department where the first item is being tested. The method for predicting the testing time can be referred to the relevant description in step S201 above, and will not be repeated here. After obtaining the required testing time for the first item, the remaining testing time for the first item is obtained based on the required testing time for the first item and the patient's already completed testing time for the first item.

[0079] For example, the status of each test item for the target patient can be monitored. If the status of the first item changes from "not yet tested" to "under testing," a timer can be started to obtain the elapsed testing time for the first item. Then, the difference between the required testing time for the first item and the elapsed testing time is calculated to obtain the remaining testing time for the first item.

[0080] Regarding step S304, the optimal check-in time is the check-in time that minimizes the waiting time for the target patient after arriving at the department to be checked in. It is understandable that since the remaining testing time for the first item determines the earliest the target patient can "escape" to the department to be checked in, the travel time required for the target patient to reach the department from their current location can be calculated based on their current location. The total queuing time of the department to be checked in affects the waiting time from check-in to being called. Therefore, considering these three factors, the optimal check-in time for the second item can be calculated, ensuring that the moment the patient completes the first item's testing and arrives at the department to be checked in is close to the time they are at the front of the queue (or just when it's their turn), thus achieving the ideal state of "arrival and immediate testing." For clarity, the specific implementation method for calculating the optimal check-in time will be described in the following embodiments, and will not be repeated here.

[0081] For example, the electronic device executing the method provided in this embodiment can send a prompt message to the target patient's user device when the optimal check-in time is reached, indicating that the current check-in time is optimal. The target patient then triggers check-in automatically through their user device based on the prompt message. For instance, upon reaching the optimal check-in time, the target patient's user device displays a prompt message, and the check-in control on the user device's front-end interface is enabled (the check-in control is inactive before the optimal check-in time is reached), allowing the user to check in automatically at any time thereafter by clicking the check-in control. After the target patient triggers the check-in operation through the check-in control on their user device, the electronic device executing the method provided in this embodiment, in response to the check-in operation triggered by the target patient on their user device, can send a check-in instruction to the self-service check-in machine in the department to be checked in, so that the self-service check-in machine completes the target patient's check-in upon receiving the instruction; alternatively, it can request the back-end management platform of the queuing system to complete the target patient's check-in in the department to be checked in, with the platform controlling the speakers and display screen in the department to call the number – both of these are reasonable.

[0082] Understandably, by sending a reminder message to the target patient at the optimal check-in time, the target patient can choose the check-in time according to the reminder message. In this way, the target patient can flexibly check in according to their own needs, making the user experience more user-friendly.

[0083] Alternatively, in another embodiment of this application, such as Figure 4As shown, step S304 above calculates the optimal check-in time for the second item based on the target patient's current location, remaining testing time, and total queuing time of the department to be checked in. This includes: S3041, calculating the travel time required for the target patient to travel from the current location to the department to be checked in; S3042, predicting the arrival time of the target patient to the department to be checked in based on the travel time and remaining testing time; S3043, if the arrival time is not later than a specified time, determining the current time as the optimal check-in time for the second item; wherein, the specified time is: the time after the current time is shifted backward by the total queuing time of the department to be checked in; S3044, if the arrival time is later than the specified time, returning to the step of obtaining the remaining testing time of the first item according to a predetermined time interval.

[0084] Understandably, if the predicted arrival time of the target patient at the designated department is no later than the specified time, immediate check-in allows them to join the waiting queue as quickly as possible. This improves patient time utilization and thus increases overall efficiency compared to arriving at the department and then checking in. If the arrival time is later than the specified time, the process of retrieving the remaining test time for the first item at predetermined time intervals and recalculating the optimal check-in time based on the re-obtained remaining test time reduces the likelihood of patients arriving too late and missing their turn, achieving the ideal state of "testing upon arrival," further improving patient time utilization and overall efficiency.

[0085] For step S3041, the shortest route between the target patient's current location and the department to be checked in can be determined based on the hospital map. Then, the length of the shortest route is divided by the preset patient movement speed (e.g., 0.5 m / s) to obtain the travel time required for the target patient to travel from the current location to the department to be checked in.

[0086] Regarding step S3042, since the target patient needs to complete the first test before proceeding to the designated department, the sum of the remaining testing time for the first test and the travel time calculated in step S3041 can be calculated. The current time can then be delayed by this sum to obtain the target patient's arrival time at the designated department. Alternatively, considering that the target patient needs rest and adjustment after completing the current test, the current time can be delayed by this sum and a preset time to obtain the target patient's arrival time at the designated department. Both are reasonable. For example, the preset time could be 2 minutes or 3 minutes. In practical applications, the preset time can be set by relevant technical personnel based on experience, or it can be based on historical data to calculate the average rest time for patients between two departments.

[0087] For step S3043, the time after shifting the current time by the total queuing time of the department to be checked in is taken as the designated time. If the predicted arrival time of the target patient in the department to be checked in is not later than the designated time, it means that after the target patient arrives at the department to be checked in, the total queuing time of the department to be checked in has not yet been exhausted. That is, the target patient can arrive at the department to be checked in before the total queuing time of the department to be checked in is exhausted. Therefore, checking in at this time allows the target patient to join the queue as soon as possible. Thus, the current time is determined as the optimal check-in time for the second item. After checking in, the patient can go to the department to be checked in while waiting in the queue, thereby improving the patient's time utilization and shortening the patient's consultation time.

[0088] Regarding step S3044, if the predicted arrival time is later than the specified time, it means that the target patient has not arrived at the department to be checked in after the total queuing time for the department has been exhausted. Checking in at this time would result in missing the required number. In this case, after the predetermined time interval has elapsed, the process returns to the step of obtaining the remaining testing time for the first item and the total queuing time for the department to be checked in, and the optimal check-in time is recalculated. In this way, it is possible to detect whether the current time is the optimal check-in time according to the predetermined time interval. By completing the check-in at the optimal check-in time, the situation of patients arriving too late and missing their required number can be reduced, achieving the ideal state of "testing upon arrival" and improving the patient's time utilization rate.

[0089] For example, the predetermined time interval could be 1 minute or 2 minutes, etc.

[0090] Alternatively, in another embodiment of this application, such as Figure 5 As shown, in step S202 above, candidate routes for the target patient to undergo various types of tests are generated based on the total queuing time of each target department, the location of each target department, and the testing constraints between the target patient's various types of tests. This includes: S2021, for each initial route that leads to each target department in different orders, based on the target patient's current location, the location of each target department, the current time, the total queuing time of each target department, and hospital information, the time of the target patient's journey to each target department according to the initial route is estimated to obtain the predicted consultation time of the target patient in each target department; S2022, based on the predicted consultation time corresponding to each target department in each initial route, candidate routes that satisfy the testing constraints between the target patient's various types of tests are determined from each initial route.

[0091] Understandably, by extrapolating the time it takes for target patients to travel to each target department along each initial route, it is possible to accurately predict the actual arrival time of the target patients at each target department, as well as the predicted consultation time at each target department. After extrapolation, combined with testing constraints (such as sequential completion or time window limitations), routes that are time-infeasible (e.g., unable to complete a certain examination within a specified time window) or that would cause serious conflicts can be eliminated, thereby improving the feasibility of candidate routes.

[0092] Regarding step S2021, the target departments can be arranged and combined in different orders to obtain each initial route to the target departments in different orders. For example, if the target departments include departments A, B, and C, then the target departments can be arranged and combined in different orders to obtain 6 initial routes, namely: A->B->C, A->C->B, B->A->C, B->C->A, C->A->B, and C->B->A.

[0093] After obtaining the initial routes, all available information can be integrated. Then, for each initial route, a timeline is generated based on the integrated information to predict key time points for the target patient's visit along that route. All available information includes the target patient's current location, the location of each target department, the current time, the total waiting time for each target department, and hospital information. For example, hospital information may include the hospital's ranking and / or peak hours.

[0094] Understandably, the number of patients at a hospital is affected by its level; generally, higher-level hospitals see more patients, leading to greater congestion and increased travel time to different departments. Furthermore, hospitals are relatively congested during peak hours, further increasing travel time to specific departments compared to off-peak times. For example, for patients with mobility issues who require elevators, waiting times are significantly longer in congested hospitals, thus greatly increasing their overall travel time.

[0095] In this embodiment, the travel time from the target patient's current location to each target department, and the travel time from one target department to another are determined based on the target patient's current location, the location of each target department, and hospital information.

[0096] For example, a congestion reduction factor C of not less than 1 can be determined based on hospital information. For instance, if the hospital information only includes hospital level, the higher the level, the larger the congestion reduction factor. For example, a Level 1 hospital might have C=1 (almost no additional reduction), a Level 2 hospital C=1.5, and a Level 3 hospital C=2. If the hospital information only includes peak hours, the peak hour congestion reduction factor can be set to be greater than the off-peak hour congestion reduction factor. For example, a peak hour congestion reduction factor C=2, and an off-peak hour congestion reduction factor C=1. If the hospital information includes both hospital level and peak hours, a factor C1 can be set separately for hospital level, and a factor C2 can be set separately for peak hours. The product of C1 and C2 can then be used as the final congestion reduction factor C.

[0097] For example, a hospital map can be obtained. Based on the hospital map, the shortest route length from the target patient's current location to each target department can be determined. Then, the quotient of this route length and a preset patient movement speed can be calculated as the ideal travel time from the target patient's current location to each target department. This calculated ideal travel time is then multiplied by a congestion reduction factor C determined based on hospital information to obtain the total travel time from the target patient's current location to each target department. Similarly, the shortest route length between two target departments can be determined based on the hospital map. The quotient of this route length and a preset patient movement speed can then be calculated as the ideal travel time from one target department to another. This ideal travel time is then multiplied by the congestion reduction factor C determined based on hospital information to obtain the total travel time from one target department to another. In this way, the travel time for a target patient to move between target departments, as well as the travel time from the current location to a specific target department, can be obtained.

[0098] After obtaining the travel times, the arrival time at the first department can be determined by shifting the current time forward by the travel time from the current location to the first department on the initial route. Then, based on the total queuing time at the first department, the waiting time after the target patient's arrival is determined, thus obtaining the predicted consultation time at the first department. For example, if the total queuing time at the first department is greater than the travel time from the current location to the first department on the initial route, the difference between the total queuing time and the travel time is determined as the target patient's waiting time at the first department. Shifting the arrival time at the first department forward by this waiting time yields the predicted consultation time at the first department.

[0099] Next, by shifting the predicted appointment time of the target patient in the first department forward by the duration of the patient's examination in the first department, we can obtain the end time of the patient's appointment in the first department. Then, by shifting the end time of the appointment forward by the travel time from the first department to the second department, we can obtain the patient's arrival time in the second department. Following this method sequentially, we can deduce the predicted appointment times of the target patient in each target department.

[0100] For example, time extrapolation can be performed by constructing a state space. Each extrapolated state is defined as a triple (current position, set of visited target departments, current time), with the initial state being (patient starting point, empty set, current time). Starting from any state, all unvisited target departments are enumerated as the next step. The remaining queue time at the current time is calculated by combining the total queue time of that department, thus extrapolating the start and end times of the examination. Simultaneously, the visited set is added to that department, forming a new state. All possible department visit sequences are traversed using breadth-first or depth-first search. The accumulated timeline information is recorded at each state transition. Ultimately, all states that have reached "all target departments visited" correspond to a complete initial route and its predicted visit time sequence, which can be used for subsequent screening of candidate routes that meet the testing constraints.

[0101] Regarding step S2022, after obtaining the predicted consultation time for each target department in each initial route, the initial routes can be screened based on whether each predicted consultation time meets the test constraints. That is, routes that do not meet the test constraints are removed from each initial route to obtain candidate routes.

[0102] For example, if the constraint includes that the ultrasound examination must be performed before 5 PM, then if department A is the ultrasound room and the predicted appointment time for department A on route 1 is 6 PM, then route 1 does not meet the constraint and needs to be eliminated. If the constraint includes that blood must be drawn before the CT scan, then if department B is the CT room and department C is the blood collection room, and department B is listed before department C on route 2, then the constraint is also not met and needs to be eliminated.

[0103] Alternatively, in another embodiment of this application, such as Figure 6As shown, step S204 above determines the target patient's medical route from the candidate routes based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's medical preferences. This includes: S2041, for each candidate route, determining the weight of the total waiting time and the weight of the distance corresponding to the candidate route according to the target patient's medical preferences, and calculating the weighted sum of the total waiting time and distance corresponding to the candidate route based on the determined weights to obtain the total cost of the candidate route; S2042, determining the target patient's medical route based on the candidate route with the lowest total cost among all candidate routes.

[0104] Understandably, selecting the final medical route based on the total waiting time, distance, and the patient's preferences for each candidate route allows for personalized decision-making, taking into account the patient's priorities regarding time or distance (e.g., elderly people may prefer shorter distances and less movement, while working professionals prioritize shorter total waiting times), thus improving the patient experience. Furthermore, by comprehensively considering both total waiting time and distance, it avoids the situation where simply pursuing the shortest distance might lead to long queues in popular departments, while simply pursuing the shortest waiting time might cause patients to run back and forth between floors. Combining these two approaches can result in a more balanced and efficient route.

[0105] The patient's preference for medical treatment can include time priority, distance priority, or a custom preference. If the preference is time priority, the weight of total waiting time is determined to be greater than the weight of distance; if the preference is distance priority, the weight of total waiting time is determined to be less than the weight of distance; if the preference is a balance between time and distance, the weight of total waiting time is determined to be equal to the weight of distance.

[0106] For example, in practical applications, a weight combination can be preset for each medical treatment preference. Each weight combination includes a weight for total waiting time and a weight for distance. For instance, in the weight combination corresponding to a time-priority medical treatment preference, the weight for total waiting time is 0.8 and the weight for distance is 0.2; in the weight combination corresponding to a distance-priority medical treatment preference, the weight for total waiting time is 0.3 and the weight for distance is 0.7.

[0107] After determining the weights of total waiting time and distance, the weighted sum of total waiting time and distance for each candidate route can be calculated to obtain the total cost of that candidate route. Then, based on the candidate route with the minimum total cost among all candidate routes, the target patient's medical route is determined. For example, in one implementation, the candidate route with the minimum total cost can be directly used as the target patient's medical route. For clarity of the solution layout, other methods for determining the target patient's medical route are described in the following embodiments.

[0108] In this embodiment, after a test begins, the process can return to steps S201-S204 to re-acquire various information, predict waiting times for each department, and re-determine the treatment route. Thus, after each test, the waiting time for the target department corresponding to subsequent tests can be re-predicted based on the current real-time information. Therefore, the latest predicted total waiting time is used each time the total cost of each candidate route is calculated, resulting in a more accurate total cost calculation.

[0109] Alternatively, in another embodiment of this application, such as Figure 7A As shown, step S2042 above, which determines the target patient's medical route based on the candidate route with the lowest total cost among all candidate routes, includes: S701, providing the target patient with a first candidate route with the lowest total cost among all candidate routes; S702, in response to the target patient's operation to adjust the order of visits to each target department in the first candidate route, obtaining an adjusted second candidate route; S703, obtaining the time cost and travel cost of the second candidate route, and feeding back the time cost and travel cost to the target patient so that the target patient can confirm the route based on the time cost and travel cost; S704, in response to the target patient's route confirmation instruction, confirming the indicated route as the target patient's medical route.

[0110] Understandably, by providing target patients with the first candidate route that has the lowest total cost among all candidate routes, target patients can confirm the planned first candidate route or flexibly adjust it according to their own needs. By providing feedback on the time and distance costs of the adjusted route, users can make choices that balance actual needs and costs, thereby improving the personalization of the final medical route and ensuring that the final determined medical route is closer to the user's needs, thus improving the user experience.

[0111] In step S701, the first candidate route with the lowest total cost among all candidate routes can be sent to the user device of the target patient. The user device displays the first candidate route on the interface for the target patient to confirm or adjust. For example, each visit node (i.e., the target department) on the first candidate route can be presented as a draggable card or icon. The target patient can press and hold and drag up and down to freely adjust their order.

[0112] Regarding step S702, if the target patient adjusts the order of visits to each target department in the first candidate route, the adjusted route is obtained as the second candidate route.

[0113] For step S703, the total waiting time corresponding to the second candidate route can be obtained by referring to the relevant description of step S203 above, and this time cost can be used as the second candidate route. The distance cost of the second candidate route is obtained by calculating the total distance. Then, the time cost and distance cost are sent to the target patient's user device so that the target patient can view the time cost and distance cost corresponding to the adjusted route, thereby helping the target patient to further adjust the route or directly confirm the route.

[0114] Regarding step S704, the target patient can confirm the displayed route on the user device. For example, a confirmation button for route confirmation can be displayed on the user device interface. When the target patient clicks the confirmation button, the electronic device executing the method provided in this application embodiment confirms the route displayed on the current user device interface as the target patient's medical appointment route.

[0115] Optionally, in another embodiment of this application, the required testing time for each queued patient in each target department is predicted through the following steps: A1, determining numerical and non-numerical parameters from the patient information of the queued patient, the item information of the items to be tested for the queued patient in the target department, and the information of the department doctor of the target department; A2, encoding the determined non-numerical parameters to obtain a first feature, and normalizing the determined numerical parameters to obtain a second feature; A3, concatenating the first and second features to obtain a feature vector; A4, predicting the required testing time for the queued patient in the target department based on the feature vector.

[0116] It is understandable that by extracting features from different dimensions of information, the original information can be transformed into numerical and scale-normalized features. The extracted first and second features are then concatenated to predict the test duration. This allows for the fusion of multi-dimensional features for prediction, thereby improving prediction accuracy.

[0117] For example, normalization methods for numerical parameters can include min-maximum normalization or Z-score standardization, while encoding methods for non-numerical parameters can include one-hot encoding or label encoding. It is understandable that normalization of numerical parameters can eliminate the influence of different units and orders of magnitude, making the features comparable. Encoding non-numerical parameters can transform non-numerical information into a numerical form that the algorithm can process, allowing the algorithm to directly utilize these encoded features for learning and analysis.

[0118] After obtaining the first and second features, they can be concatenated in a predetermined order to obtain a feature vector. This predetermined order can be set by relevant technical personnel based on experience. Understandably, this concatenation operation is simple and maintains the independence of each feature dimension, avoiding mutual interference or information loss between features, and providing a unified and complete feature representation for subsequent model inputs (such as neural networks, logistic regression, etc.).

[0119] For example, after obtaining the feature vector, it can be input into a pre-trained time prediction model to obtain the required testing time for the queued patient in the target department. The training process of the time prediction model can be as follows: Numerical and non-numerical parameters are determined from patient information (obtainable from historical patients), the test information for the sample patient in the target department, and the doctor information of the target department. The determined non-numerical parameters are encoded to obtain the first sample feature. The determined numerical parameters are normalized to obtain the second sample feature. The first and second sample features are concatenated to obtain the sample feature vector. This sample feature vector is input into the initial time prediction model to obtain the predicted testing time. The difference between the predicted testing time and the actual testing time of the sample patient is calculated to obtain the model loss value. The parameters of the time prediction model are adjusted by backpropagation to minimize the model loss value until the model converges, and training ends.

[0120] For example, an initial time prediction model can consist of fully connected layers, activation functions, and Dropout layers. Understandably, by introducing Dropout layers into the model, a portion of neurons and their connections can be randomly "dropped" in each training iteration, forcing the network to avoid over-reliance on co-adaptation relationships between specific neurons or features, thereby significantly reducing the risk of overfitting. Introducing the ReLU (Rectified Linear Unit) activation function into the model introduces non-linearity, improving the model's expressive power.

[0121] This application does not limit the method of predicting test duration based on feature vectors. For example, in one implementation, the K-nearest neighbor algorithm can also be used to predict test duration based on feature vectors. That is, the similarity (such as Euclidean distance or Manhattan distance) between the feature vector of the current patient and the feature vectors of all patients who have completed the test in the historical data is calculated, and the K most similar historical samples are selected. Then, the average or weighted average (weighted by the inverse of distance) of the actual test durations corresponding to these K samples is used as the predicted duration of the current patient.

[0122] Optionally, in another embodiment of this application, the step S203 above, which predicts the total waiting time of the target patient in each target department according to the candidate route for testing, includes: B1, predicting the time consumed by the target patient to reach each target department according to the candidate route based on the total queuing time of each target department and the location of each target department; B2, for each target department, if the consumption time corresponding to the target department is not greater than the total queuing time of the target department, then calculating the difference between the total queuing time of the target department and the consumption time corresponding to the target department to obtain the waiting time of the target department; if the consumption time corresponding to the target department is greater than the total queuing time of the target department, then determining the preset value as the waiting time of the target department; B3, calculating the sum of the waiting times of each target department to obtain the total waiting time of the target patient in each target department according to the candidate route for testing.

[0123] Understandably, by analyzing the location and total queuing time of each target department, the time it takes for a target patient to reach each target department can be predicted. Based on the relationship between the time it takes for a patient to reach each target department and the total queuing time of each target department, the waiting time for a patient in each target department can be accurately determined.

[0124] The implementation of step B1 may include: predicting the arrival time of the target patient to each target department according to the candidate route, referring to step S2021 above, and then using the time between the initial time (i.e. the time when the target patient starts to go to the first target department according to the candidate route) and the arrival time of the target patient to each target department as the time consumed by the target patient to each target department.

[0125] If the time taken for a target patient to reach any target department is no greater than the total queuing time for that department, the patient can be tested as soon as the total queuing time runs out. In this case, the difference between the total queuing time and the time taken to reach that department can be used as the patient's waiting time in that department. If the time taken for a target patient to reach any target department is greater than the total queuing time for that department, the patient can be tested as soon as they arrive at that department. In this case, a preset value can be used as the patient's waiting time in that department.

[0126] For example, the preset value could be 0, or the average service time for patients in the target department, which is also reasonable.

[0127] Alternatively, in another embodiment of this application, based on any of the above embodiments, such as Figure 7BAs shown, the above-mentioned method for generating a medical route may further include: S711, determining the next target department for the target patient based on the medical route and the status of each test item; S712, determining the navigation start time based on the waiting time of the target patient in the next target department and the travel time required for the target patient to travel from the current location to the next target department; S713, pushing a guidance map about the medical route to the target patient when the determined navigation start time is reached.

[0128] Understandably, by determining the navigation activation time based on the target patient's waiting time in the next target department and the travel time required for the target patient to reach the next target department from the current location, navigation can be activated at the appropriate time. This allows patients to begin their examinations almost immediately upon arriving at the next department after navigation is activated, thereby improving their medical experience.

[0129] For example, the status of each test item can include "Pending Check-in," "In Queue," "Under Examination," "Report Not Yet Available," "Completed," and so on. It can be understood that, among the target departments along the treatment route, the department corresponding to the first test item in the "Pending Check-in" status is the next target department the patient should visit. For instance, if the treatment route is A->B->C, and the test item in department A is in the "Completed" status, while the test items in departments B and C are both in the "Pending Check-in" status, then department B is the next target department the patient should visit.

[0130] In this embodiment, the waiting time for the target patient in the next target department is the waiting time for the target patient to be called after completing the check-in for the next target department.

[0131] For example, the waiting time for the target patient in the next target department can be obtained by referring to steps B1-B2 above. Then, the difference between the obtained waiting time in the next target department and the travel time required for the target patient to travel from the current location to the next target department is calculated. The time obtained by shifting this difference forward from the current time is determined as the navigation start time. In this way, when the patient arrives at the next department after navigation is activated, there is almost no extra waiting time before the examination can begin, thereby improving the patient's medical experience.

[0132] When the designated navigation start time is reached, a guide map of the medical route can be pushed to the target patient's user device, so that the target patient's user device can directly display the guide map and the target patient can go to the target department according to the guide map.

[0133] For example, the guidance map can be a dynamic navigation map that integrates augmented reality information, i.e., an AR real-world map based on the patient's real-time location and treatment route, overlaid with interactive elements such as virtual directional arrows, distance prompts, and landmarks. Specifically, upon navigation initiation, the user's device can be triggered to activate its camera and capture the surrounding environment's real-world view. Based on the captured surrounding environment's real-world view, the patient's real-time location within the hospital is determined. The treatment route is determined based on the patient's real-time location, and the route guidance is precisely overlaid onto the real-world view in the form of 3D virtual signs, ground arrows, or floating light strips, forming an AR navigation interface. The target patient can intuitively and continuously travel to each target department simply by following the virtual guidance overlaid on the screen.

[0134] To better understand the embodiments of this application, the method for generating medical routes provided by the embodiments of this application will be introduced below with reference to a specific example.

[0135] The method for generating a patient's medical route provided in this example is implemented by various functional modules within IMCS (Intelligent Medical Companion System). The overall process for a patient to seek medical care using this method is as follows: Figure 8 As shown. First, IMCS will synchronize the queue status in real time, that is, obtain the real-time queue information of each laboratory through periodic polling. The queue information of each laboratory includes the patient information of each queued patient and the specific test items of each queued patient in that laboratory.

[0136] Once the doctor prescribes tests for the patient, the IMCS (In-Process Clinical Review System) notifies the patient that the order is complete. Specifically, IMCS filters the logged-in user's information to retrieve a list of all pending tests for that user (corresponding to the target patient mentioned above), including test name, queue status, and completion status. The system then notifies the patient that the order is complete. Next, the patient needs to personalize their settings based on their preferences (time priority, route priority, priority for a specific test, custom settings, etc.). IMCS analyzes these preferences and sends planning information to the patient based on a multi-factor feature fusion-based test time prediction model and the patient's preferences. This planning information includes the test list and real-time queue information for each laboratory. The test list is generated by arranging the tests in the planned order.

[0137] When the test list is not empty, the next test is planned after the current test begins (based on patient preferences and departmental queue information). The IMCS performs intelligent background check-in based on the current test and the model-predicted time information, effectively checking the patient in for the next test. After checking in to the corresponding laboratory for the next test, the laboratory reports successful check-in to the IMCS and returns the queue number. The IMCS then pushes the queue information for the next department to the patient.

[0138] In IMCS, the intelligent planning module can predict the completion time of the current test by combining the patient's current medical information; at the same time, it can predict the travel time to the next department by combining the hospital map model and the patient's travel time (corresponding to the travel time mentioned above). Then, based on the predicted completion time and travel time, AR navigation is activated at an appropriate time, that is, the patient is guided to the department where the next test is located by using the hospital AR map. After arriving, the patient can directly wait and undergo the test.

[0139] After completing a patient's examination, the laboratory can send a confirmation message to the patient confirming completion. The patient can then report the completion of the current test to the IMCS (Integrated Medical Systems Center). This process continues until all tests are completed, at which point the IMCS sends a notification to the patient confirming completion of all tests. Once the patient receives this notification, they can return to their doctor for a follow-up appointment.

[0140] like Figure 9 and Figure 10 As shown, the IMCS service includes displaying a list of test items, intelligently planning the test process, updating the queue, AI (Artificial Intelligence) interaction, intelligent check-in for test items, WeChat mini-programs, route guidance, and push notifications of test results. The third-party services required to implement the IMCS service include HIS (Hospital Information System), LIS (Laboratory Information Management System), PACS (Picture Archiving and Communication System), WeChat, maps, and AI interaction.

[0141] After obtaining a lab test order from the outpatient clinic, patients access IMCS via a WeChat mini-program and complete identity authentication and login. Once logged in, patients can view their lab test list, which includes the status of each test (including attendance and completion status), current waiting time, and test results. Patients can check the progress of their tests during their appointment and view the results immediately afterward (i.e., query the results). IMCS proactively pushes the test results to patients. For different patient preferences, patients can set their own preferences (including time priority, location priority, priority of specific tests, and custom sorting). After setting these preferences, IMCS will plan the patient's appointment based on established principles. It will then perform intelligent check-in and navigation to lab departments according to the plan. Furthermore, if a patient's waiting time in a particular department exceeds a preset threshold (e.g., 20 minutes), IMCS will activate an intelligent companion assistant to interact with the patient via AI, including explaining the planned process, providing psychological support, offering emotional companionship and comfort, and alleviating anxiety.

[0142] like Figure 11 As shown, IMCS comprises three modules: a user management module, a predictive model management module, and an IMCS management module. These three modules interact with external third-party software systems to achieve their respective functions. The user management module includes user login, user authentication, and status query. Specifically, user login allows users to log in to IMCS via a WeChat mini-program and set their medical preferences; user authentication ensures privacy by confirming logged-in patient information; and status query retrieves relevant examination information for patients and allows them to check the status of examinations and estimated queueing information on the mini-program. The external third-party software systems include a WeChat mini-program, an AI engine, a map engine, and laboratory software. The laboratory software includes LIS, HIS, and PACS from the hospital system.

[0143] The predictive model management module includes the following functions: using an open-source AI interaction model to conduct patient dialogues, providing psychological comfort, health knowledge dissemination, and entertainment interaction; using an open-source map navigation engine to assist navigation, providing patients with navigation functions accurate to the department and room level, supporting AR guidance, and providing relevant time information for patient movement during their visit; analyzing external influencing factors through a time prediction model to calculate the precise visit time for each patient currently visiting the laboratory department; and using a treatment optimization model based on the time prediction model to perceive patient visit preferences, and combining departmental route optimization to output an optimized visit sequence.

[0144] The IMCS management module includes: an interface management module for interacting with external third-party software systems, handling unpacking and repackaging protocols; a system monitoring module for managing real-time data from laboratories and polling the laboratory interfaces; a laboratory status management module for processing real-time data from the current laboratory system and managing the status of each laboratory and laboratory item for user query; a patient visit planning module for intelligent planning of patient visits based on a treatment optimization model, and displaying the planned visits; and intelligent check-in, which uses current visit planning information, predicted laboratory item times, and route planning times to calculate check-in time and proactively completes the patient's check-in for the next laboratory.

[0145] Time forecasting models, as an important component of IMCS, provide fundamental services for the planning and calculation of other modules. For example... Figure 12 As shown, the time prediction model uses deep learning to fit a model of the expected time to be used for a single test item based on input information from multiple dimensions.

[0146] The input layer of the time prediction model can accept information from multiple dimensions, including patient information, patient medical history, patient age, doctor information, doctor's operating habits, doctor's seniority, specific test item names, hospital level, and hospital peak hours.

[0147] The feature processing layer of the time prediction model transforms various types of raw data into features that the model can process. Specifically, non-numerical parameters are encoded / embedded, and numerical parameters are normalized. For example, patient information, patient medical history, doctor information, doctor's operating habits, specific test names, hospital level, and peak hospital periods are encoded as features, and patient age is normalized.

[0148] The feature concatenation layer of the time prediction model concatenates all features to perform multi-factor feature fusion and capture the interaction relationships between features.

[0149] The time prediction model consists of a multi-layer neural network structure, including a fully connected layer 1 (128 dimensions), a ReLU activation function, a Dropout layer (dropout rate 0.2), a fully connected layer 2 (64 dimensions), another ReLU activation function, and an output layer (1 dimension). The Dropout layer prevents overfitting, and the ReLU activation function introduces non-linearity.

[0150] The output layer of the time prediction model outputs a single numerical value, which is the predicted time for a single test (the unit can be set to minutes); the model can be trained using historical data and the model parameters can be optimized through a loss function.

[0151] like Figure 13As shown, the diagnosis and treatment optimization model outputs a planned list of test order by predicting the waiting time of different departments, the patient's set consultation preferences, and the specific location of the departments.

[0152] The model's input layer includes: patient visit preferences, predicted waiting times for individual tests (corresponding to the waiting time mentioned above), and travel time. Patient visit preferences include time preferences and travel preferences; the predicted waiting time for individual tests is the predicted waiting time for each test output by the time prediction model; and travel time is the inter-departmental travel time provided by the map engine.

[0153] The model's data processing layer transforms the input data into a matrix form usable by the planning algorithm. This includes: quantifying patient visit preferences into weights and constructing a preference weight vector P=[w1,w2], where w1 and w2 are the weights corresponding to waiting time and distance, respectively; constructing a waiting time matrix based on the waiting time of each test item, for example, if the waiting time for test item A is t1, the waiting time for test item B is t2, and the waiting time for test item C is t3, then the waiting time matrix W=[t1,t2,t3] is constructed; and constructing a distance-time matrix containing the travel time between departments, for example, if the distance between departments AB is d1, the distance between departments BC is d2, and the distance between departments AC is d3, then the distance-time matrix D=[d1,d2,d3].

[0154] The planning engine constructs the state space by integrating all waiting times, travel times, and constraints (corresponding to the testing constraints mentioned above). These constraints include project dependencies (e.g., an ultrasound can only be performed after blood collection); equipment availability time periods (affected by hospital level); doctor scheduling times; and department service times. Candidate routes are generated based on the state space by permuting and combining all possible testing sequences and selecting those that satisfy the constraints. Then, a multi-objective evaluation function is constructed based on the preference weight vector, which serves as the evaluation criterion for comprehensively assessing each candidate route. Finally, dynamic programming / reinforcement learning is used as the "optimization engine" to select the optimal path, efficiently searching the complex decision space to find the action sequence that truly optimizes the multi-objective evaluation function.

[0155] The output layer outputs a list of planned test order sequences. For example, Option 1: A->C->B, total time T1; Option 2: B->A->C, total time T2; Option 3: C->B->A, total time T3. Based on this list, the optimal and alternative options can be recommended. Furthermore, the test order sequence supports real-time adjustments, meaning it acquires real-time queue changes and tracks patient locations, dynamically replanning the test order based on the real-time queue situation and patient location.

[0156] The workflow of the IMCS system monitoring module is as follows: Figure 14 As shown, after monitoring begins, the monitoring module is initialized, i.e., the system monitoring module is started, to check the connection status of each testing system. If the status is normal, the testing system management module begins the polling query process. The polling scheduler starts according to the set period, obtains the list of testing systems to be queried, and iterates through each testing system, calling the interface of the testing interface management module, receiving and parsing the real-time data returned by the testing interface management module. Simultaneously, local cache information is updated. The testing interface management module provides a unified interface service, receiving query requests and returning the testing system status.

[0157] The system monitoring module updates the status information of each testing system in real time and maintains real-time information databases for different departments, such as real-time information for Department A, Department B, and Department C.

[0158] In the system monitoring module, real-time information from each department is stored independently. The information aggregation processor integrates multi-source data to detect status changes, record changes, and persist the data. This means that status changes are recorded in a historical status database, and statistical analysis reports can be generated based on these changes. Users can initiate query requests through the user interaction layer. Upon receiving the request, the information aggregation processor pushes updates, allowing users to view statistical analysis reports on the real-time monitoring dashboard displayed in the user interaction layer. The monitoring dashboard provides detailed test status information, allows filtering by department, and supports historical data tracing and analysis. If the connection status of any testing system becomes abnormal, an anomaly alarm is triggered. The user interaction layer receives the alarm and the process ends.

[0159] The core function of IMCS's laboratory status management module is to accurately maintain the current status of each laboratory department and patient items by integrating real-time information from the system monitoring module.

[0160] like Figure 15 As shown, the laboratory status management module first acquires all departmental information from the system monitoring module, including three key real-time data categories: doctor on-duty status, equipment status, and number of people waiting in line. After acquiring the information, the module integrates and analyzes this data to determine if the laboratory status has changed. When a status change is detected, the module triggers updates in two dimensions: patient item status updates and laboratory department status updates. The laboratory department status reflects the department's workload and equipment availability; the patient item status update drives the status progression of patient tests, such as from "pending testing" to "in progress," then to "completed" and "report issued." All status changes are recorded and stored, forming a traceable status change log. If there is no status change, the current update cycle ends, awaiting the next trigger.

[0161] Through the test status management module, IMCS can reflect the dynamics of the entire test process in real time and accurately, providing patients and medical staff with reliable status query services.

[0162] In IMCS, the intelligent planning module is the implementation layer module for the diagnosis and treatment optimization model. For example... Figure 16 As shown, after receiving a patient's planning request, the intelligent planning module simultaneously retrieves information from three data sources: the user management module provides patient information and preference settings, the map engine module provides travel time between departments, and the system monitoring module provides real-time queuing status for each department. The intelligent planning module integrates this multi-source data to construct model input features, calls the diagnosis and treatment optimization model for inference calculations, and finally generates the optimal treatment sequence plan and returns the planning results to the front end for display. Through this process, the system can provide patients with personalized intelligent medical route planning that considers real-time conditions.

[0163] The IMCS system also includes a smart check-in module. For example... Figure 17 As shown, after a patient begins their current test, the system automatically retrieves their next scheduled test and uses a time prediction model to calculate the remaining time for the current test (corresponding to the remaining test duration mentioned above) and the waiting time for the next test. Simultaneously, it retrieves the travel time to the next department from the map engine. The intelligent check-in module calculates the optimal check-in time based on the current time, the remaining time for the current test, and the travel time, and continuously monitors the real-time status for dynamic adjustments. When the calculated optimal check-in time is reached, a prompt message indicating the optimal check-in time is sent to the patient's user device, guiding the patient to actively trigger check-in on their device. After the patient triggers check-in, the intelligent check-in module automatically completes the check-in for the next test in the background, updates the status of the next test to "checked in and awaiting test," and pushes a notification to the patient reminding them to go to the corresponding department. If the optimal check-in time has not been reached, the system returns to retrieve the waiting time for the next test. This mechanism avoids patients arriving too early to queue and also prevents them from missing their turn due to lateness, achieving an intelligent check-in experience.

[0164] This solution achieves global dynamic path optimization for patients' multiple testing tasks. By collecting real-time queuing status of each laboratory, spatial distance between departments, and patient preferences (time priority / space priority / custom), combined with a trained planning model, it dynamically generates the optimal testing order and walking path for patients, effectively avoiding unnecessary backtracking and aimless movement within the hospital. The intelligent check-in module monitors the remaining completion time of the current laboratory and the queue change trend of the next laboratory in real time, calculates the optimal check-in time, and automatically completes the check-in operation in the background, reducing the burden of repetitive operations for patients and avoiding the problems of arriving too early leading to long waits or arriving too late leading to missed appointments. It also improves the depth of queuing information utilization, moving beyond a static display of real-time numbers to a system that trains the status information of each testing item, combining historical data and real-time trends to optimize queuing. Queue management predicts and analyzes queue patterns, allowing patients to anticipate queue conditions upon arrival at the next department and dynamically adjust testing order based on changes, thus proactively optimizing the testing process. An AI interaction module proactively provides emotional support and psychological comfort to patients while they wait or move, alleviating anxiety and improving their overall experience. The system can monitor patients' test completion status in real time and proactively push the optimal time to visit the next testing department, achieving real-time linkage between patient status and testing progress. AR navigation provides intuitive guidance to the testing department, enhancing convenience and user-friendliness. When all tests are completed, patients are automatically notified of the end of the testing process and informed of the estimated result times for each item, proactively pushing results to patients upon arrival, forming a closed-loop information system for the entire testing process.

[0165] Corresponding to the above method embodiments, this application provides a medical appointment route generation device, such as... Figure 18 As shown, the device includes:

[0166] The first prediction module 1810 is used to predict the required testing time for each queued patient in each target department based on the patient information, the test information of each queued patient in the target department, and the doctor information of the department. Based on the testing time, the total queuing time of the target department is obtained. Each target department is the department that conducts tests on each category of test items for the target patients.

[0167] The generation module 1820 is used to generate candidate routes for the target patient to perform each category of tests based on the total queuing time of each target department, the location of each target department, and the test constraints between each category of tests to be performed by the target patient.

[0168] The second prediction module 1830 is used to predict the total waiting time of the target patient in each target department according to the candidate route, based on the total queuing time of each target department.

[0169] The first determining module 1840 is used to determine the target patient's medical route from among the candidate routes based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's medical preferences.

[0170] Optionally, the device further includes:

[0171] The second determining module is used to return to the steps of predicting the required testing time for each patient in the target department after the first test for the target patient has started, based on the patient information of each patient in the target department, the test information of each patient in the target department, and the doctor information of the department, and obtaining the total waiting time of the target department based on the testing time, and redetermine the medical route to obtain the current medical route of the target patient.

[0172] The third determining module is used to determine the second item to be checked by the target patient in the test items based on the target patient's current medical route;

[0173] The acquisition module is used to acquire the remaining inspection time of the first item;

[0174] The calculation module is used to calculate the optimal check-in time for the second item based on the target patient's current location, remaining testing time, and total queuing time of the department to be checked in, and to execute a check-in task based on the optimal check-in time to check in the department to be checked in for the target patient. The check-in task is triggered by the target patient based on a prompt message indicating the optimal check-in time. The department to be checked in is the target department that will perform the second item test.

[0175] Optionally, the computing module includes:

[0176] The first calculation submodule is used to calculate the travel time required for the target patient to travel from the current location to the department to be checked in;

[0177] The first prediction submodule is used to predict the arrival time of the target patient to the department to be checked in based on the passage time and the remaining inspection time.

[0178] The first determining submodule is used to determine the current time as the optimal check-in time for the second item if the arrival time is not later than a specified time; wherein, the specified time is: the time after shifting the total queuing time of the departments to be checked in backwards from the current time;

[0179] The return submodule is used to return the step of obtaining the remaining inspection time of the first item according to a predetermined time interval if the arrival time is later than the specified time.

[0180] Optionally, the generation module 1820 includes:

[0181] The extrapolation submodule is used to extrapolate the time of the target patient's journey to each target department according to each initial route in different orders, based on the target patient's current location, the location of each target department, the current time, the total queuing time of each target department, and hospital information, so as to obtain the predicted consultation time of the target patient in each target department.

[0182] The second determining submodule is used to determine candidate routes from each initial route that meet the testing constraints between the various categories of test items for the target patient, based on the predicted consultation time corresponding to each target department in each initial route.

[0183] Optionally, the first determining module 1840 includes:

[0184] The second calculation submodule is used to determine the weight of the total waiting time and the weight of the distance for each candidate route according to the patient's medical preferences, and calculate the weighted sum of the total waiting time and distance for the candidate route based on the determined weights to obtain the total cost of the candidate route.

[0185] The third determination submodule is used to determine the medical route for the target patient based on the candidate route with the lowest total cost among all candidate routes.

[0186] Optionally, the third determining submodule includes:

[0187] A providing unit is configured to provide the target patient with the first candidate route that has the lowest total cost among all candidate routes;

[0188] An adjustment unit is used to respond to the target patient's operation of adjusting the order of visits to each target department in the first candidate route, and to obtain an adjusted second candidate route;

[0189] The feedback module is used to obtain the time cost and distance cost of the second candidate route, and to feed back the time cost and distance cost to the target patient so that the target patient can confirm the route based on the time cost and distance cost;

[0190] The confirmation module is used to confirm the indicated route as the target patient's medical route in response to the route confirmation instruction from the target patient.

[0191] Optionally, the first prediction module 1810 predicts the required testing time for each queued patient in each target department in the following manner:

[0192] Numerical and non-numerical parameters are determined from the patient information of the queued patients, the test information of the queued patients in the target department, and the information of the doctors in the target department.

[0193] The determined non-numerical parameters are encoded to obtain the first feature, and the determined numerical parameters are normalized to obtain the second feature.

[0194] By concatenating the first feature and the second feature, a feature vector is obtained.

[0195] Based on the feature vector, the required testing time for the queued patient in the target department is predicted.

[0196] Optionally, the second prediction module 1830 includes:

[0197] The second prediction submodule is used to predict the time it takes for the target patient to reach each target department according to the candidate route, based on the total queuing time of each target department and the location of each target department.

[0198] The third calculation submodule is used to calculate the waiting time of each target department if the consumption time corresponding to the target department is not greater than the total queuing time of the target department. If the consumption time corresponding to the target department is greater than the total queuing time of the target department, the preset value is determined as the waiting time of the target department.

[0199] The fourth calculation submodule calculates the sum of the waiting times in each target department to obtain the total waiting time of the target patient in each target department when performing the test according to the candidate route.

[0200] Optionally, the device further includes:

[0201] The fourth determination module is used to determine the next target department for the target patient based on the medical route and the status of each test item;

[0202] The fifth determining module is used to determine the navigation start time based on the waiting time of the target patient in the next target department and the travel time required for the target patient to travel from the current location to the next target department.

[0203] The push module is used to push a guidance map about the medical route to the target patient when the determined navigation start time is reached.

[0204] This application also provides an electronic device, such as... Figure 19As shown, it includes a processor 1901, a communication interface 1902, a memory 1903, and a communication bus 1904. The processor 1901, the communication interface 1902, and the memory 1903 communicate with each other through the communication bus 1904.

[0205] Memory 1903 is used to store computer programs;

[0206] When the processor 1901 executes the program stored in the memory 1903, it implements any of the above-described methods for generating a medical route.

[0207] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0208] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0209] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0210] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0211] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described methods for generating a medical route.

[0212] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the above-described methods for generating a medical route.

[0213] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0214] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0215] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, electronic devices, and computer-readable storage media are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0216] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A medical route generating method characterized by comprising: The method includes: For each target department, based on the patient information of each queued patient in that target department, the test information of each queued patient in that target department, and the doctor information of the department, the required testing time for each queued patient in that target department is predicted, and the total waiting time for that target department is obtained based on the testing time of each test; each target department is: the department that performs the tests for each category of test items for the target patients; Based on the total queuing time of each target department, the location of each target department, and the testing constraints between the various categories of tests to be tested for the target patient, candidate routes for the target patient to undergo each category of tests are generated. For each candidate route, based on the total queuing time in each target department, predict the total waiting time of the target patient in each target department when undergoing examination according to the candidate route; Based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the patient's medical preferences, the patient's medical route is determined from the candidate routes.

2. The method of claim 1, wherein, The method further includes: After the first test for the target patient begins, the process returns to the steps of predicting the required testing time for each patient in the target department based on the patient information, the test information for each patient in the target department, and the doctor information of each patient in the target department, and obtaining the total waiting time for the target department based on the testing time, and then redetermines the medical route to obtain the current medical route for the target patient. Based on the target patient's current medical route, determine the second item that the target patient needs to sign in for in the tests; Obtain the remaining inspection time for the first item; Based on the target patient's current location, remaining testing time, and total queuing time of the department to be checked in, the optimal check-in time for the second item is calculated, and a check-in task based on the optimal check-in time is executed to check the target patient in the department to be checked in. The check-in task is triggered by the target patient based on a prompt message indicating the optimal check-in time. The department to be checked in is the target department that will perform the second item test.

3. The method of claim 2, wherein, The calculation of the optimal check-in time for the second item, based on the target patient's current location, remaining testing time, and total queuing time for the department to be checked in, includes: Calculate the travel time required for the target patient to travel from their current location to the department where they need to check in; Based on the passage time and remaining inspection time, predict the arrival time of the target patient at the department to be checked in; If the arrival time is not later than the specified time, the current time is determined as the optimal check-in time for the second item; wherein, the specified time is: the time after shifting the total queuing time of the departments to be checked in backwards from the current time; If the arrival time is later than the specified time, return to the step of obtaining the remaining inspection time of the first item according to the predetermined time interval.

4. The method of claim 1, wherein, The process of generating candidate routes for each category of tests for the target patient based on the total queuing time of each target department, the location of each target department, and the testing constraints between different categories of tests for the target patient includes: For each initial route leading to each target department in different orders, based on the target patient's current location, the location of each target department, the current time, the total queuing time of each target department, and hospital information, the time of the target patient's journey to each target department according to the initial route is estimated to obtain the predicted consultation time of the target patient in each target department. Based on the predicted consultation time corresponding to each target department in each initial route, candidate routes that meet the testing constraints among the various categories of test items for the target patient are determined from each initial route.

5. The method of claim 1, wherein, The process of determining the target patient's medical route from the candidate routes based on the total waiting time, the distance of each candidate route, and the target patient's medical preferences includes: For each candidate route, based on the target patient's medical preferences, the weights of the total waiting time and the distance corresponding to the candidate route are determined respectively. Based on the determined weights, the weighted sum of the total waiting time and the distance corresponding to the candidate route is calculated to obtain the total cost of the candidate route. The patient's medical route is determined based on the candidate route with the lowest total cost among all candidate routes.

6. The method of claim 5, wherein, The process of determining the patient's treatment route based on the candidate route with the lowest total cost among all candidate routes includes: Provide the target patient with the first candidate route that has the lowest total cost among all candidate routes; In response to the target patient's operation of adjusting the order of visits to each target department in the first candidate route, an adjusted second candidate route is obtained; The time cost and distance cost of the second candidate route are obtained, and the time cost and distance cost are fed back to the target patient so that the target patient can confirm the route based on the time cost and distance cost; In response to the route confirmation instruction from the target patient, the indicated route is confirmed as the target patient's medical route.

7. The method of claim 1, wherein, The required testing time for each patient in each target department is predicted as follows: numerical and non-numerical parameters are determined from the patient information, the test information of the patient in the target department, and the doctor information of the target department; the determined non-numerical parameters are encoded to obtain a first feature, and the determined numerical parameters are normalized to obtain a second feature; the first and second features are concatenated to obtain a feature vector. Based on the feature vector, predict the required testing time for the queued patient in the target department; And / or, The step of predicting the total waiting time of the target patient in each target department according to the candidate route, based on the total queuing time of each target department, includes: predicting the time consumed by the target patient to reach each target department according to the candidate route based on the total queuing time of each target department and the location of each target department; for each target department, if the consumption time corresponding to the target department is not greater than the total queuing time of the target department, then the difference between the total queuing time of the target department and the consumption time corresponding to the target department is calculated to obtain the waiting time of the target department; if the consumption time corresponding to the target department is greater than the total queuing time of the target department, then a preset value is determined as the waiting time of the target department; the sum of the waiting times of each target department is calculated to obtain the total waiting time of the target patient in each target department according to the candidate route. And / or, The method further includes: determining the next target department for the target patient based on the medical route and the status of each test item; determining the navigation start time based on the waiting time of the target patient in the next target department and the travel time required for the target patient to travel from the current location to the next target department; and pushing a guidance map about the medical route to the target patient when the determined navigation start time is reached.

8. A visit route generating apparatus characterized by comprising: The device includes: The first prediction module is used to predict the required testing time for each queued patient in each target department based on the patient information, the test information of each queued patient in the target department, and the doctor information of the department. Based on the testing time, the total queuing time of the target department is obtained. Each target department is the department that conducts tests on each category of test items for the target patients. The generation module is used to generate candidate routes for the target patient to perform each category of tests based on the total queuing time of each target department, the location of each target department, and the test constraints between each category of tests to be performed by the target patient. The second prediction module is used to predict the total waiting time of the target patient in each target department according to the candidate route, based on the total queuing time of each target department. The first determining module is used to determine the target patient's medical route from among the candidate routes based on the total waiting time corresponding to each candidate route, the distance of each candidate route, and the target patient's medical preferences.

9. An electronic device, comprising: It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images