Intelligent hospital guide system for outpatient triage
By integrating dialect databases and Chinese-foreign language switching models, combined with multimodal interaction modes and in-hospital navigation, the problem of communication barriers between patients who only speak dialects and the triage system has been solved, improving triage efficiency and patient experience.
Patent Information
- Application Number
- CN202511121045.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-12
- Publication Date
- 2025-11-28
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing outpatient smart triage system has communication barriers when interacting with patients who only speak dialects, which prevents patients from receiving timely system suggestions and prolongs the triage time.
By integrating dialect data and language data from around the world, a dialect database and a Chinese-foreign language switching model are established. Combined with multimodal interaction modes and in-hospital navigation models, intelligent triage services are provided, supporting patient interaction from multiple language backgrounds.
It reduces communication barriers caused by language barriers, improves the patient interaction experience, reduces time spent in the triage process, and optimizes the hospital's operational efficiency.
Smart Images
Figure CN121034572A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the medical technical field, and in particular to an outpatient triage intelligent guiding system. BACKGROUND
[0002] The outpatient triage intelligent guiding system is a system that integrates various data resources by means of modern information technology to provide intelligent services for outpatient patients from triage, guiding to the whole process of medical treatment. It aims to optimize the outpatient medical treatment process, improve the efficiency of patient medical treatment, improve the medical treatment experience, reasonably allocate medical resources, and improve the overall operation efficiency of the hospital.
[0003] In the prior art, such as the outpatient intelligent guiding system (general type), it has a voice interaction mode with Putonghua as the core, which can interact with patients through standard Putonghua, and through voice recognition technology and natural language processing technology, the system can obtain, analyze and understand the language of the patient, and realize the understanding and feedback of the patient's semantics.
[0004] The above system can use standard Putonghua for voice interaction when interacting with patients to ensure smooth communication with patients in Putonghua. However, for patients who only speak dialects, there may be communication barriers, which makes it difficult for the system to accurately understand the patient's intention, so that the patient cannot get the system's suggestion in time, and the patient's guiding time is prolonged.
[0005] In summary, the existing system may cause patients who only speak dialects to be unable to get the system's suggestion in time when using voice interaction to interact with patients, which prolongs the patient's guiding time. This problem has become a difficult problem that needs to be solved in the field, so it is necessary to propose an outpatient triage intelligent guiding system. SUMMARY
[0006] To solve the above problems, the present application provides an outpatient triage intelligent guiding system, which integrates dialect data and language data of countries around the world through a model establishment module, establishes a dialect database, a Chinese and foreign language database and a Chinese and foreign language switching model, so that the system can better serve patients with different language backgrounds, reduce the communication barriers caused by language barriers, and reduce the time consumption of patients in the guiding process.
[0007] In order to achieve the above purpose, the technical scheme of the present application is as follows: an outpatient triage intelligent guiding system, comprising the following modules:
[0008] A data acquisition module is used to acquire patient case data, local dialect data, language data of countries around the world, 3D map data of the hospital and disease data of existing diseases.
[0009] The model establishing module is configured to establish a disease database based on disease data of existing diseases, establish a disease model based on the disease database, establish a dialect database based on dialect data, establish a Chinese and foreign language database based on language data of countries around the world, and establish a Chinese and foreign language switching model based on the Chinese and foreign language database; and establish a multi-layer weighted directed graph based on 3D map data of the hospital, and establish an in-hospital navigation model based on the multi-layer weighted directed graph.
[0010] The case searching module is configured to allow a patient to input a disease, search for a same case through the disease model according to the disease input by the patient, and dock the found case with medical staff to verify the accuracy of the case searching. The case searching module is also configured to allow information exchange between medical staff and confirm a department to which the found case belongs.
[0011] The intelligent guidance module is configured to integrate a multi-modal interaction mode of physical buttons, touch interaction and voice interaction, switch the interaction mode according to a specific condition of a patient, and provide intelligent navigation for the patient based on AR technology and the in-hospital navigation model to plan an optimal travel route.
[0012] The intelligent guidance module is also configured to provide an online number calling service for the patient, and display a predicted arrival time and a predicted waiting time for the patient in real time, and associate and match the predicted arrival time and the predicted waiting time.
[0013] The real-time shunting module is configured to monitor a passenger flow of each department and each channel, generate a heat distribution map, and send a real-time shunting suggestion to a patient according to the heat distribution map; and the real-time shunting module is also configured to send detailed information of an expanded outpatient service to the patient when the outpatient service of the same department is expanded.
[0014] The emergency processing module is configured to locally cache a 24-hour scheduling table and map data, and formulate and take a corresponding emergency processing scheme when one or more of a network disconnection, a queue exceeding a limit or a device failure occurs.
[0015] The intelligent suggestion module is configured to provide customized suggestions based on individual characteristics of a patient.
[0016] Further, the data collection module includes the following units:
[0017] The case data collection unit is configured to collect patient case data by docking a HIS system.
[0018] The dialect data collection unit is configured to collect dialect data of a current region through an ASR voice recognition engine.
[0019] The Chinese and foreign language data collection unit is configured to collect language data of countries around the world through ParseHub.
[0020] The map data collection unit is configured to collect 3D map data of the hospital by using a laser scanning point cloud data technology.
[0021] The disease data collection unit is configured to collect disease data of existing diseases by using an Internet large model.
[0022] Further, the data collection module is also configured to collect and organize high-frequency words in the patient's inquiry process, so that the model establishment module can dialectize the high-frequency words when establishing the dialect database, and translate the high-frequency words when establishing the Chinese-foreign language database.
[0023] Further, when the intelligent guide diagnosis module switches the interaction mode, the operation process thereof includes the following steps:
[0024] The patient's case data and specific age are obtained by face recognition.
[0025] The interaction mode is adaptively adjusted according to the patient's age. When the patient's age is between 15 and 60 years old, the touch interaction mode is adopted; when the patient's age is between 60 and 75 years old, the physical button interaction mode is adopted; when the patient's age is more than 75 years old or less than 15 years old, the voice interaction mode is adopted.
[0026] When the patient communicates with the system by using dialect, the intelligent guide diagnosis module identifies the dialect type of the patient through the dialect database and switches to the corresponding dialect type to communicate with the patient by voice; when the patient communicates with the system by using foreign language, the intelligent guide diagnosis module identifies the foreign language of the patient through the Chinese-foreign language switching model, and converts the operation mode according to the foreign language of the patient, so that the touch interaction content and the voice interaction content are converted into the same foreign language as the patient.
[0027] Further, when the intelligent guide diagnosis module associates and matches the predicted arrival time and the predicted waiting time, the operation logic thereof is as follows:
[0028] When the predicted waiting time of the department the patient has made an appointment for is less than 15 minutes, and the patient's predicted arrival time at the department is less than 15 minutes, the patient is suggested to go to the department immediately.
[0029] When the predicted waiting time of the department the patient has made an appointment for is between 15 and 30 minutes, and the patient's arrival time at the department is less than 30 minutes, the patient is suggested to go to the waiting area to wait.
[0030] When the predicted waiting time of the department the patient has made an appointment for is more than 30 minutes, the patient is suggested to continue to pay attention to the change of the predicted waiting time.
[0031] During this period, when the serial number changes, the intelligent guide diagnosis module immediately sends a notification push to the patient for prompt.
[0032] Further, the operation process of the real-time diversion module includes the following steps:
[0033] Real-time monitoring of the flow of each department and each channel, generating a heat distribution map based on the flow of each department and each channel, and displaying it in real time for patients, while setting a flow threshold to determine the specific state of the department and the channel. When the flow exceeds the flow threshold, the department is in full load state and the channel is in impassable state.
[0034] When the same department's outpatient service expands, the location of the expanded outpatient service is monitored in real time, and the patients are assigned weights according to their registration types. Based on the weight distribution result, the specific information of the expanded outpatient service is pushed to the patients, and the patients are asked whether they are willing to go to the expanded outpatient service. When the patient is willing to go to the expanded outpatient service, the patient is guided to the expanded outpatient service.
[0035] Further, the emergency handling scheme in the emergency handling module is as follows:
[0036] When the network is disconnected, the emergency handling module switches the system to offline mode, changes the positioning mode of intelligent navigation to Bluetooth Beacon fingerprint positioning, and stores the data during the network interruption locally. When the network is restored, the offline operation log is automatically uploaded.
[0037] When the queue is over limit, the emergency handling module first determines whether there is a standby outpatient service of the same department on the same floor. When there is a standby outpatient service of the same department on the same floor, the emergency diversion suggestion of the same floor is pushed to the patient. When the patient accepts the emergency diversion suggestion of the same floor, the patient is guided to the standby outpatient service through AR navigation technology. When there is no standby outpatient service on the same floor, the emergency handling module queries the nearest idle outpatient service of the same department, plans the best path to the idle outpatient service, and pushes the emergency diversion suggestion to the idle outpatient service. When the patient accepts the emergency diversion suggestion to the idle outpatient service, the patient is guided to move to the idle outpatient service through AR navigation technology.
[0038] When the equipment fails, the emergency handling module automatically switches to the adjacent equipment networking and generates a maintenance work order, while scanning the non-faulty equipment, generating the best path to the non-faulty equipment, and guiding the patient to the location of the non-faulty equipment through voice broadcast.
[0039] Further, it also includes an online medicine pickup module for providing online medicine pickup service for patients. After the patient pays online, the online medicine pickup module sends the medicine pickup information to the medicine pickup point, and the medical staff at the medicine pickup point dispenses medicine for the patient according to the medicine pickup information.
[0040] Further, the real-time shunting module adopts a linear weight distribution method to distribute weights, and the initial weights of the n channels are respectively denoted as w1, w2, …, wn. n , and make When the weight needs to be adjusted according to the index x, the specific formula of the linear weight distribution is as follows:
[0041]
[0042] Wherein, the index x is the registration type index of the patient, w i ′ is the adjusted weight, w i is the initial weight, and a is used to control the influence degree of the index x on the weight, a>0, x i is the index value corresponding to the i th channel, is the average value of the index values of all channels, and the specific formula is as follows:
[0043]
[0044] Wherein, x j is the index value corresponding to the j th channel.
[0045] After the weight adjustment is completed, the weight is normalized to meet The normalization formula is as follows:
[0046]
[0047] Wherein, w i " is the adjusted and normalized weight, is the sum of the adjusted weights of all channels.
[0048] Further, the intelligent guide diagnosis module adopts the shortest path algorithm to calculate the shortest path to the target point when planning the route, and the specific formula is as follows:
[0049] f(n)=g(n)+h(n)(4)。
[0050] Wherein, g(n) is the actual cost from the starting point to node n, and h(n) is the heuristic function from node n to the end point, and the specific representation is as follows:
[0051] h(n)=|dx|+|dy|+k*|dz| (5)。
[0052] Wherein, dx, dy and dz all represent the coordinate difference between the node and the end point, and k represents the weight factor of the floor height difference.
[0053] The above scheme has the following beneficial effects:
[0054] 1、The application, by the model establishing module, dialect data and language data of countries in the world are integrated, dialect database, Chinese and foreign language database and Chinese and foreign language switching model are established, so that the system can better serve patients with different language backgrounds, reduces the communication barriers caused by language barrier, thereby reducing the time consumption of patients in the process of guiding diagnosis.
[0055] 2、The application, the intelligent guiding diagnosis module supports the multi-modal interaction mode of physical button, touch interaction and voice interaction, and can switch the interaction mode according to the specific condition of the patient, fully considers the use habit and demand of different age groups, and improves the interaction experience of the system.
[0056] 3、The application, the real-time shunting module can generate the human flow heat distribution map of each department and each channel and display it to the patient in real time, so that the patient can intuitively understand the personnel density of different areas in the hospital, and then the patient can plan the action route in advance and avoid crowded areas, thereby reducing the time of blindly searching for departments in the hospital and the waiting time, and improving the overall operation efficiency of the hospital.
[0057] Additional aspects and advantages of the application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS
[0058] Figure 1 It is a structure schematic view of the outpatient triage intelligent guiding diagnosis system of the application.
[0059] Figure 2 It is a structure schematic view of the data acquisition module in the outpatient triage intelligent guiding diagnosis system of the application.
[0060] Figure 3 It is an operation flow schematic view of the intelligent guiding diagnosis module switching interaction mode in the outpatient triage intelligent guiding diagnosis system of the application.
[0061] Figure 4 It is a running logic schematic view of the outpatient triage intelligent guiding diagnosis system of the application when the time node of calling number is associated and matched with the time node of the patient's expected arrival at the diagnosis room.
[0062] Figure 5 It is an operation flow schematic view of the real-time shunting module in the outpatient triage intelligent guiding diagnosis system of the application.
[0063] Figure 6 It is a running logic schematic view of the emergency treatment scheme in the outpatient triage intelligent guiding diagnosis system of the application. DETAILED DESCRIPTION
[0064] The following will be further described in detail through specific embodiments:
[0065] Example 1:
[0066] As attached Figure 1 The system includes a data acquisition module for data collection, a model building module for establishing disease models, dialect databases, Chinese-foreign language switching models, and in-hospital navigation models, a case search module for finding patients with the same symptoms, an intelligent triage module for providing intelligent triage functions to patients, a real-time triage module for generating heat maps and sending real-time triage suggestions to patients, an emergency response module for dealing with emergencies, an intelligent suggestion module for providing customized suggestions to patients, and an online medication retrieval module for providing online medication retrieval services to patients.
[0067] The functions of each module are explained in detail below:
[0068] The data acquisition module is used to acquire patient medical records, local dialect data, language data from around the world, 3D map data of the hospital, and symptom data of existing diseases.
[0069] like Figure 2 As shown, the data acquisition module includes a case data acquisition unit, a dialect data acquisition unit, a Chinese and foreign language data acquisition unit, a map data acquisition unit, and a symptom data acquisition unit. The case data acquisition unit is used to collect patient case data by connecting to the HIS system; the dialect data acquisition unit is used to collect dialect data of the current location using an ASR speech recognition engine; the Chinese and foreign language data acquisition unit is used to collect language data from around the world using ParseHub; the map data acquisition unit is used to collect 3D map data of the hospital using laser scanning point cloud data technology; and the symptom data acquisition unit is used to collect symptom data of existing diseases using large-scale internet models.
[0070] The model building module is used to establish a disease database based on existing disease data, and to build disease models based on this database. It also establishes a dialect database based on dialect data, a Chinese-foreign language database based on language data from around the world, and a Chinese-foreign language switching model based on these databases. Furthermore, it constructs a multi-layered weighted directed graph based on the hospital's 3D map data, and uses this as the basis for building an in-hospital navigation model. Simultaneously, the data acquisition module collects and organizes high-frequency words from patient consultations, enabling the model building module to perform targeted dialectization of high-frequency words when building the dialect database, and targeted translation of high-frequency words when building the Chinese-foreign language database.
[0071] The case searching module is used for the patient to input a disease, and according to the disease input by the patient, the same case is searched through a disease model, and the searched case is connected with the medical staff to verify the accuracy of the case searching. In the embodiment, the patient can select one or more input modes of touch input, voice input and photograph input to input the disease.
[0072] For example, when the patient inputs the disease by touch input, and the input disease is fever and dizziness, the case searching module searches the case according to the disease of fever and dizziness, and after the searching is completed, the searched case and the input disease of the patient are connected with the medical staff, and the medical staff judges the accuracy of the case searching. When the medical staff judges that the searching result is accurate, the case searching module displays the case to the patient. When the medical staff judges that the searching result is inaccurate, the medical staff manually corrects the searching result, and the corrected searching result is displayed to the patient through the case searching module. At the same time, the case searching module also enters the corrected searching result into the disease model for retraining. In this process, the medical staff can interact with each other through the case searching module to confirm the accuracy of the case searching and the department to which the searched case belongs, and after confirming the department to which the searched case belongs, the case is displayed to the patient.
[0073] In the embodiment, when the patient is not in the hospital, the patient can still perform touch input through the mobile phone. By clicking the body pain point position and the pain time of the pain point position on the screen of the mobile phone, the touch input operation can be realized. At this time, the case searching module searches the case according to the input of the patient to obtain the searching result. After the searching result is confirmed by the medical staff, the case searching module pushes the searching result to the mobile phone of the patient.
[0074] The intelligent guidance module is used for the multimodal interaction mode integrating physical buttons, touch interaction and voice interaction, switches the interaction mode according to the specific condition of the patient, and provides intelligent navigation for the patient based on the AR technology and the hospital navigation model to plan the best travel route. At the same time, the intelligent guidance module is also used for providing online calling service for the patient, and real-time displaying the predicted arrival time and the predicted waiting time for the patient, and matching the predicted arrival time and the predicted waiting time.
[0075] As shown in Figure 3 The operation process of the intelligent guidance module when switching the interaction mode includes the following steps:
[0076] The intelligent guide diagnosis module first acquires the case data and specific age of the patient through face recognition. Then, the intelligent guide diagnosis module adaptively adjusts the interaction mode according to the age of the patient. When the patient is 15-60 years old, the touch interaction mode is adopted; when the patient is 60-75 years old, the physical key interaction mode is adopted; and when the patient is more than 75 years old or less than 15 years old, the voice interaction mode is adopted.
[0077] In this process, when the patient communicates with the system through dialect, the intelligent guide diagnosis module identifies the dialect type of the patient through the dialect database and switches to the corresponding dialect type to communicate with the patient through voice; when the patient communicates with the system through foreign language, the intelligent guide diagnosis module identifies the foreign language of the patient through the Chinese-foreign language switching model and converts the operation mode according to the foreign language of the patient, so that the touch interaction content and the voice interaction content are both converted into the same foreign language as the patient.
[0078] As shown in Figure 4 , the running logic of the intelligent guide diagnosis module when matching the predicted arrival time and the predicted waiting time is as follows:
[0079] When the predicted waiting time of the department the patient is scheduled to visit is less than 15 minutes and the patient is scheduled to arrive at the department in less than 15 minutes, the patient is advised to go to the department immediately. When the predicted waiting time of the department the patient is scheduled to visit is 15-30 minutes and the patient is scheduled to arrive at the department in less than 30 minutes, the patient is advised to go to the waiting area to wait. When the predicted waiting time of the department the patient is scheduled to visit is more than 30 minutes, the patient is advised to continue to pay attention to the change of the predicted waiting time. During this period, when the serial number changes, the intelligent guide diagnosis module immediately sends a notification push to the patient for prompt.
[0080] For example, when the serial number of the patient is advanced, the intelligent guide diagnosis module sends a voice broadcast to the patient, "Your serial number has been advanced. It is recommended that you arrive at the examination room within 10 minutes", wherein "10 minutes" in the broadcast is the predicted waiting time of the patient.
[0081] The real-time shunting module is used to monitor the passenger flow of each department and each channel, generate a heat distribution map, and send real-time shunting suggestions to the patient according to the heat distribution map. It is also used to send detailed information of the expanded outpatient department to the patient when the same department has an expansion. In this embodiment, the heat distribution map is displayed in real time for the patient through the guide diagnosis screen, and the patient can query the heat distribution map at the current time point in real time through the mobile phone, so that the patient can better plan the route.
[0082] As shown in Figure 5 , the operation process of the real-time shunting module includes the following steps:
[0083] The real-time shunting module first monitors the flow of each department and each channel in real time, generates a heat distribution map based on the flow of each department and each channel, and displays the heat distribution map in real time through the guidance screen. Patients can also query the heat distribution map at the current time point through their mobile phones. Subsequently, the real-time shunting module sets a flow threshold, and determines the specific state of the department and the channel according to the flow threshold. When the flow exceeds the flow threshold, the department is determined to be in a full-load state, and the channel is determined to be in an impassable state.
[0084] During this process, when the out-patient department of the same department expands, the real-time shunting module also monitors the location of the expanded out-patient department in real time, and assigns weights to patients according to their registration types. Based on the weight distribution result, the specific information of the expanded out-patient department is pushed to the patient, and the patient is asked whether he or she is willing to go to the expanded out-patient department. When the patient is willing to go to the expanded out-patient department, the real-time shunting module guides the patient to the expanded out-patient department through the AR navigation mode.
[0085] For example, when the patient is an emergency patient of the cardiology department, the patient's weight distribution result is 85%. At this time, the real-time shunting module pushes the information of the expanded out-patient department to the emergency patient and asks whether he or she is willing to go to the expanded out-patient department. When the patient indicates that he or she is willing to go, the real-time shunting module guides the patient to the expanded out-patient department through AR navigation.
[0086] The emergency processing module is used to locally cache the 24-hour schedule and map data. When one or more of the following situations occurs: network interruption, queue overflow, or device failure, the corresponding emergency processing scheme is formulated and implemented. The specific operation process is as follows:
[0087] As shown in Figure 6 When the network is interrupted, the emergency processing module first switches the system to offline mode, loads and displays the locally cached 24-hour schedule and map data, changes the positioning mode of the intelligent navigation to Bluetooth Beacon fingerprint positioning, and stores the data during the network interruption locally. When the network is restored, the offline operation log is automatically uploaded.
[0088] When the queue is overflowed, the emergency processing module first determines whether there is a standby out-patient department of the same department on the same floor. When there is a standby out-patient department of the same department on the same floor, the emergency processing module pushes the emergency shunting suggestion of the same floor to the patient. After the patient accepts the emergency shunting suggestion of the same floor, the patient is guided to the standby out-patient department of the same department through AR navigation technology. When there is no standby out-patient department of the same department on the same floor, the emergency processing module queries the nearest idle out-patient department of the same department, pushes the emergency shunting suggestion of the idle out-patient department to the patient, and plans the best path to the idle out-patient department after the patient accepts the emergency shunting suggestion of the idle out-patient department. The patient is guided to move to the idle out-patient department through AR navigation technology.
[0089] When a device failure occurs, the emergency handling module first automatically switches to the adjacent device networking and generates a maintenance work order, while scanning the non-faulty device to generate the best path to the non-faulty device, and then guiding the patient to the location of the non-faulty device through voice broadcast.
[0090] The intelligent suggestion module is used to provide customized suggestions based on the individual characteristics of the patient, for example, when the patient's input symptoms include fever, the patient is advised to walk through a well-ventilated passage.
[0091] The online dispensing module is used to provide online dispensing services for patients, and after the patient pays online, the online dispensing module sends dispensing information to the dispensing point, and the medical staff at the dispensing point dispenses the medicine according to the dispensing information, at which time the patient can directly go to the dispensing point to take the medicine, or the dispensing point sends the dispensed medicine to the patient.
[0092] The model establishment module integrates the dialect data and language data of countries around the world to establish a dialect database, a Chinese and foreign language database and a Chinese and foreign language switching model, so that the system can better serve patients with different language backgrounds and reduce the communication barriers caused by language barriers, thereby reducing the time consumption of patients in the process of guiding diagnosis.
[0093] Example 2:
[0094] Different from example 1, the real-time shunting module adopts a linear weight distribution method when distributing weights, and the initial weights of the n channels are w1, w2, …, w n , and When the weight needs to be adjusted according to the index x, the specific formula of the linear weight distribution is as follows:
[0095]
[0096] Wherein, the index x is the registration type index of the patient, w i ′ is the adjusted weight, w i is the initial weight, and alpha is used to control the influence degree of the index x on the weight, alpha>0, x i is the index value corresponding to the i th channel, is the average value of the index values of all channels, and the specific formula is as follows:
[0097]
[0098] Wherein, x j is the index value corresponding to the j th channel.
[0099] After the weight adjustment is completed, the weight is normalized to satisfy The normalization formula is as follows:
[0100]
[0101] wherein w i is the adjusted and normalized weight, is the sum of the adjusted weights of all channels.
[0102] Embodiment 3:
[0103] The difference between the embodiment 2 and the embodiment 3 is that the intelligent diagnosis guiding module adopts the shortest path algorithm to calculate the shortest path to the target point when planning the route, and the specific formula is as follows:
[0104] f(n) = g(n) + h(n) (4).
[0105] wherein g(n) is the actual cost from the starting point to the node n, and h(n) is the heuristic function from the node n to the end point, and the specific representation is as follows:
[0106] h(n) = |dx| + |dy| + k * |dz| (5).
[0107] wherein dx, dy and dz all represent the coordinate difference between the node and the end point, and k represents the weight factor of the floor height difference.
[0108] Obviously, the above embodiments are only examples for clearly illustrating, and are not the limitation to the embodiments. For the ordinary skilled in the art, other different forms of changes or variations can be made on the basis of the above description. Here, it is not necessary and also impossible to enumerate all the embodiments. The obvious changes or variations derived therefrom are still within the protection scope of the present application.
Claims
1. An intelligent triage and guidance system for outpatient clinics, characterized in that, Includes the following modules: The data acquisition module is used to acquire patient medical records, local dialect data, language data from around the world, 3D map data of the hospital, and symptom data of existing diseases. The model building module is used to build a disease database based on existing disease data, and to build disease models based on the disease database; to build a dialect database based on dialect data; to build a Chinese and foreign language database based on language data from around the world, and to build a Chinese and foreign language switching model based on the Chinese and foreign language database; and to construct a multi-layer weighted directed graph based on the hospital's 3D map data, and to build an in-hospital navigation model based on this graph. The case search module allows patients to input their symptoms. Based on the symptoms entered by the patient, the module searches for similar cases through a disease model and connects the found cases with medical staff to verify the accuracy of the case search. The case search module is also used for information exchange between medical staff to confirm the department to which the found cases belong. The intelligent triage module integrates multimodal interaction modes, including physical buttons, touch interaction, and voice interaction. It switches the interaction mode according to the patient's specific situation. At the same time, it provides intelligent navigation for patients based on AR technology and in-hospital navigation models, planning the best route. The intelligent triage module also provides patients with online queuing services and displays the estimated arrival time and estimated waiting time to patients in real time, and correlates and matches the estimated arrival time and estimated waiting time. The real-time triage module is used to monitor the flow of people in each department and each passage, generate a heat map, and send real-time triage suggestions to patients based on the heat map; The real-time triage module is also used to send detailed information about the expanded outpatient clinic to patients when the outpatient clinic of the same department expands its capacity. The emergency response module is used to locally cache 24-hour shift schedules and map data. When one or more of the following situations occur, such as network outage, queue exceeding limits, or equipment failure, corresponding emergency response plans will be formulated and implemented. The intelligent suggestion module is used to provide customized suggestions based on the patient's individual characteristics.
2. The outpatient triage intelligent guidance system according to claim 1, characterized in that, The data acquisition module includes the following units: The case data collection unit is used to collect patient case data by connecting to the HIS system; The dialect data acquisition unit is used to collect dialect data of the current region through the ASR speech recognition engine; The Chinese and foreign language data acquisition unit is used to collect language data from around the world through ParseHub; The map data acquisition unit is used to acquire 3D map data of the hospital through laser scanning point cloud data technology; The disease data collection unit is used to collect disease data of existing diseases through a large-scale Internet model.
3. The outpatient triage intelligent guidance system according to claim 2, characterized in that, The data acquisition module is also used to collect and organize high-frequency words during patient consultations, enabling the model building module to perform targeted dialectization of high-frequency words when building a dialect database, and to perform targeted translation of high-frequency words when building a Chinese-foreign language database.
4. The outpatient triage intelligent guidance system according to claim 3, characterized in that, When the intelligent triage module switches between interaction modes, its operation process includes the following steps: S1: Obtain patient medical records and their specific age through facial recognition. S2: Adaptively adjust the interaction mode according to the patient's age. When the patient is 15-60 years old, use the touch interaction mode; when the patient is 60-75 years old, use the physical button interaction mode; when the patient is over 75 years old or under 15 years old, use the voice interaction mode. S3: When a patient communicates with the system via voice in their dialect, the intelligent triage module identifies the patient's dialect type through the dialect database and switches to the corresponding dialect type to communicate with the patient via voice. When a patient communicates with the system via a foreign language, the intelligent triage module identifies the patient's foreign language through a Chinese-foreign language switching model and switches the operation mode accordingly, so that both the touch interaction content and the voice interaction content are converted to the same foreign language as the patient.
5. The outpatient triage intelligent guidance system according to claim 4, characterized in that, The intelligent triage module operates as follows when it correlates and matches estimated arrival time with estimated waiting time: If the estimated waiting time for the department a patient has booked is less than 15 minutes, and the patient expects to arrive at the department in less than 15 minutes, it is recommended that the patient go to the department immediately. If the estimated waiting time for the department a patient has booked is 15-30 minutes, and the patient's arrival time at the department is less than 30 minutes, it is recommended that the patient wait in the waiting area. When the estimated wait time for the department a patient has booked is greater than 30 minutes, it is recommended that the patient continuously monitor changes in the estimated wait time. During this period, when the serial number changes, the intelligent triage module immediately sends a notification push to the patient to provide a reminder.
6. The outpatient triage intelligent guidance system according to claim 5, characterized in that, The operation process of the real-time traffic splitting module includes the following steps: S01: Real-time monitoring of patient flow in each department and passageway, generating a heat map based on patient flow in each department and passageway, providing real-time display for patients, and setting a flow threshold to determine the specific status of departments and passageways based on the flow threshold. When the patient flow exceeds the flow threshold, it is determined that the department is at full capacity and the passageway is impassable. S02: When the outpatient department of the same department expands its capacity, the location of the expanded outpatient department is monitored in real time. Patients are assigned weights according to their registration type. Based on the weight allocation results, specific information about the expanded outpatient department is pushed to patients, and patients are asked if they are willing to go to the expanded outpatient department. When patients are willing to go to the expanded outpatient department, they are guided to the expanded outpatient department.
7. The outpatient triage intelligent guidance system according to claim 6, characterized in that, The emergency response plan in the emergency response module is as follows: When a network outage occurs, the emergency handling module will switch the system to offline mode, change the positioning mode of the intelligent navigation to Bluetooth Beacon fingerprint positioning, and store the data during the network outage locally. Once the network connection is restored, the offline operation log will be automatically uploaded. When queues exceed the limit, the emergency response module first checks if there is a backup outpatient clinic for the same department on the same floor. If there is, it sends an emergency diversion suggestion to the patient. If the patient accepts the suggestion, AR navigation technology guides the patient to the backup outpatient clinic. If there is no backup outpatient clinic on the same floor, the emergency response module searches for the nearest available outpatient clinic for the same department, plans the best route to the available outpatient clinic, and sends an emergency diversion suggestion to the patient. If the patient accepts the suggestion, AR navigation technology guides the patient to the available outpatient clinic. When equipment failure occurs, the emergency response module automatically switches to the network of adjacent devices and generates a maintenance work order. At the same time, it scans for non-faulty devices, generates the best path to the non-faulty devices, and guides the patient to the location of the non-faulty devices through voice broadcast.
8. The outpatient triage intelligent guidance system according to claim 7, characterized in that, It also includes an online medication pickup module, which provides patients with online medication pickup services. After the patient pays online, the online medication pickup module sends the medication pickup information to the pharmacy, and the medical staff at the pharmacy will dispense the medication for the patient based on the medication pickup information.
9. The outpatient triage intelligent guidance system according to claim 8, characterized in that, The real-time splitting module uses a linear weighting method to allocate weights, denoting the initial weights of the n channels as w1, w2, ..., w... n And make When it is necessary to adjust the weights according to the indicator x, the specific formula for linear weight allocation is as follows: Wherein, indicator x represents the patient's registration type, and w′ i For the adjusted weights, w i Let x be the initial weight, and α be the value used to control the influence of the index x on the weight. If α > 0, x i The index value corresponding to the i-th channel. The average value of all channel indicators is calculated using the following formula: Where, x j This is the index value corresponding to the j-th channel; After the weights are adjusted, they are normalized to meet the following conditions: Its normalization formula is as follows: Among them, w" i To adjust and normalize the weights, This is the sum of the adjusted weights for all channels.
10. The outpatient triage intelligent guidance system according to claim 9, characterized in that, When planning routes, the intelligent triage module uses a shortest path algorithm to calculate the shortest path to the target point. The specific formula is as follows: f(n)=g(n)+h(n)(4); Where g(n) is the actual cost from the starting point to node n, and h(n) is the heuristic function from node n to the endpoint, which is specifically represented as follows: h(n)=|dx|+|dy|+k*|dz| (5); Where dx, dy, and dz represent the coordinate differences between the node and the endpoint, and k represents the weighting factor for the floor height difference.