Indoor and outdoor integrated medical navigation system for hospital

Abstract

The invention relates to the technical field of intelligent medical treatment, in particular to a hospital indoor and outdoor integrated medical treatment navigation system, and provides the following scheme aiming at the problems that in the prior art, a hospital navigation system is low in indoor and outdoor positioning precision, and the coupling degree of a hospital real-time resource state and a patient personalized diagnosis and treatment process is low. The system comprises a special portable intelligent navigation terminal, a hospital intelligent resource scheduling server cluster, a multi-level fusion positioning infrastructure network and a distributed edge computing architecture, sub-meter precise positioning is realized through a multi-source fusion positioning technology, a diagnosis and treatment task chain is generated and optimized in real time in combination with a dynamic resource scheduling algorithm, multi-mode intelligent navigation interaction is realized based on context sensing, and functions of payment, near field communication and the like are integrated to form a closed-loop service. The problems that existing hospital navigation is low in positioning precision, single in function and disjointed with the diagnosis and treatment process are solved, intelligence, initialization and individuation of the medical treatment process are achieved, and the medical treatment experience of patients and hospital operation efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical technology, and in particular to an integrated indoor and outdoor medical navigation system for hospitals. Background Technology

[0002] Large general hospitals typically consist of multiple buildings with different functions, numerous departments, complex layouts, and crisscrossing corridors. For patients visiting for the first time or unfamiliar with the hospital environment, "difficulty finding their way" is a common and prominent problem. Patients often spend a lot of time asking for directions and searching for departments, which not only exacerbates their anxiety but also leads to chaos in the hospital's public areas.

[0003] Furthermore, the traditional medical process is linear and rigid, with patients holding paper guides and following a fixed order to complete steps such as registration, waiting, examination, payment, and medication collection. However, hospital operations are dynamic and subject to change. Unexpected events such as equipment malfunctions in a particular examination department, a surge in queues, or emergency patients interrupting the original process can lead to long periods of aimless waiting, resulting in the problem of "complex procedures" and "long waiting times."

[0004] Currently, several hospital navigation solutions exist on the market. One approach is a two-dimensional planar navigation app, which uses Wi-Fi or Bluetooth beacons for coarse-grained positioning and provides planar map navigation on a mobile app. However, its positioning accuracy is low (typically 3-10 meters), failing to provide precise guidance for the "last mile." Furthermore, its functionality is limited to route planning, completely disconnected from the hospital's core treatment processes, and unable to perceive or respond to real-time status changes. Another approach is static guidance systems, such as electronic screens and signs. These methods suffer from untimely information updates and cannot provide personalized, dynamic route planning for individual patients.

[0005] Therefore, existing technologies lack a closed-loop intelligent system that can deeply couple high-precision indoor and outdoor positioning, real-time hospital resource status, and personalized patient treatment processes, and perform global dynamic optimization and scheduling. To address this, this solution proposes an integrated indoor and outdoor hospital navigation system. Summary of the Invention

[0006] The present invention proposes an integrated indoor and outdoor medical navigation system for hospitals, which solves the problems of low indoor and outdoor positioning accuracy and low coupling between real-time hospital resource status and personalized diagnosis and treatment process of patients in the existing hospital navigation system.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A hospital indoor and outdoor integrated medical navigation system includes: multiple dedicated portable intelligent navigation terminals, a hospital intelligent resource scheduling server cluster, a multi-level fusion positioning infrastructure network, and a distributed edge computing architecture;

[0009] The hardware architecture of the dedicated portable intelligent navigation terminal includes a multi-layer motherboard integrating a main processor and a coprocessor, a multi-source fusion positioning module, a context-aware user interaction module, a secure communication module, a near-field interaction module, and an integrated payment module.

[0010] The multi-source fusion positioning module is used to seamlessly receive and process data from GPS / BeiDou satellite signals, UWB ultra-wideband base stations, Bluetooth beacons, Wi-Fi RTT access points and built-in IMU inertial sensors. It uses a hybrid algorithm based on extended Kalman filtering and particle filtering for data fusion and calculation, and has three adaptive working modes: outdoor dominant mode, indoor seamless mode and visual enhancement mode.

[0011] The context-aware user interaction module includes a high-resolution touch screen, a high-fidelity speaker, a noise-canceling microphone array, and a wide-angle camera for AR navigation. Its interaction logic is driven by a rule-based engine, which can automatically and intelligently switch between 2D planar navigation, AR real-scene navigation, and voice command navigation modes based on environmental complexity, patient movement status, and the urgency of the current task.

[0012] The hospital's intelligent resource scheduling server cluster is used to create, manage, and dynamically optimize treatment task chains for patients.

[0013] The multi-level fusion positioning infrastructure network is used to provide continuous positioning signals from meter level to sub-meter level, and achieves full indoor and outdoor coverage by deploying UWB base stations, Bluetooth beacons, Wi-Fi RTT access points and visual reference codes;

[0014] The distributed edge computing architecture includes edge computing nodes deployed on each main floor or in each department to handle real-time computing tasks in the area.

[0015] The above technical solutions enable deep fusion and adaptive switching of multi-source positioning data. Combined with a high-precision positioning infrastructure network, they break through the bottleneck of low accuracy in traditional navigation and positioning. At the same time, through the multi-module integration of smart terminals, they provide patients with a comprehensive service carrier that integrates navigation, interaction, and payment, laying the hardware foundation for a closed-loop service throughout the entire process.

[0016] As a further improvement to the above solution, the secure communication module supports multiple communication protocols such as 5G / 4G, Wi-Fi 6, and Bluetooth; the near-field interaction module supports NFC, QR code scanning, and RFID reading; and the integrated payment module has a built-in security chip to support contact / contactless payment.

[0017] Through the above technical solutions, the support of multiple communication protocols ensures stable connection of the terminal in the complex network environment of the hospital, the near-field interaction module enables rapid data interaction between the terminal and various hospital devices, and the integrated payment module ensures the security and convenience of the payment process during medical treatment, avoids patients queuing repeatedly to pay, and improves medical efficiency.

[0018] As a further improvement to the above solution, the software architecture of the hospital intelligent resource scheduling server cluster includes:

[0019] A patient task management engine for creating, updating, and managing a structured “treatment task chain” for each patient;

[0020] A hospital-wide real-time status monitoring and prediction module is used to continuously collect and analyze the number of people queuing in each department, the working status of doctors, and the operating status of equipment.

[0021] A dynamic path and process optimization algorithm library, whose core algorithm is an improved mixed integer programming model, used to globally and dynamically reconstruct the patient's "diagnosis and treatment task chain" based on real-time status and prediction information;

[0022] A high-precision indoor and outdoor integrated map service engine.

[0023] Through the above technical solutions, the core decision-making and scheduling system of the system was constructed. The patient task management engine realizes the structured management of the diagnosis and treatment process. The real-time status monitoring and prediction module of the whole hospital provides data support for dynamic optimization. The dynamic path and process optimization algorithm library ensures the scientific and optimal nature of task chain adjustment. The high-precision map service engine provides basic guarantee for accurate navigation.

[0024] As a further improvement to the above solution, the patient task management engine is used to create, update and manage a structured "treatment task chain" data structure for each patient, which includes task sequences, logical dependencies between tasks and binding relationships with physical resources.

[0025] The above technical solutions clarify the inherent logic and resource association of diagnosis and treatment tasks, enabling the dynamic adjustment of the task chain to strictly follow medical process specifications and avoid medical logic confusion caused by process reconstruction. At the same time, it achieves precise binding between tasks and hospital physical resources (such as consultation rooms, equipment, and windows), providing data support for the automatic confirmation mechanism.

[0026] As a further improvement to the above solution, the hospital-wide real-time status monitoring and prediction module is used to continuously collect and analyze the number of people queuing in each department, the working status of doctors, and the operating status of equipment, and predict the congestion situation at key nodes in the future through time series analysis or machine learning models.

[0027] The above technical solutions enable real-time perception and trend prediction of hospital operation status. This not only allows for timely response to disturbances that have occurred, but also for early prediction of potential congestion risks, providing a basis for proactive optimization of the task chain and shifting from passive response to proactive prevention.

[0028] As a further improvement to the above scheme, the core algorithm of the dynamic path and process optimization algorithm library is an improved mixed integer programming model, which is solved by a hybrid heuristic algorithm based on rolling time domain. It responds to resource unavailability, congestion, high-priority insertion and patient behavior deviation events and dynamically reconstructs the "diagnosis and treatment task chain".

[0029] Through the above technical solutions, the improved mixed integer programming model can quickly solve the optimal task chain scheme under multiple constraints (such as medical logic, resource capacity, and patient priority). The solution method based on rolling time domain ensures that the algorithm responds quickly to real-time state changes and effectively copes with various dynamic disturbance events in the hospital.

[0030] A hospital indoor and outdoor integrated medical navigation method includes the following steps: S1: The patient receives the dedicated portable smart navigation terminal and completes identity binding; the intelligent resource scheduling server generates an initial "treatment task chain" based on the patient information and the overall hospital resource status and pushes it to the dedicated portable smart navigation terminal.

[0031] S2: The dedicated portable intelligent navigation terminal calculates the optimal path based on the first task and the real-time fusion positioning results, and dynamically selects the most suitable navigation mode to guide the patient;

[0032] S3: When the patient arrives at the task node, the "arrival" is confirmed through an automatic confirmation mechanism, and the status of the dedicated portable intelligent navigation terminal and the hospital information system is updated simultaneously.

[0033] S4: The intelligent resource scheduling server continuously monitors the status of the entire hospital. When a disturbance event is detected, it triggers a dynamic optimization algorithm to calculate and push a new optimal task chain for the patient.

[0034] S5: After receiving the new task chain from the intelligent resource scheduling server, the dedicated portable intelligent navigation terminal actively notifies the patient and updates the navigation interface to guide the patient to the new destination.

[0035] S6: When encountering a payment step in the process, the patient can directly complete the payment through the integrated payment module of the dedicated portable smart navigation terminal, and the smart resource scheduling server can unlock subsequent tasks accordingly.

[0036] S7: After all tasks are completed, the dedicated portable smart navigation terminal guides the patient to return the device, and automatically performs a data erasure operation upon return.

[0037] Through the above technical solutions, a closed-loop service is built from terminal activation to device recycling, realizing personalized generation, dynamic optimization and precise execution of diagnosis and treatment task chains. Combined with context-aware navigation and seamless payment, it significantly reduces patient waiting time and unnecessary movements, and improves the smoothness and intelligence of the medical process.

[0038] As a further improvement to the above scheme, the real-time reconstruction of the task chain in step S4 follows a hierarchical response strategy, specifically including:

[0039] For minor disturbances, perform local fine-tuning, adjusting only the expected time of the task;

[0040] For moderate perturbations, perform sequence reconstruction and reorder tasks while satisfying medical logic.

[0041] For severe disturbances, implement resource replacement and find alternative resources within the facility that can provide the same services;

[0042] When the system cannot resolve the issue automatically, the process is interrupted and manual intervention is initiated, prompting the patient to seek human assistance.

[0043] Through the above technical solutions, differentiated response strategies are adopted for disturbance events of different degrees. While ensuring the continuity of the diagnosis and treatment process, the impact of disturbances on the patient's medical experience is minimized, and process chaos caused by excessive adjustments is avoided. At the same time, the robustness of the system is ensured through a manual intervention mechanism.

[0044] As a further improvement to the above solution, a group behavior analysis and resource prediction subsystem is also included. This system uses a machine learning model to learn the hospital's patient flow patterns and resource consumption patterns over a long period of time, in order to predict departmental congestion in advance and provide resource allocation suggestions to hospital management.

[0045] The above technical solutions extend the optimization from individual tasks to the scheduling of group resources. They can not only provide patients with more accurate and personalized navigation services, but also provide hospital management with data-driven resource allocation decision support, helping hospitals optimize resource allocation and improve overall operational efficiency.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] 1. By integrating GPS / BeiDou, UWB, Bluetooth, Wi-FiRTT and IMU data through multi-source fusion positioning technology and using a hybrid filtering algorithm, sub-meter level continuous positioning is achieved from outdoors to indoors, solving the problem of accurate guidance in the "last meter" and thus achieving the goal of improving positioning accuracy.

[0048] 2. Based on the real-time resource status and prediction information of the hospital, the diagnosis and treatment task chain is generated and dynamically reconstructed through an improved mixed integer programming model. A hierarchical response strategy is adopted to deal with various disturbance events, avoid patients waiting blindly, and thus achieve optimized intelligent dynamic process.

[0049] 3. By automatically switching between 2D, AR real-scene and voice navigation modes through a context-aware multimodal interaction engine, the system meets the usage needs of different patients, improves ease of use and inclusivity, and achieves personalized adaptation of interaction modes.

[0050] 4. It integrates functions such as navigation, task management, payment, and equipment recycling, covering the entire process from patient admission to discharge, forming a smooth and secure service loop, and achieving the goal of providing a closed-loop service throughout the entire process.

[0051] 5. By using global resource scheduling and group behavior analysis, the resource load of each department can be balanced, congestion can be predicted in advance, and resource allocation can be optimized to improve the hospital's service throughput and management level, thereby achieving the goal of optimizing hospital operation efficiency. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the composition of the medical navigation system in an embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram illustrating the composition of a dedicated portable intelligent navigation terminal in an embodiment of the present invention;

[0054] Figure 3 This is a schematic diagram illustrating the composition of a hospital intelligent resource scheduling server cluster in an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram illustrating the composition of a multi-layered fusion positioning infrastructure network in an embodiment of the present invention;

[0056] Figure 5 This is a flowchart illustrating the medical navigation method in an embodiment of the present invention.

[0057] Explanation of key symbols:

[0058] 100. Navigation System; 110. Dedicated Portable Intelligent Navigation Terminal; 111. Multi-layer Motherboard; 112. Multi-source Fusion Positioning Module; 113. Context-aware User Interaction Module; 114. Secure Communication Module; 115. Near-field Interaction Module; 116. Integrated Payment Module; 120. Hospital Intelligent Resource Scheduling Server Cluster; 121. Patient Task Management Engine; 122. Hospital-wide Real-time Status Monitoring and Prediction Module; 123. Dynamic Path and Process Optimization Algorithm Library; 130. Multi-layer Fusion Positioning Infrastructure Network; 131. UWB Base Station; 132. Bluetooth Beacon; 133. Wi-Fi RTT Access Point; 134. Visual Reference Code; 140. Distributed Edge Computing Architecture. Detailed Implementation

[0059] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0060] Example 1:

[0061] Please combine Figure 1 This embodiment provides a hospital indoor-outdoor integrated medical navigation system 100, such as... Figure 1 As shown, the system includes multiple dedicated portable smart navigation terminals 110, a hospital intelligent resource scheduling server cluster 120, a multi-level fusion positioning infrastructure network 130, and a distributed edge computing architecture 140. The multiple dedicated portable smart navigation terminals 110 serve as personal mobile guidance and service centers for patients within the hospital, referred to as "Medical Guide-Tong Pro." The hospital intelligent resource scheduling server cluster 120 serves as the system's decision center, used to create, manage, and dynamically optimize treatment task chains for patients. The multi-level fusion positioning infrastructure network 130 is physically deployed in the hospital's indoor and outdoor environment to provide continuous positioning signals from meter-level to sub-meter-level accuracy. The distributed edge computing architecture 140 includes edge computing nodes deployed on major floors or in departments to handle real-time computing tasks within their respective areas.

[0062] like Figure 2 As shown, the hardware architecture of the dedicated portable intelligent navigation terminal 110 includes a multi-layer motherboard 111 integrating a main processor and a coprocessor, a multi-source fusion positioning module 112, a context-aware user interaction module 113, a secure communication module 114, a near-field interaction module 115, and an integrated payment module 116. The main processor of the multi-layer motherboard 111 is a Qualcomm QCS6490 with a built-in AI engine; the coprocessor of the multi-layer motherboard 111 is a specially programmed FPGA used to process multiple sensor signals in parallel to achieve low-latency sensor fusion.

[0063] The multi-source fusion positioning module 112 seamlessly receives and processes data from GPS / BeiDou satellite signals, UWB (Ultra-Wideband) base stations, Bluetooth beacons, Wi-Fi RTT (Wi-Fi Round-Trip Time) access points, and the built-in IMU (Inertial Measurement Unit) inertial sensor. The workflow of the multi-source fusion positioning module 112 includes data acquisition, EKF (Extended Kalman Filter) prediction, EKF update, and particle filter assistance. Data acquisition receives UWB ToF, Bluetooth AoA, Wi-Fi RTT distance, and IMU acceleration / angular velocity. EKF prediction: Based on the optimal estimate from the previous moment and IMU data, it predicts the current state (position, velocity, attitude). EKF update uses absolute position measurements provided by UWB, Wi-Fi RTT, etc., as observations to correct the predicted values. UWB data is given the highest observation weight due to its high accuracy. Particle filtering is used in areas with complex signals or severe multipath effects (such as elevator entrances). Particle filtering uses a set of random samples (particles) to represent the posterior probability distribution of the state, which better handles non-Gaussian noise. The particles are resampled based on observations, such as Bluetooth beacon area constraints and feature matching scores from visual SLAM (Simultaneous Localization and Mapping), resulting in a more robust localization output.

[0064] The multi-source fusion positioning module 112 employs a hybrid algorithm combining extended Kalman filtering and particle filtering to fuse and calculate position data from different sources. It features three adaptive operating modes: outdoor dominant mode, indoor seamless mode, and visual enhancement mode. In outdoor dominant mode, satellite signals are the primary source, supplemented by IMU data for dead reckoning. In indoor seamless mode, upon entering indoor environments, it switches to using UWB signals as a high-precision position reference, fusing Bluetooth beacon area markers, IMU data, and relative position corrections provided by visual SLAM. In visual enhancement mode, in areas with signal obstruction, the camera is activated to perform visual SLAM calculations, assisting in positioning by recognizing environmental features or visual reference codes.

[0065] The context-aware user interaction module 113 includes a high-resolution touchscreen display, high-fidelity speakers, a noise-canceling microphone array, and a wide-angle camera for AR (Augmented Reality) navigation. The interaction logic of the context-aware user interaction module 113 is driven by a rule-based engine, capable of automatically and intelligently switching between 2D planar navigation, AR real-scene navigation, and voice command navigation modes based on environmental complexity, patient movement status, and the urgency of the current task. The secure communication module 114 supports multiple communication protocols including 5G / 4G, Wi-Fi 6, and Bluetooth. The near-field interaction module 115 supports NFC (Near Field Communication), QR code scanning, and RFID (Radio Frequency Identification) reading. The integrated payment module 116 incorporates a security chip to support contact / contactless payment.

[0066] The software architecture of the hospital's intelligent resource scheduling server cluster 120 acts as the system's brain, responsible for optimizing global resources and dynamically reconstructing task chains; such as Figure 3 As shown, the hospital intelligent resource scheduling server cluster 120 includes a patient task management engine 121, a hospital-wide real-time status monitoring and prediction module 122, and a dynamic path and process optimization algorithm library 123. The patient task management engine 121 is used to create, update, and manage a structured "treatment task chain" data structure for each patient. This structure includes task sequences, logical dependencies between tasks, and binding relationships with physical resources. The hospital-wide real-time status monitoring and prediction module 122 is used to continuously collect and analyze the number of people queuing in each department, the doctors' work status, and the equipment operating status, and utilize time series analysis or machine learning... The model predicts congestion at key nodes over a future period; the core algorithm of the dynamic path and process optimization algorithm library 123 is an improved mixed integer programming model used to globally and dynamically reconstruct the patient's "treatment task chain" based on real-time status and prediction information. It employs a hybrid heuristic algorithm based on rolling time domain to solve the problem, responding to resource unavailability, resource congestion, high-priority insertion, and patient behavior deviation events, and dynamically reconstructing the "treatment task chain"; the software architecture of the hospital intelligent resource scheduling server cluster 120 can also provide a high-precision indoor and outdoor integrated map service engine.

[0067] like Figure 4 As shown, the multi-level integrated positioning infrastructure network 130, by deploying UWB base stations 131, Bluetooth beacons 132, Wi-Fi RTT access points 133 and visual reference codes 134, constitutes a high-precision, seamless, and continuous positioning network covering both indoor and outdoor environments.

[0068] Specifically, UWB base stations 131, Bluetooth beacons 132, and Wi-Fi 6RTT access points 133 are deployed both indoors and outdoors in the hospital, and a high-precision map is established. UWB base stations 131 are deployed in the hospital's outpatient hall, corridors on each floor, elevator lobbies, and entrances to key departments, with a grid interval of 15-20 meters. Each base station 131 is powered via Ethernet and synchronized with the edge computing nodes in the equipment room, achieving nanosecond-level accuracy, which is the basis for achieving decimeter-level positioning. The coordinates of base stations 131 are precisely measured using a total station during deployment and entered into the high-precision map service engine. Bluetooth beacons 132 supporting AoA technology are co-located next to all UWB base stations 131, and additional beacons are deployed at precise locations such as each consultation room door, medication window, and blood collection station for area wake-up and task confirmation.

[0069] Beacon 132 not only broadcasts its own ID, but also transmits information containing "spatial semantics" (such as "Internal Medicine Clinic 301" and "Blood Collection Window No. 3"). Wi-Fi 6RTT access point 133 has been upgraded to support the IEEE 802.11mc protocol in the entire hospital's Wi-Fi network. In areas where UWB signals may be weak, such as stairwells and restrooms, the density of access points has been increased to ensure the continuity of RTT positioning.

[0070] Visual reference codes 134 are placed in specific locations on walls or ceilings in long corridors lacking obvious features and at the connection points between different buildings. These QR codes 134 have precise 3D coordinates in the system map. When the terminal camera recognizes them, it can provide absolute pose correction for visual SLAM and prevent the accumulation of errors in IMU dead reckoning.

[0071] The distributed edge computing architecture 140 deploys computing nodes at the network edge to handle local high-concurrency, low-latency computing tasks, and works in collaboration with the cloud center to improve system response speed and reliability.

[0072] Edge servers are deployed in the main building's computer rooms for local positioning calculation, data fusion, and AR rendering. The "Medical Guide-Tong Pro" terminal uses a Qualcomm QCS6490 main processor and an FPGA coprocessor to run a hybrid fusion positioning algorithm based on EKF and PF (Particle Filter). Each server fuses raw UWB, Bluetooth, and Wi-Fi RTT data from all terminals in the building in real time to calculate the precise location, providing real-time rendering and push of AR navigation paths for terminals in the building. It also enables low-latency communication with the building's HIS (Hospital Information System) subsystem, queuing system, and smart gates.

[0073] The edge computing node deployment consists of one industrial-grade edge server each in the outpatient building, medical technology building, and inpatient department. Taking the edge server in the outpatient building as an example, its core tasks include real-time positioning and calculation, data fusion, and AR rendering services.

[0074] Real-time positioning calculation includes receiving raw signals from all UWB base stations 131 within the building, running the TDOA algorithm to initially calculate the terminal's position, and data fusion combining the initial UWB position, Bluetooth beacon area information, Wi-FiRTT ranging data, and raw IMU data uploaded by the terminal, running extended Kalman filtering locally to output a smooth, jitter-resistant final positioning result; AR rendering service generates virtual arrows and markers for AR navigation in real time based on the terminal's position and orientation, and streams them to the terminal, reducing the terminal's own computational load.

[0075] The system implements a deep binding and confirmation mechanism of "task chain - resource". Specifically, each sub-task in the "diagnosis and treatment task chain" is bound to a specific physical resource and time window on the server side. When the user or patient navigates to the vicinity of the target resource point through the terminal, the terminal can automatically confirm with the resource point through at least one of the following methods: precise positioning trigger, NFC contact, and QR code scanning, and upload the confirmation signal to the server in real time to update the task status.

[0076] Example 2:

[0077] Combination Figure 5 This embodiment is an improvement on embodiment 1, further described in the following aspects:

[0078] This embodiment provides a method for integrated indoor and outdoor medical navigation in hospitals. Please refer to... Figure 5 As shown, this medical navigation method includes the following steps:

[0079] S1: Terminal activation and personalized task chain initialization: The patient receives the dedicated portable smart navigation terminal and completes identity binding. The intelligent resource scheduling server generates an initial "treatment task chain" based on the patient information and the overall hospital resource status and pushes it to the dedicated portable smart navigation terminal.

[0080] When a patient swipes their medical insurance card to collect medication at the terminal, the server retrieves their appointment information from the HIS (Hospital Information System) and, combined with real-time status (predicting congestion in the ultrasound department), generates an initial optimal treatment task chain using a mixed integer programming model: outpatient -> electrocardiogram -> payment -> medication collection.

[0081] S2: Context-aware intelligent navigation execution: The dedicated portable intelligent navigation terminal calculates the optimal path based on the first task and the real-time fusion positioning results, and dynamically selects the most suitable navigation mode to guide the patient;

[0082] When in use, the terminal guides patients to the outpatient clinic, uses 2D+voice navigation in the lobby, and automatically switches to AR real-view navigation at complex intersections, displaying virtual arrows on the ground.

[0083] S3: Automatic perception and status synchronization of task nodes: When a patient arrives at a task node, an automatic confirmation mechanism is used to confirm the "arrival" and the status of the dedicated portable intelligent navigation terminal and the hospital information system is updated synchronously.

[0084] When a patient arrives at the clinic door, the terminal automatically triggers an "arrival" confirmation via the UWB electronic fence and notifies the doctor's workstation.

[0085] S4: Global dynamic monitoring and real-time reconstruction of task chains: The intelligent resource scheduling server continuously monitors the status of the entire hospital. When a disturbance event is detected, it triggers a dynamic optimization algorithm to calculate and push a new optimal task chain for the patient.

[0086] In step S4, the real-time reconstruction of the task chain follows a hierarchical response strategy, including local fine-tuning, sequence reconstruction, resource replacement, and process interruption and manual intervention. For minor disturbances, local fine-tuning is performed, adjusting only the expected time of the tasks; for moderate disturbances, sequence reconstruction is performed, reordering the tasks while satisfying medical logic; for severe disturbances, resource replacement is performed, searching for alternative resources within the hospital that can provide the same service; when the system cannot resolve the issue automatically, process interruption and manual intervention are performed, prompting the patient to seek human assistance.

[0087] S5: Terminal active intervention and guidance: After receiving a new task chain from the intelligent resource scheduling server, the dedicated portable intelligent navigation terminal actively notifies the patient and updates the navigation interface to guide the patient to the new destination.

[0088] If a diagnostic device malfunctions while a patient is waiting for their appointment, the server will detect this and trigger an optimization algorithm to redesign the patient's task chain (such as changing the diagnostic item or location). The terminal will then vibrate and display a prompt to guide the patient to the new destination.

[0089] S6: Integrated Services and Seamless Payment: When encountering a payment step in the process, the patient can directly complete the payment through the integrated payment module of the dedicated portable smart navigation terminal, and the smart resource scheduling server can unlock subsequent tasks accordingly;

[0090] After the doctor writes a prescription, a payment notification pops up on the terminal, and the user patient completes the medical insurance settlement through the terminal's built-in secure payment module.

[0091] S7: Process closed loop and data security recycling: After all tasks are completed, the dedicated portable intelligent navigation terminal guides the patient to return the device, and automatically performs data erasure operation when returning it;

[0092] After all tasks are completed, the terminal guides the user / patient to the return point, and the device automatically erases the usage data and disinfects.

[0093] The method also includes a group behavior analysis and resource prediction subsystem, which uses a machine learning model to learn the hospital's patient flow patterns and resource consumption patterns over a long period of time, in order to predict departmental congestion in advance and provide resource allocation suggestions to hospital management.

[0094] Let's take a specific user patient as an example for a detailed explanation:

[0095] S1: Terminal activation and personalized task chain initialization;

[0096] Patients swipe their medical insurance cards at the self-service terminal in the outpatient hall to receive a "Medical Guide-Pass Pro" device. After successfully binding their identity, the hospital's intelligent resource scheduling server cluster 120 receives the terminal activation event. The patient task management engine retrieves the patient's appointment information (9:00, Department of Cardiology, Director Wang) from the HIS system via API and checks the status of their registration slip. At the same time, the hospital's real-time status monitoring and prediction module starts working (it obtains that there are currently 12 people queuing in the ultrasound department and 3 people queuing in the electrocardiogram room at 8:45). Based on the LSTM (Long Short-Term Memory) model, it predicts that if the patient goes for a cardiac ultrasound after 9:30, the waiting time will exceed 40 minutes.

[0097] Simultaneously, the dynamic path and process optimization algorithm library is invoked, and its improved mixed-integer programming model begins calculation. The objective function is minZ = Σ(waiting time) + Σ(movement distance) + α*(resource load balancing index). After calculation, the model outputs the initial optimal task chain as follows:

[0098] [T1] 08:50-09:10: Cardiovascular Medicine Outpatient Clinic (Room 401, Area A, 4th Floor) → [T2] 09:20-09:40: Electrocardiogram (ECG) Examination (ECG Examination, Area B, 3rd Floor) Figure 2 [T3]: Payment -> [T4]: Pharmacy (1st floor)

[0099] Decision logic: To avoid long waiting times for color Doppler ultrasound, the system prioritizes scheduling relatively free electrocardiogram (ECG) examinations after the doctor's consultation. This decision is made during initialization.

[0100] S2: Context-aware intelligent navigation execution:

[0101] Upon receiving the task chain, the terminal immediately begins guiding the user / patient to consultation room 401 on the 4th floor: Initial stage (lobby): The environment is open and there is a large flow of people. The context-aware engine (based on the rule: IF area == "lobby" AND flow of people > threshold THEN mode = 2D + voice) activates 2D planar navigation and provides a clear voice prompt: "Please go to the elevator hall ahead and take the elevator to the 4th floor." Complex intersection (4th floor elevator entrance): The user / patient exits the elevator and faces an intersection with three forks. The engine detects the semantic label of the complex intersection and automatically switches to AR real-scene navigation. Through the camera, a green virtual arrow and sign are superimposed on the real scene. Specifically, the user / patient sees a glowing green virtual arrow superimposed on the ground in the real scene, clearly pointing to the middle passage, while a sign for "Cardiovascular Medicine Area A" is also floating in the air.

[0102] S3: Automatic detection and state synchronization of task nodes:

[0103] When the user patient follows the AR guidance to the door of clinic room 401, the terminal enters the electronic fence (within 3 meters) of the UWB base station deployed next to the door. The terminal triggers an automatic "arrival" signal to the edge node through precise positioning. At the same time, the edge node forwards this signal to the central server, causing the server to update the T1 task status to "arrived" and push a message to the computer in Dr. Wang's clinic room through the HIS interface: "Patient Li XX is waiting for the consultation"; at the same time, the user patient's terminal screen displays: "You have arrived, please wait patiently for your number to be called".

[0104] S4: Global dynamic monitoring and real-time reconstruction and proactive guidance of task chains:

[0105] If a sudden malfunction occurs in the ultrasound equipment during the waiting period (resource unavailability event), causing a backlog of patients originally scheduled for the ultrasound department, the hospital-wide status monitoring module will detect this and predict that the ECG room will become crowded in 20 minutes due to patient diversion. The optimization algorithm library will be triggered again to dynamically reconstruct the task chain for all affected patients. For the patient, the algorithm will determine that she should go to the ECG room immediately after completing her consultation to avoid the upcoming congestion peak.

[0106] S5: Terminal Active Intervention and Guidance: After the user / patient finishes seeing the doctor, the terminal vibrates gently, and an optimization prompt pops up on the screen, specifically: "Plan has been optimized! An ECG examination has been scheduled for you earlier. Please proceed to the ECG area on the 3rd floor, section B, immediately." Figure 2 The system displays the name of the ECG room and provides voice prompts while simultaneously updating navigation paths to guide patients to the ECG room.

[0107] S6: Integrated Services and Seamless Payments

[0108] The doctor issues a prescription in the HIS, and the user's task chain is automatically added to the [T3] payment task. A payment notification pops up on the terminal screen, showing the drug list and amount. Clicking "Pay Now" completes the medical insurance settlement through fingerprint verification using the terminal's integrated payment module. After successful payment, the server unlocks the [T4] medication pickup task.

[0109] S7: Process Closed Loop and Secure Data Recovery

[0110] After the patient completes all examinations and picks up their medication, the terminal displays a summary interface stating "All treatment tasks have been completed" and prompts the user to "Return the device to any self-service return locker." The terminal then guides the patient to place the device into the return locker at the exit. The locker automatically disinfects the contact areas of the terminal with ultraviolet light and communicates with the server to trigger a security wipe command, deleting all temporary data from this use (such as location tracking and payment tokens). The device is then restored to factory settings and ready to serve the next patient.

[0111] In this embodiment, local fine-tuning can be performed in the graded response strategy (to deal with different disturbances in S4). Specifically, the graded response strategy for task chain reconstruction manifests as follows: For minor disturbances (two additional people queuing in the ECG room), the system only updates the estimated waiting time on the user's terminal from "5 minutes" to "8 minutes"; for moderate disturbances (temporary upgrade of the pharmacy system), sequence reconstruction is performed, with an estimated interruption of 15 minutes. The system adjusts the order of the user's [T3] payment and [T4] medication pickup, guiding her to undergo the examination first and then pick up the medication. For severe disturbances (ECG room...), the system adjusts the order of the patient's [T3] payment and [T4] medication pickup, guiding her to undergo the examination first and then pick up the medication. Figure 2 (For temporary room disinfection), the system can also perform resource replacement, reassigning the user's patient's tasks to an idle "ECG" slot. Figure 3 In addition to the "ECG room", the system can also interrupt the process and allow manual intervention. If all ECG rooms are full, the system will prompt: "All ECG examination resources are full. Please go to the service center or click here to connect to the online triage for help."

[0112] The group behavior analysis and resource prediction subsystem learns the hospital's patient flow patterns through machine learning models. For example, it discovers that the peak time for the laboratory is from 10:00 to 11:00 a.m. on Mondays. Therefore, when generating task chains for patients on Monday mornings, it proactively avoids scheduling laboratory tasks during this time period and prioritizes recommending imaging examinations that are available at that time, thus achieving "prevention of congestion before it occurs".

[0113] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A hospital indoor and outdoor integrated medical navigation system, characterized in that, include: Multiple dedicated portable intelligent navigation terminals, hospital intelligent resource scheduling server clusters, multi-level fusion positioning infrastructure network, and distributed edge computing architecture; The hardware architecture of the dedicated portable intelligent navigation terminal includes a multi-layer motherboard integrating a main processor and a coprocessor, a multi-source fusion positioning module, a context-aware user interaction module, a secure communication module, a near-field interaction module, and an integrated payment module. The multi-source fusion positioning module is used to seamlessly receive and process data from GPS / BeiDou satellite signals, UWB ultra-wideband base stations, Bluetooth beacons, Wi-Fi RTT access points and built-in IMU inertial sensors. It uses a hybrid algorithm based on extended Kalman filtering and particle filtering for data fusion and calculation, and has three adaptive working modes: outdoor dominant mode, indoor seamless mode and visual enhancement mode. The context-aware user interaction module includes a high-resolution touch screen, a high-fidelity speaker, a noise-canceling microphone array, and a wide-angle camera for AR navigation. Its interaction logic is driven by a rule-based engine, which can automatically and intelligently switch between 2D planar navigation, AR real-scene navigation, and voice command navigation modes based on environmental complexity, patient movement status, and the urgency of the current task. The hospital's intelligent resource scheduling server cluster is used to create, manage, and dynamically optimize treatment task chains for patients. The multi-level fusion positioning infrastructure network is used to provide continuous positioning signals from meter level to sub-meter level, and achieves full indoor and outdoor coverage by deploying UWB base stations, Bluetooth beacons, Wi-Fi RTT access points and visual reference codes; The distributed edge computing architecture includes edge computing nodes deployed on each main floor or in each department to handle real-time computing tasks in the area.

2. The hospital indoor-outdoor integrated medical navigation system according to claim 1, characterized in that, The secure communication module supports multiple communication protocols such as 5G / 4G, Wi-Fi 6, and Bluetooth. The near-field interaction module supports NFC, QR code scanning, and RFID reading. The integrated payment module has a built-in security chip to support contact / contactless payment.

3. The hospital indoor-outdoor integrated medical navigation system according to claim 1, characterized in that, The software architecture of the hospital intelligent resource scheduling server cluster includes: A patient task management engine for creating, updating, and managing a structured "treatment task chain" for each patient; A hospital-wide real-time status monitoring and prediction module is used to continuously collect and analyze the number of people queuing in each department, the working status of doctors, and the operating status of equipment. A dynamic path and process optimization algorithm library, whose core algorithm is an improved mixed integer programming model, used to globally and dynamically reconstruct the patient's "diagnosis and treatment task chain" based on real-time status and prediction information; A high-precision indoor and outdoor integrated map service engine.

4. The hospital indoor and outdoor integrated medical navigation system according to claim 1, characterized in that, The patient task management engine is used to create, update and manage a structured "treatment task chain" data structure for each patient. This structure includes task sequences, logical dependencies between tasks and binding relationships with physical resources.

5. The hospital indoor and outdoor integrated medical navigation system according to claim 1, characterized in that, The hospital-wide real-time status monitoring and prediction module is used to continuously collect and analyze the number of people queuing in each department, the working status of doctors, and the operating status of equipment, and to predict the congestion situation at key nodes in the future through time series analysis or machine learning models.

6. The hospital indoor and outdoor integrated medical navigation system according to claim 1, characterized in that, The core algorithm of the dynamic path and process optimization algorithm library is an improved mixed integer programming model, which is solved using a hybrid heuristic algorithm based on rolling time domain. It responds to resource unavailability, congestion, high-priority insertion, and patient behavior deviation events and dynamically reconstructs the "diagnosis and treatment task chain".

7. A medical navigation method based on the medical navigation system according to any one of claims 1-4, characterized in that, Includes the following steps: S1: The patient receives the dedicated portable smart navigation terminal and completes identity binding. The smart resource scheduling server generates an initial "treatment task chain" based on the patient information and the overall hospital resource status and pushes it to the dedicated portable smart navigation terminal. S2: The dedicated portable intelligent navigation terminal calculates the optimal path based on the first task and the real-time fusion positioning results, and dynamically selects the most suitable navigation mode to guide the patient; S3: When the patient arrives at the task node, the "arrival" is confirmed through an automatic confirmation mechanism, and the status of the dedicated portable intelligent navigation terminal and the hospital information system is updated synchronously. S4: The intelligent resource scheduling server continuously monitors the status of the entire hospital. When a disturbance event is detected, it triggers a dynamic optimization algorithm to calculate and push a new optimal task chain for the patient. S5: After receiving the new task chain from the intelligent resource scheduling server, the dedicated portable intelligent navigation terminal actively notifies the patient and updates the navigation interface to guide the patient to the new destination. S6: When encountering a payment step in the process, the patient can directly complete the payment through the integrated payment module of the dedicated portable smart navigation terminal, and the smart resource scheduling server can unlock subsequent tasks accordingly. S7: After all tasks are completed, the dedicated portable smart navigation terminal guides the patient to return the device, and automatically performs a data erasure operation upon return.

8. The hospital indoor and outdoor integrated medical navigation method according to claim 5, characterized in that, The real-time reconstruction of the task chain in step S4 follows a hierarchical response strategy, specifically including: For minor disturbances, perform local fine-tuning, adjusting only the expected time of the task; For moderate perturbations, perform sequence reconstruction and reorder tasks while satisfying medical logic. For severe disturbances, implement resource replacement and find alternative resources within the facility that can provide the same services; When the system cannot resolve the issue automatically, the process is interrupted and manual intervention is initiated, prompting the patient to seek human assistance.

9. The hospital indoor and outdoor integrated medical navigation method according to claim 5, characterized in that, It also includes a group behavior analysis and resource prediction subsystem. This system uses machine learning models to learn the hospital's patient flow patterns and resource consumption patterns over a long period of time, in order to predict departmental congestion in advance and provide resource allocation suggestions to hospital management.

Images