An online diagnosis and treatment system

Abstract

This invention relates to the field of medical emergency and telemedicine technology, and in particular to an online diagnosis and treatment system. Addressing the problems of lagging information transmission and lack of intelligent data analysis in existing traditional pre-hospital emergency care models, this invention proposes the following solution: a portable emergency intelligent terminal, a cloud-based big data and artificial intelligence analysis platform, a hospital-side command and preparation system, and a high-speed mobile communication network. The portable emergency intelligent terminal is deployed at the emergency site to collect and preprocess patient physiological and environmental data and transmit it in real time. After receiving the data, the cloud platform retrieves the patient's historical health data, diagnoses and evaluates the patient using an AI model, and generates a personalized dynamic emergency care plan based on a knowledge graph. This invention achieves real-time, multi-dimensional collection and intelligent analysis of pre-hospital emergency care data, breaks down information barriers between pre-hospital and in-hospital care, provides dynamic optimization solutions for on-site emergency care, guides precise in-hospital preparation, significantly improves emergency care efficiency and accuracy, and reduces the prognostic risks for critically ill patients.

Description

Technical Field

[0001] This invention relates to the field of medical emergency and telemedicine technology, and in particular to an online diagnosis and treatment system. Background Technology

[0002] Pre-hospital emergency care is the most critical link in the chain of survival for critically ill patients; its response speed and accuracy directly determine the patient's prognosis. Currently, the traditional pre-hospital emergency care model has many significant shortcomings, as follows: 1. Delayed and incomplete information transmission: Rescuers can only verbally describe the patient's condition to the hospital via radio or telephone. The information is incomplete, inaccurate, and discontinuous. Furthermore, it cannot transmit real-time physiological waveforms such as electrocardiogram and blood oxygenation, and the hospital cannot obtain complete on-site vital signs information of the patient. 2. Diagnosis relies on personal experience: The diagnosis of on-site emergency physicians is highly dependent on personal clinical experience. When faced with complex, rare or multi-system injuries, misjudgment or omission is very likely to occur, affecting the effectiveness of emergency treatment. 3. Lack of historical data support: At the emergency scene, it is impossible to immediately access the patient's complete electronic health record, past medical history, allergy history, medication history and other key information, which leads to blindness in the treatment work and is prone to secondary risks due to improper medication and treatment; 4. Passive in-hospital preparation: Hospital emergency departments usually only obtain limited patient information a few minutes before the patient arrives, making it difficult to prepare targeted and sufficient rescue resources based on the patient's condition, such as the allocation of specific operating rooms, specialist doctors, special drugs or equipment, resulting in a serious waste of the golden time for emergency treatment; 5. Lack of coordination between the scene and the hospital: There is a lack of real-time and efficient linkage between on-site emergency measures and subsequent in-hospital treatment plans. Pre-hospital and post-hospital diagnosis and treatment behaviors are difficult to form a unified treatment system, which affects the overall rescue effect.

[0003] In recent years, although remote monitoring devices capable of transmitting a limited number of vital signs have emerged, these devices can only achieve simple data telemetry and lack in-depth data fusion, intelligent analysis, and forward-looking decision support. They have failed to form a closed-loop intelligent emergency medical ecosystem from the scene to the hospital, and cannot fundamentally solve the aforementioned shortcomings of the traditional pre-hospital emergency care model. Therefore, this solution proposes an online diagnosis and treatment system. Summary of the Invention

[0004] The online diagnosis and treatment system proposed in this invention solves the problems of information lag and lack of intelligent data analysis in the traditional pre-hospital emergency care model.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: An online medical system includes: a portable emergency intelligent terminal, a cloud-based big data and artificial intelligence analysis platform, a hospital-side command and preparation system, and a high-speed mobile communication network; The portable emergency intelligent terminal is deployed at the emergency site to collect the patient's physiological and environmental data, and to perform local preprocessing and real-time transmission. The cloud-based big data and artificial intelligence analysis platform is deployed in the cloud and is used to receive and process data streams from the portable emergency intelligent terminal, retrieve patients' historical health data, run artificial intelligence models for real-time diagnosis and risk assessment, and generate personalized dynamic emergency rescue plans based on knowledge graphs. The hospital-side command and preparation system is deployed at the hospital and is used to receive early warning information, diagnostic results and emergency plans pushed by the cloud-based big data and artificial intelligence analysis platform, and to coordinate in-hospital emergency resources for proactive preparation. The high-speed mobile communication network connects the portable emergency medical terminal, the cloud-based big data and artificial intelligence analysis platform, and the hospital's command and preparation system for data transmission.

[0006] Through the above technical solutions, a closed-loop diagnosis and treatment system is built, covering the entire chain from the emergency scene to the cloud and then to the hospital. This system enables real-time collection, transmission, and intelligent analysis of pre-hospital emergency data, as well as seamless linkage between pre-hospital and in-hospital information. It breaks down the information silos of traditional pre-hospital emergency care, allowing on-site emergency care to receive cloud-based AI intelligent diagnostic support. At the same time, it enables hospitals to make targeted rescue preparations in advance, fundamentally solving the core problems of delayed information transmission and lack of intelligent analysis, and significantly improving the accuracy and efficiency of emergency care.

[0007] As a further improvement to the above solution, the portable emergency intelligent terminal includes a multimodal physiological data acquisition module, an edge computing and preprocessing module, an interaction and display unit, and a communication module; The multimodal physiological data acquisition module integrates medical-grade sensors for collecting electrocardiogram, blood pressure, blood oxygen, respiration and body temperature, and supports connection to portable ultrasound probes and end-tidal carbon dioxide monitors. The edge computing and preprocessing module has a built-in embedded AI chip, which is used to perform local preprocessing and feature extraction on the collected data, and run a lightweight AI model for local emergency event early warning. The interaction and display unit is used to display physiological waveforms, AI diagnostic prompts, emergency rescue plans and augmented reality animation guidance, and supports voice interaction; The communication module integrates a 5G or 6G communication module and supports network slicing and multi-link aggregation technologies.

[0008] The above technical solution enables medical-grade accurate collection of multi-dimensional physiological data of patients at the emergency scene. At the same time, it relies on edge computing capabilities to complete local data preprocessing and emergency early warning, ensuring basic emergency judgment even when the network is interrupted.

[0009] As a further improvement to the above solution, the cloud-based big data and artificial intelligence analysis platform includes a real-time streaming data processing engine, a multimodal AI diagnostic and prediction model library, a global health big data center, and an emergency knowledge graph and decision reasoning engine. The real-time streaming data processing engine is used to parse, align, and standardize the received real-time data stream to construct a real-time digital twin of the patient. The multimodal AI diagnostic and prediction model library includes deep learning-based pathology recognition models, injury assessment models, and dynamic risk assessment models. The comprehensive health big data center is used to securely retrieve patients' historical electronic health records, medication history, and allergy history through standard medical data interfaces; The emergency medical knowledge graph and decision reasoning engine are used to integrate real-time data, AI diagnostic results, and historical data to generate phased, personalized emergency medical treatment plans through reasoning.

[0010] Through the above technical solutions, high-speed processing and multi-dimensional fusion analysis of real-time pre-hospital data transmission are achieved. The patient's physiological state is presented intuitively by relying on the digital twin. Accurate real-time diagnosis and risk prediction are completed based on the well-trained multimodal AI model. Personalized treatment risks are avoided by combining the patient's historical health data. Finally, an executable emergency plan tailored to the individual patient's situation is generated through logical reasoning of the emergency knowledge graph.

[0011] As a further improvement to the above solution, the hospital-side command and preparation system includes a panoramic command dashboard, an intelligent early warning and resource allocation module, a digital handover system, and a remote expert consultation interface. The panoramic command dashboard is used to display the location, estimated arrival time, vital signs, AI diagnostic summary, and risk assessment level of emergency patients in real time. The intelligent early warning and resource allocation module is used to automatically send early warnings to relevant departments and generate a resource preparation list based on the AI ​​diagnosis results. The digital handover system is used to automatically connect all data and treatment records generated before hospitalization to the hospital's electronic medical record system. The remote expert consultation interface is used to support in-hospital experts to remotely view the scene and coordinate treatment through a secure video link.

[0012] Through the above technical solutions, hospitals can achieve real-time monitoring of the pre-hospital emergency care process in all dimensions. Relying on the intelligent early warning and resource allocation module, the in-hospital emergency preparation is transformed from "passive waiting" to "proactive planning". The digital handover system realizes seamless connection of pre-hospital and post-hospital data, avoiding the time waste caused by repeated data collection. The remote expert consultation interface allows high-quality medical resources in the hospital to directly empower on-site emergency care, making the hospital command and preparation system a "resource scheduling and coordination center" for pre-hospital and in-hospital collaborative treatment.

[0013] An online diagnosis and treatment method includes the following steps: S1: Emergency personnel arrive at the scene, use the portable emergency smart terminal to identify the patient and create an emergency conversation; S2: The portable emergency intelligent terminal continuously collects the patient's physiological data and uploads it to the cloud big data and artificial intelligence analysis platform in real time, while inputting on-site observation information; S3: The cloud-based big data and artificial intelligence analysis platform receives real-time data, retrieves the patient's historical health data, analyzes and diagnoses through AI models, and comprehensively generates the first version of a personalized emergency rescue plan. S4: Push the emergency plan to the portable emergency smart terminal to guide the on-site emergency personnel to execute it, and feed back the execution confirmation information to the cloud big data and artificial intelligence analysis platform; S5: The cloud-based big data and artificial intelligence analysis platform reassesses the patient's condition and the effectiveness of the plan based on the new data after execution, and dynamically generates and pushes the adjusted emergency plan; S6: The hospital-side command and preparation system receives information and coordinates the preparation of hospital resources starting from step S3, and completes the digital handover after the patient arrives. Through the above technical solutions, standardized and full-process online diagnosis and treatment execution steps have been established, realizing closed-loop management from patient identification and data collection and transmission, to cloud-based intelligent diagnosis, plan generation and push, to on-site execution and dynamic adjustment of the plan, and finally to in-hospital preparation and digital handover. This ensures that every link of pre-hospital emergency care is supported by data, guided by intelligence and coordinated, guaranteeing the standardization, continuity and accuracy of the emergency process, and maximizing the use of the golden time for emergency care.

[0014] As a further improvement to the above scheme, the specific process of generating the first version of the personalized emergency plan in step S3 is as follows: the decision reasoning engine integrates the analysis results of the real-time data stream by AI and the key information extracted from historical health data, and performs reasoning by searching the emergency knowledge graph to output an emergency plan that includes specific drugs, dosages, operations and in-hospital preparation suggestions, ensuring the completeness and feasibility of the emergency plan.

[0015] The above technical solution clarifies the generation logic and specific content of personalized emergency care plans. Through deep integration of multi-source data and professional reasoning of knowledge graphs, the generated emergency care plan not only matches the patient's real-time physiological state and personal health history, but also conforms to the professional standards of clinical emergency care. At the same time, the clear medications, dosages, operating procedures, and in-hospital preparation suggestions in the plan enable both on-site emergency personnel and hospital staff to execute it quickly, avoiding treatment errors caused by ambiguity in the plan.

[0016] As a further improvement to the above scheme, in step S5, if the patient's physiological indicators deviate from the expected response to the implementation plan by more than a set threshold, the AI ​​is triggered to re-analyze, consider other differential diagnoses, and generate a second version of the adjustment plan, forming a closed-loop optimization process of monitoring-analysis-decision-execution-re-monitoring, so as to realize the dynamic adjustment of the emergency plan and adapt to the changes in the patient's condition.

[0017] The above technical solutions enable emergency plans to move beyond static execution standards and instead dynamically optimize them based on real-time changes in the patient's condition. By setting thresholds, the plan adjustment can be automatically triggered, and the limitations of a single plan can be avoided by combining re-analysis and re-diagnosis processes.

[0018] As a further improvement to the above scheme, the threshold setting can be personalized according to the patient's acute and critical illness type, age, and underlying diseases, and can be fine-tuned by the hospital according to the needs of diagnosis and treatment.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Achieve accurate and intelligent on-site diagnosis: Extend cloud-based AI diagnostic capabilities to emergency sites, breaking through the limitations of human experience-based diagnosis and achieving "diagnosis upon boarding," significantly improving the accuracy and efficiency of on-site diagnosis, especially suitable for on-site identification of complex and rare diseases.

[0020] 2. Generate personalized dynamic emergency rescue plans: Integrate real-time physiological data and historical health big data of patients, generate exclusive emergency rescue plans based on emergency rescue knowledge graph and AI reasoning, and dynamically adjust them according to the patient's response to the plan to achieve precise emergency rescue with "one plan for one person" and avoid secondary risks caused by blind treatment; 3. Break down information barriers between pre-hospital and in-hospital care: Real-time linkage between pre-hospital emergency data and in-hospital treatment systems is achieved through high-speed mobile communication networks. In-hospital teams can obtain complete and accurate patient information before the patient arrives, and make precise allocation of rescue resources in advance, realizing "preparation before the patient arrives". This transforms rescue from "passive waiting" to "proactive response", greatly reducing the waste of golden rescue time.

[0021] 4. Enhance the operational capabilities of on-site emergency responders: Portable emergency smart terminals provide comprehensive operational support to on-site emergency responders through AR animation guidance, voice interaction, and edge intelligence warnings, enabling them to perform more complex and standardized emergency procedures and bridging the gap in professional skills among on-site emergency responders.

[0022] 5. Forming a virtuous cycle of technological iteration: High-quality emergency data generated during system operation can feed back into and optimize the AI ​​diagnostic model and emergency knowledge graph in the cloud, making the system smarter with use, continuously improving the overall level of emergency care, and promoting the intelligent development of the medical emergency industry.

[0023] 6. Achieve seamless data integration between pre-hospital and post-hospital care: The digital handover system automatically connects all pre-hospital emergency data and treatment records to the hospital's electronic medical record system, avoiding duplicate data collection after hospitalization, improving in-hospital reception efficiency, and providing complete and continuous emergency data support for patients' subsequent diagnosis and treatment. Attached Figure Description

[0024] Figure 1 is a schematic diagram of the overall composition of the online diagnosis and treatment system of the present invention; Figure 2 is a schematic diagram of the components of the portable emergency rescue smart terminal of the present invention; Figure 3 is a schematic diagram of the composition of the cloud-based big data and artificial intelligence analysis platform of the present invention; Figure 4 is a schematic diagram of the composition of the hospital-side command and preparation system of the present invention; Figure 5 is a schematic flowchart of the online diagnosis and treatment method of the present invention.

[0025] Explanation of key symbols: 100. Online diagnosis and treatment system; 110. Portable emergency intelligent terminal; 111. Multimodal physiological data acquisition module; 112. Edge computing and preprocessing module; 113. Interaction and display unit; 114. Communication module; 120. Cloud big data and artificial intelligence analysis platform; 121. Real-time streaming data processing engine; 122. Multimodal AI diagnosis and prediction model library; 123. Global health big data center; 124. Emergency knowledge graph and decision reasoning engine; 130. Hospital-side command and preparation system; 131. Panoramic command dashboard; 132. Intelligent early warning and resource allocation module; 133. Digital handover system; 134. Remote expert consultation interface; 140. High-speed mobile communication network. Detailed Implementation

[0026] 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.

[0027] Example 1: An online diagnosis and treatment system 100 in this example, such as Figure 1 As shown, the system includes a portable emergency medical terminal 110, a cloud-based big data and artificial intelligence analysis platform 120, a hospital-side command and preparation system 130, and a high-speed mobile communication network 140. The high-speed mobile communication network 140 enables high-speed and stable data interaction among the portable emergency medical terminal 110, the cloud-based big data and artificial intelligence analysis platform 120, and the hospital-side command and preparation system 130, thus constructing a closed-loop diagnosis and treatment system from the emergency scene to the cloud and then to the hospital.

[0028] A portable emergency medical terminal 110 is deployed at the emergency scene to collect patients' physiological and environmental data, perform local preprocessing, and transmit the data in real time. A cloud-based big data and artificial intelligence analysis platform 120 is deployed in the cloud to receive and process data streams from the portable emergency medical terminal 110, retrieve patients' historical health data, run artificial intelligence models for real-time diagnosis and risk assessment, and generate personalized dynamic emergency medical plans based on knowledge graphs. A hospital-side command and preparation system 130 is deployed at the hospital to receive early warning information, diagnostic results, and emergency medical plans pushed by the cloud-based big data and artificial intelligence analysis platform 120, and coordinate in-hospital rescue resources for proactive preparation. A high-speed mobile communication network 140 connects the portable emergency medical terminal 110, the cloud-based big data and artificial intelligence analysis platform 120, and the hospital-side command and preparation system 130 for data transmission.

[0029] The portable emergency medical terminal 110 includes a multimodal physiological data acquisition module 111, an edge computing and preprocessing module 112, an interaction and display unit 113, and a communication module 114; such as Figure 2 As shown, the multimodal physiological data acquisition module 111 integrates medical-grade sensors for collecting electrocardiogram, blood pressure, blood oxygen, respiration, and body temperature, and supports connection to a portable ultrasound probe and an end-tidal carbon dioxide monitor to achieve multi-dimensional vital sign acquisition; the edge computing and preprocessing module 112 has a built-in embedded AI chip for local preprocessing and feature extraction of the collected data, and runs a lightweight AI model for local emergency event warning; the interaction and display unit 113 provides visualization display and voice and AR interaction functions; the communication module 114 adopts a 5G / 6G communication module to ensure the priority and stability of data transmission.

[0030] like Figure 3As shown, the cloud-based big data and artificial intelligence analysis platform 120 includes a real-time streaming data processing engine 121, a multimodal AI diagnostic and prediction model library 122, a comprehensive health big data center 123, and an emergency knowledge graph and decision reasoning engine 124. The real-time streaming data processing engine 121 is used to parse, align, and standardize the received real-time data stream to construct a real-time digital twin of the patient. The multimodal AI diagnostic and prediction model library 122 includes deep learning-based pathology recognition models, injury assessment models, and dynamic risk assessment models. The comprehensive health big data center 123 is used to securely retrieve the patient's historical electronic health records, medication history, and allergy history through a standard medical data interface. The emergency knowledge graph and decision reasoning engine 124 is used to integrate real-time data, AI diagnostic results, and historical data to generate a phased, personalized emergency treatment plan through reasoning.

[0031] like Figure 4 As shown, the hospital-side command and preparation system 130 includes a panoramic command dashboard 131, an intelligent early warning and resource allocation module 132, a digital handover system 133, and a remote expert consultation interface 134. The panoramic command dashboard 131 is used to display the location, estimated arrival time, vital signs, AI diagnosis summary, and risk assessment level of emergency patients in real time. The intelligent early warning and resource allocation module 132 is used to automatically send early warnings to relevant departments and generate resource preparation lists based on AI diagnosis results. The digital handover system 133 is used to automatically connect all data and treatment records generated before hospitalization to the hospital's electronic medical record system. The remote expert consultation interface 134 is used to support in-hospital experts to remotely view the scene and coordinate treatment through a secure video link.

[0032] Example 2: Figure 5 As shown, this embodiment, based on embodiment 1, further improves upon the following: an online diagnosis and treatment method using the online diagnosis and treatment system in embodiment 1, comprising the following steps: S1: When emergency personnel arrive at the scene, they use the portable emergency smart terminal 110 to identify the patient's identity by scanning medical insurance cards, ID cards, etc., and create a unique emergency conversation to achieve personalized management of patient data.

[0033] S2: Emergency personnel continuously collect patient physiological data through the multimodal physiological data acquisition module 111 of the portable emergency intelligent terminal 110. After local preprocessing by the edge computing and preprocessing module 112, the data is uploaded in real time to the cloud big data and artificial intelligence analysis platform 120 by the communication module 114. At the same time, emergency personnel upload on-site observation information to the cloud platform through voice or manual input.

[0034] S3: The cloud-based big data and artificial intelligence analysis platform 120 receives and processes real-time data through the real-time streaming data processing engine 121. The whole-domain health big data center 123 retrieves the patient's historical health data. The multimodal AI diagnosis and prediction model library 122 runs AI models to perform real-time diagnosis and risk assessment. Finally, the emergency knowledge graph and decision reasoning engine 124 integrates multi-source data to generate the first version of a personalized emergency plan that includes medication, dosage, operation, and in-hospital preparation suggestions.

[0035] S4: The cloud platform pushes the generated emergency plan to the interaction and display unit 113 of the portable emergency smart terminal 110 to guide the on-site emergency personnel to perform the operation according to the plan; after the emergency personnel complete the operation, they will send the execution confirmation information back to the cloud big data and artificial intelligence analysis platform 120 through the terminal.

[0036] S5: The cloud-based big data and artificial intelligence analysis platform 120 reassesses the patient's condition and the effectiveness of the plan based on the new physiological data and on-site information uploaded after the plan is implemented. If the deviation between the patient's physiological indicators and the expected values ​​does not exceed the set threshold, the plan is fine-tuned. If the deviation exceeds the set threshold, AI is triggered to re-analyze and consider other differential diagnoses, generating a second version of the adjusted emergency plan and pushing it to the on-site terminal, forming a closed-loop optimization process.

[0037] S6: Starting from step S3, the hospital-side command and preparation system 130 synchronously receives patient warning information, diagnosis results, and emergency plans pushed by the cloud platform. The panoramic command dashboard 131 displays patient information in real time. The intelligent warning and resource allocation module 132 automatically sends warnings to relevant departments and generates a resource preparation list, coordinating in-hospital rescue resources to carry out proactive preparation. After the patient arrives at the hospital, the digital handover system 133 automatically connects all pre-hospital data and treatment records to the hospital's electronic medical record system to complete the digital information handover.

[0038] Example 3: This example is an improvement on Examples 1 and 2, focusing on its emergency application in the treatment of patients with acute ST-segment elevation myocardial infarction (STEMI). S1: The city's emergency medical center received a call for help from a 65-year-old male patient experiencing chest pain and profuse sweating. After the ambulance arrived at the scene, the emergency physician scanned the patient's medical insurance card using a portable emergency smart terminal 110. The identification was completed within 2 seconds. The portable emergency smart terminal 110 automatically initiated a binding request to the cloud-based big data and artificial intelligence analysis platform 120, creating a unique emergency session ID for this incident. Based on the location information, the estimated arrival time was initially set to 18 minutes.

[0039] S2: The emergency nurse puts an integrated wearable sensor patch and cuff on the patient. The multimodal physiological data acquisition module 111 continuously collects physiological data such as the patient's 12-lead electrocardiogram, non-invasive blood pressure (90 / 55 mmHg), blood oxygen saturation (95%), and respiratory rate (24 breaths / min). The environmental perception unit continuously uploads the ambulance location via GPS and the patient's general condition captured by the camera. The emergency physician verbally describes the patient's complaints through the voice interaction function and manually inputs the findings of the physical examination. All raw data are initially filtered and denoised by the edge computing and preprocessing module 112. The waveform is displayed locally and uploaded in real time to the cloud big data and artificial intelligence analysis platform 120 via 5G network slicing.

[0040] S3: The real-time streaming data processing engine of the cloud-based big data and artificial intelligence analysis platform 120 receives data streams in milliseconds, triggering parallel processes for AI pathological identification, historical data retrieval, dynamic risk assessment, and decision reasoning generation. The electrocardiogram analysis model in the multimodal AI diagnostic model library 122 outputs analysis results within 1.5 seconds, suggesting a possible acute inferior myocardial infarction complicated by right ventricular infarction. The whole-domain health big data center 123, under authorization, retrieves the patient's electronic health record in seconds, extracting information such as a 10-year history of type 2 diabetes, moderate stenosis in the mid-segment of the right coronary artery, aspirin allergy, and long-term use of aspirin. Key information such as clopidogrel; a dynamic risk assessment model predicts that the patient's risk of cardiogenic shock in the next 10 minutes is 35%; the emergency knowledge graph and decision reasoning engine 124 integrate the above information to generate the first version of a personalized emergency plan within 3 seconds, including on-site treatment plan (oral administration of ticagrelor 180mg, rapid infusion of normal saline, low-dose morphine analgesia, etc.) and in-hospital preparation suggestions (activating the emergency PCI process, prioritizing the radial artery route, preparing specific drugs, notifying relevant teams, etc.), and simultaneously push it to the portable emergency smart terminal 110 and the hospital command and preparation system 130.

[0041] S4: The emergency plan is displayed on the touchscreen of the portable emergency smart terminal 110 in the form of a list and flowchart. Augmented reality animation demonstrates the drug dispensing and verification operations. After the emergency doctor performs operations such as drug administration and establishing intravenous access, the execution confirmation information is fed back to the cloud big data and artificial intelligence analysis platform 120 through the portable emergency smart terminal 110. The system continuously monitors and shows that the patient's blood pressure rises to 105 / 65 mmHg and the heart rate drops to 90 beats / min. The cloud platform evaluates the plan as effective and reduces the risk of shock to 20%.

[0042] S5: The patient responded positively to the initial protocol, and the deviation of physiological indicators from the expected values ​​did not exceed the set threshold. The cloud-based big data and artificial intelligence analysis platform 120 only made minor adjustments to the monitoring strategy and did not trigger major changes in the protocol. It continuously monitored changes in the patient's physiological indicators in real time.

[0043] S6: The hospital-side command and preparation system 130 receives information from S3 onwards. The panoramic command dashboard 131 highlights the patient. The intelligent early warning and resource allocation module 132 automatically sends an activation command to the catheterization lab and generates a resource preparation list. The interventional team is assembled 10 minutes before the patient's arrival and reviews the patient's historical coronary CTA images. After the ambulance arrives at the hospital, the emergency physician initiates a handover request through the portable emergency smart terminal 110. The digital handover system 133 imports the complete pre-hospital emergency electronic medical record into the hospital's EMR system with one click. The patient is directly sent to the catheterization lab through a green channel. The time from the hospital gate to the guidewire passing through the infarcted blood vessel in this emergency was only 19 minutes, far below the international standard of 90 minutes.

[0044] Example 4: This example is an improvement upon Examples 1-3, focusing on the emergency application for patients with severe multiple injuries (traffic accidents). S1: At the scene of a high-speed car accident, emergency personnel found a 40-year-old male driver who was incoherent. The portable emergency smart terminal 110 quickly identified the patient's identity through the vehicle's ID card reader, created an emergency conversation, and initiated basic life support and advanced life support procedures.

[0045] S2: The multimodal physiological data acquisition module 111 collects physiological data such as the patient's heart rate of 132 beats / min, blood pressure of 85 / 50 mmHg, and blood oxygen saturation of 89% (under oxygen inhalation). The emergency physician connects the portable ultrasound probe to the portable emergency smart terminal 110 to complete the ultrasound assessment of key traumas. The device automatically records the accident location and time and marks the injury mechanism. The doctor verbally describes the on-site observation information and inputs it into the device. All data is preprocessed and uploaded in real time to the cloud big data and artificial intelligence analysis platform 120.

[0046] S3: The cloud-based big data and artificial intelligence analysis platform 120's image AI analysis module identifies ultrasound images in real time, indicating that the patient has intra-abdominal organ bleeding and the amount of bleeding is moderate; the physiological data integration and prediction model calculates the patient's trauma severity score as 29 points, predicting an 88% probability of hemorrhagic shock and the need for emergency surgery within the next 20 minutes; the whole-domain health big data center 123 retrieves information such as the patient's type O blood and no history of drug allergies; the emergency knowledge graph and decision reasoning engine 124 integrate the above information to generate the highest priority treatment plan. The on-site treatment plan includes initiating restrictive fluid resuscitation, external pelvic fixation, infusion of tranexamic acid, and preparation for endotracheal intubation, while the in-hospital preparation suggestions include activating the trauma team, implementing a large-volume transfusion plan, and allocating operating room and radiology resources.

[0047] S4: The on-site emergency team followed the instructions of the portable emergency smart terminal 110 to carry out the treatment plan. After tying the pelvic belt, administering tranexamic acid, and restricting fluid resuscitation, the patient's blood pressure only rose slightly and the heart rate remained fast. This signal was captured by the cloud big data and artificial intelligence analysis platform 120, which pushed supplementary prompts and suggested shortening the transfer time and sending the patient directly to the operating room.

[0048] S5: The cloud-based big data and artificial intelligence analysis platform 120, based on the patient's physiological indicators, determined that the patient was not responding well to the initial resuscitation plan, raised the urgency of the surgery to "extremely high", dynamically adjusted the in-hospital preparation recommendations, and required the blood bank and operating room to further accelerate resource preparation.

[0049] S6: The hospital-side command and preparation system 130 marks the patient as red with the highest priority. The intelligent early warning and resource allocation module 132 automatically completes operations such as calling the trauma team, preparing blood products, and allocating surgical equipment. After the ambulance arrives at the hospital, it stops directly at the entrance of the operating room. The emergency physician completes the handover with the trauma team by scanning a code through the portable emergency intelligent terminal 110. All information is synchronized in seconds. The time from patient admission to surgical incision is shortened to 22 minutes, realizing rapid treatment of patients with severe multiple injuries and greatly improving rescue efficiency.

[0050] 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. An online diagnosis and treatment system, characterized in that, include: Portable emergency medical terminal, cloud-based big data and artificial intelligence analysis platform, hospital command and preparation system, and high-speed mobile communication network; The portable emergency intelligent terminal is deployed at the emergency site to collect the patient's physiological and environmental data, and to perform local preprocessing and real-time transmission. The cloud-based big data and artificial intelligence analysis platform is deployed in the cloud and is used to receive and process data streams from the portable emergency intelligent terminal, retrieve patients' historical health data, run artificial intelligence models for real-time diagnosis and risk assessment, and generate personalized dynamic emergency rescue plans based on knowledge graphs. The hospital-side command and preparation system is deployed at the hospital and is used to receive early warning information, diagnostic results and emergency plans pushed by the cloud-based big data and artificial intelligence analysis platform, and to coordinate in-hospital emergency resources for proactive preparation. The high-speed mobile communication network connects the portable emergency medical terminal, the cloud-based big data and artificial intelligence analysis platform, and the hospital's command and preparation system for data transmission.

2. The online diagnosis and treatment system according to claim 1, characterized in that, The portable emergency intelligent terminal includes a multimodal physiological data acquisition module, an edge computing and preprocessing module, an interaction and display unit, and a communication module; The multimodal physiological data acquisition module integrates medical-grade sensors for collecting electrocardiogram, blood pressure, blood oxygen, respiration and body temperature, and supports connection to portable ultrasound probes and end-tidal carbon dioxide monitors. The edge computing and preprocessing module has a built-in embedded AI chip, which is used to perform local preprocessing and feature extraction on the collected data, and run a lightweight AI model for local emergency event early warning. The interaction and display unit is used to display physiological waveforms, AI diagnostic prompts, emergency rescue plans and augmented reality animation guidance, and supports voice interaction; The communication module integrates a 5G or 6G communication module and supports network slicing and multi-link aggregation technologies.

3. The online diagnosis and treatment system according to claim 1, characterized in that, The cloud-based big data and artificial intelligence analysis platform includes a real-time streaming data processing engine, a multimodal AI diagnostic and prediction model library, a global health big data center, and an emergency knowledge graph and decision reasoning engine. The real-time streaming data processing engine is used to parse, align, and standardize the received real-time data stream to construct a real-time digital twin of the patient. The multimodal AI diagnostic and prediction model library includes deep learning-based pathology recognition models, injury assessment models, and dynamic risk assessment models. The comprehensive health big data center is used to securely retrieve patients' historical electronic health records, medication history, and allergy history through standard medical data interfaces; The emergency medical knowledge graph and decision reasoning engine are used to integrate real-time data, AI diagnostic results, and historical data to generate phased, personalized emergency medical treatment plans through reasoning.

4. The online diagnosis and treatment system according to claim 1, characterized in that, The hospital-side command and preparation system includes a panoramic command dashboard, an intelligent early warning and resource allocation module, a digital handover system, and a remote expert consultation interface. The panoramic command dashboard is used to display the location, estimated arrival time, vital signs, AI diagnostic summary, and risk assessment level of emergency patients in real time. The intelligent early warning and resource allocation module is used to automatically send early warnings to relevant departments and generate a resource preparation list based on the AI ​​diagnosis results. The digital handover system is used to automatically connect all data and treatment records generated before hospitalization to the hospital's electronic medical record system. The remote expert consultation interface is used to support in-hospital experts to remotely view the scene and coordinate treatment through a secure video link.

5. The online diagnosis and treatment method of the online diagnosis and treatment system according to any one of claims 1-4, characterized in that, Includes the following steps: S1: Emergency personnel arrive at the scene, use the portable emergency smart terminal to identify the patient and create an emergency conversation; S2: The portable emergency intelligent terminal continuously collects the patient's physiological data and uploads it to the cloud big data and artificial intelligence analysis platform in real time, while inputting on-site observation information; S3: The cloud-based big data and artificial intelligence analysis platform receives real-time data, retrieves the patient's historical health data, analyzes and diagnoses through AI models, and comprehensively generates the first version of a personalized emergency rescue plan. S4: Push the emergency plan to the portable emergency smart terminal to guide the on-site emergency personnel to execute it, and feed back the execution confirmation information to the cloud big data and artificial intelligence analysis platform; S5: The cloud-based big data and artificial intelligence analysis platform reassesses the patient's condition and the effectiveness of the plan based on the new data after execution, and dynamically generates and pushes the adjusted emergency plan; S6: The hospital-side command and preparation system receives information and coordinates the preparation of hospital resources starting from step S3, and completes the digital handover after the patient arrives.

6. The online diagnosis and treatment method according to claim 5, characterized in that, The specific process of generating the first version of the personalized emergency plan in step S3 is as follows: the decision reasoning engine integrates the analysis results of the real-time data stream by AI and the key information extracted from historical health data, and performs reasoning by searching the emergency knowledge graph to output an emergency plan that includes specific drugs, dosages, operations and in-hospital preparation suggestions, ensuring the completeness and feasibility of the emergency plan.

7. The online diagnosis and treatment method according to claim 5, characterized in that, In step S5, if the patient's physiological indicators deviate from the expected response to the execution plan by more than a set threshold, the AI ​​is triggered to re-analyze, consider other differential diagnoses, and generate a second version of the adjustment plan. This forms a closed-loop optimization process of monitoring, analysis, decision-making, execution, and re-monitoring, enabling dynamic adjustment of the emergency plan to adapt to changes in the patient's condition.

8. The online diagnosis and treatment method according to claim 7, characterized in that, The threshold values ​​can be set individually based on the patient's type of acute or critical illness, age, and underlying diseases, and can be fine-tuned by the hospital according to the needs of diagnosis and treatment.

Images