A care bed monitoring system

Abstract

The application discloses a kind of nursing sickbed monitoring systems, including sickbed main body, multi-source sensing monitoring module, bed posture adjustment module, edge processing control module, cloud data management module and terminal early warning interaction module.The nursing sickbed monitoring system provided by the application can adapt to the nursing needs of different groups of people such as critically ill patients, postoperative patients, elderly disabled patients and rehabilitation patients, supports real-time updating of medical orders, and when clinical order parameters change, the AI model can automatically adapt to update the preliminary review rules and alarm thresholds without manual debugging.The whole process realizes intelligent, accurate and closed-loop nursing monitoring, effectively reduces the nursing error rate, saves nursing labor costs, and significantly improves the safety and intelligence level of bedridden patient care.

Description

Technical Field

[0001] This invention relates to a nursing bed monitoring system. Background Technology

[0002] Currently, most intelligent bed monitoring systems used in clinical and elderly care settings employ fixed parameter thresholds for monitoring and judgment. These system parameters are standardized and fixed at the factory, making it impossible to personalize and adjust them according to different patients' conditions, medical orders, and physical states. Existing equipment can only achieve simple data collection and over-limit alarm functions; that is, it directly triggers an alarm when the monitored data exceeds a fixed threshold. It lacks the ability for autonomous data preliminary review, anomaly classification and judgment, and risk categorization analysis, resulting in extremely low levels of intelligence.

[0003] In actual clinical nursing practice, there are significant differences in monitoring standards for patients with different conditions and nursing levels. For example, the bed pressure tolerance, bed exit control standards, and vital sign warning intervals are all different for postoperative immobilization patients, elderly disabled patients, and general rehabilitation patients. Traditional fixed threshold monitoring systems cannot match the individualized medical orders of healthcare professionals, and are prone to frequent false alarms, missed alarms, and invalid alarms. On the one hand, a large number of invalid alarms will increase the workload of healthcare professionals and waste nursing resources; on the other hand, key abnormal risks cannot be accurately identified and graded for early warning, which can easily lead to the inability to intervene in nursing risks such as pressure ulcers, falls from bed, and abnormal vital signs in a timely manner, posing a great risk to nursing safety.

[0004] Meanwhile, the existing monitoring system lacks a complete closed-loop logic for initial data review. It only completes data collection and simple comparison, and cannot conduct compliance review, anomaly quantification scoring, or risk classification of the monitoring data. It also cannot dynamically update the monitoring rules according to medical orders. The entire process relies on manual verification of data, review of anomalies, and judgment of risk levels. The manual initial review is labor-intensive and inefficient. Furthermore, manual judgment is subject to strong subjectivity, significant lag, and low error tolerance, making it difficult to meet the development needs of modern precision nursing and intelligent nursing.

[0005] In summary, existing nursing bed monitoring systems suffer from core deficiencies such as fixed and unadaptable parameters, lack of AI autonomous judgment capabilities, lack of a systematic preliminary review mechanism, single alarm mode, and inability to connect with clinical medical orders. There is an urgent need for a fully closed-loop intelligent nursing bed monitoring system that can adapt to individualized medical orders and has AI autonomous preliminary review and hierarchical progressive alarm capabilities. Summary of the Invention

[0006] This invention aims to provide a nursing bed monitoring system that solves the technical problems of traditional equipment, such as fixed threshold monitoring, inability to adapt to medical orders, lack of autonomous preliminary review capability, low alarm accuracy, and high dependence on manual intervention. It realizes fully automatic data preliminary review, risk classification, progressive alarm and adaptive nursing intervention based on medical order adaptation and AI intelligent judgment, and constructs a closed-loop intelligent nursing monitoring system for the entire process.

[0007] To address the above problems, the present invention provides a nursing bed monitoring system that adopts the following technical solution: A nursing bed monitoring system includes a bed body, a multi-source sensing and monitoring module, a bed posture adjustment module, an edge processing control module, a cloud data management module, and a terminal early warning and interaction module; The multi-source sensing and monitoring module and the bed posture adjustment module are both installed on the main body of the hospital bed. The edge processing control module is electrically connected to the multi-source sensing and monitoring module and the bed posture adjustment module respectively. The cloud data management module is wirelessly connected to the edge processing control module for receiving information. The terminal early warning interaction module is bidirectionally connected to the cloud data management module and the edge processing control module for displaying and issuing detection results. The multi-source sensing and monitoring module is used to collect real-time data on patient bed pressure distribution, vital signs, patient body movement and posture, bed posture, and ward environment. The edge processing control module is used to receive various raw data collected by the multi-source sensing and monitoring module, complete data filtering, noise reduction, feature extraction and threshold determination, identify pressure ulcer risk, bed fall risk, abnormal vital signs, bed rest time exceeding the limit and abnormal bed posture, and generate posture adjustment instructions and early warning signals. The bed posture adjustment module is used to receive posture adjustment commands from the edge processing control module and adaptively adjust the backrest lift angle, leg lift angle, and overall tilt height of the bed body. The cloud data management module is used to receive and store the monitoring data, abnormal records, and adjustment logs uploaded by the edge processing control module, complete the long-term care data trend analysis, risk level rating, and nursing report generation, and link with the medical record system to complete data archiving. The terminal early warning interaction module is used to receive early warning information from the edge processing control module and the cloud data management module, and to provide graded sound and light warnings, pop-up prompts and message pushes according to the risk level. It also supports medical staff to manually control the posture of the hospital bed and view real-time monitoring data and historical records.

[0008] Furthermore, the multi-source sensing and monitoring module includes a pressure sensing unit, a vital signs sensing unit, an attitude detection unit, and an environmental monitoring unit. The pressure sensing unit is a flexible pressure sensing array that is fitted and installed inside the mattress of the main body of the hospital bed, covering the pressure areas of the patient's head, back, buttocks and legs. It is used to collect global pressure distribution data and pressure duration data to identify local pressure overload and long-term unchanged body position of the patient. The vital signs sensing unit includes a heart rate sensor, a respiration sensor, and a body temperature sensor; all of which are fixed inside the headboard of the main body of the hospital bed, eliminating the need for patients to wear devices and enabling non-contact collection of real-time heart rate, respiratory rate, and body surface temperature data. The posture detection unit includes a six-axis gyroscope and an accelerometer, which are installed in the middle of the bed frame of the main body of the hospital bed. It is used to collect real-time tilt angle, lifting height data of the hospital bed, as well as patient body movement, turning over, and getting up movement data. The environmental monitoring unit includes a temperature and humidity sensor, an air quality sensor, and a noise sensor; it is installed on the headboard of the main body of the hospital bed and is used to monitor the temperature, humidity, air quality, and environmental noise parameters of the ward in real time.

[0009] Furthermore, the pressure sensing unit has a sampling frequency of 10-30Hz, which can accurately identify single-point pressure overload, bilateral pressure unevenness, and bed rest without body movement for an extended period. The pressure detection accuracy error does not exceed 3%, and it can distinguish between three basic states of the patient: bed rest, sitting up, and getting out of bed.

[0010] Furthermore, the bed posture adjustment module includes multiple sets of electric push rods, an angle detection component, and a drive control unit; the multiple sets of electric push rods correspond to the back lifting mechanism, leg lifting mechanism, and overall lifting mechanism of the main body of the hospital bed, respectively; the drive control unit receives the adjustment command from the edge processing control module and drives the electric push rods to extend and retract, precisely adjusting the posture of the hospital bed, with an angle adjustment accuracy of ≤0.5° and a height adjustment accuracy of ≤5mm.

[0011] Furthermore, the edge processing control module is internally equipped with a data preprocessing unit, an intelligent judgment unit, an instruction output unit, and a local storage unit; The data preprocessing unit is used to perform noise reduction filtering, outlier removal, and data normalization on the raw sensor data. The intelligent judgment unit incorporates an AI adaptive judgment model and a medical order parameter adaptation database, abandoning the traditional fixed threshold judgment mode. It supports the manual import of clinical medical order indicator parameters and can autonomously train and correct the judgment thresholds and review standards of various monitoring indicators based on patient age, symptoms, nursing level, and historical monitoring data. It completes individualized parameter adaptation for core monitoring indicators such as heart rate, respiration, body temperature, bed rest pressure, time out of bed, and duration of static posture. The AI ​​adaptive judgment model has complete preliminary data review logic, performs compliance review, quantitative scoring of abnormality, and accurate classification of risk types on real-time collected multi-source data, and triggers corresponding subsequent graded alarms and intelligent intervention strategies based on the preliminary review results, realizing unattended intelligent review and progressive early warning. The instruction output unit is used to output attitude adjustment instructions and graded early warning instructions based on the judgment results; The local storage unit is used to temporarily store real-time monitoring data and anomaly records within 24 hours. It can operate independently when the network is disconnected and automatically synchronizes to the cloud data management module when the network is connected.

[0012] Furthermore, the cloud-based data management module includes a data storage unit, a trend analysis unit, a risk rating unit, and a data archiving unit; The data storage unit is used for long-term storage of patient care monitoring data, bed operation data, early warning records, and adjustment logs; The trend analysis unit is used to generate patterns of patient position changes, vital sign fluctuation curves, and pressure ulcer risk trends based on historical data. The risk rating unit is used to combine real-time data and historical data to rate patient care risk into three levels: low, medium, and high, and generate personalized care recommendations. The data archiving unit is used to connect to the hospital's HIS and EMR systems, automatically complete the electronic archiving of nursing data, and generate daily nursing reports.

[0013] Furthermore, the terminal early warning interaction module includes a bedside touch terminal, a nurse station monitoring host, and a mobile app; The bedside touch terminal is installed at the head of the main body of the hospital bed and is used to display monitoring data in real time, provide local sound and light alarms, manually adjust the posture of the hospital bed, and view local nursing records. The aforementioned nurse station monitoring host is used to centrally display the monitoring status of all beds in the entire hospital or ward, uniformly receive early warning information, and view nursing data in batches; The aforementioned mobile app allows medical staff and family members to remotely view patient status, receive abnormal alerts, and achieve mobile nursing supervision.

[0014] Furthermore, the monitoring and early warning methods of this system include the following steps: S1. System initialization: Complete self-test of each module, sensor calibration, risk threshold parameter initialization, and bind corresponding patient information and nursing level. S2. The multi-source sensing and monitoring module continuously collects real-time data on patient pressure distribution, vital signs, body posture, bed posture, and ward environment, and transmits it to the edge processing control module. S3. The edge processing control module preprocesses the collected data and performs risk identification through the built-in intelligent judgment unit: when it detects that the patient's local pressure has exceeded the time limit or that the pressure is overloaded, it is judged as a risk of pressure ulcers, and a positional adjustment instruction and warning prompt are generated; when it detects that the patient gets up, moves around the bedside, or gets out of bed, it is judged as a risk of falling out of bed, and a graded warning is triggered; when the vital signs data exceed the normal threshold, a vital signs abnormality warning is triggered; when the duration of bed rest without body movement exceeds the standard, a turning reminder is triggered. S4. The edge processing control module, in conjunction with the bed posture adjustment module, completes adaptive posture adjustment based on the risk type, and simultaneously uploads abnormal data and early warning information to the cloud data management module. S5. The cloud-based data management module completes data storage, trend analysis, and risk rating, generates nursing suggestions, and synchronizes them to the terminal early warning interaction module. S6. The terminal early warning interaction module executes different early warning strategies according to the risk level: low-risk pop-up prompts, medium-risk sound and light warnings, and high-risk emergency message pushes. It also supports manual intervention by medical staff.

[0015] Furthermore, in step S3, the pressure ulcer risk determination logic is as follows: when the patient's continuous pressure on the same area exceeds a preset threshold and the pressure value is higher than the standard threshold, and there is no significant fluctuation in the data from three consecutive samplings, it is determined to be a high-risk state for pressure ulcers. The system automatically fine-tunes the bed angle to change the pressure point of the patient and pushes a turning care reminder. The bed-leaving risk determination logic is as follows: when the pressure sensing unit detects the disappearance of pressure across the entire mattress area and the posture detection unit detects the patient's body movement and getting up signal, a bed-leaving warning is immediately triggered, and the warning level is continuously upgraded when no one responds.

[0016] Furthermore, in step S3, the system completes a full preliminary review and subsequent alarm logic based on medical order parameters and AI model, specifically including three stages: indicator adaptation, data preliminary review, and graded alarm. Indicator adaptation process: The system reads the patient-specific medical order indicator range, contraindication parameters, and key nursing indicators entered by medical staff. The AI ​​model automatically replaces the system's default fixed thresholds and generates individualized monitoring and review standards for patients. In the preliminary data review stage, the AI ​​model conducts a dimensional compliance review of real-time stress data, vital sign data, body movement and posture data, and environmental data, distinguishing between four states: normal fluctuations, minor abnormalities, moderate abnormalities, and severe abnormalities. It then generates a preliminary review report, marking abnormal indicators, the degree of deviation, and potential nursing risks. Tiered alarm mechanism: Based on the initial review results, a progressive tiered alarm is triggered. When the initial review indicates normal fluctuations, the system silently records and continuously monitors. When the initial review indicates minor abnormalities, the system retains the record locally and generates a nursing task reminder, without triggering an emergency alarm. When the initial review indicates moderate abnormalities, a bedside audio-visual alarm and a pop-up notification at the nurse station are triggered to remind medical staff to review the information. When the initial review indicates severe abnormalities, an emergency alert is immediately pushed to the medical staff's mobile device, and the entire abnormal process data is recorded simultaneously. The logic for determining pressure ulcer risk is as follows: AI combines the turning interval set by the doctor's orders with the patient's pressure tolerance threshold to determine the state of continuous excessive pressure on the same area, automatically adjusts the bed angle and pushes a turning reminder; the logic for determining the risk of getting out of bed is as follows: AI combines the patient's condition and doctor's orders (postoperative immobilization, bed rest, etc.) to distinguish between the patient's normal sitting up, unintentional slipping, and unauthorized getting out of bed, and triggers corresponding level warnings accordingly.

[0017] The beneficial effects of this invention are as follows: The nursing bed monitoring system provided by this invention can be adapted to the nursing needs of different groups such as critically ill patients, postoperative patients, elderly disabled patients, and rehabilitation patients. It supports real-time updates of medical orders. When clinical medical order parameters change, the AI ​​model can automatically adapt and update the initial review rules and alarm thresholds without manual debugging. It realizes intelligent, precise, and closed-loop nursing monitoring throughout the process, effectively reducing the nursing error rate, saving nursing labor costs, and significantly improving the safety and intelligence level of nursing care for bedridden patients. Attached Figure Description

[0018] Appendix Figure 1 This is a schematic diagram of the system structure of the present invention.

[0019] Appendix Figure 2 This is a flowchart of the monitoring process of the present invention.

[0020] The system includes: 1. Bed body; 2. Multi-source sensing and monitoring module; 2-1. Pressure sensing unit; 2-2. Vital signs sensing unit; 2-3. Posture detection unit; 2-4. Environmental monitoring unit; 3. Bed posture adjustment module; 3-1. Electric push rod; 3-2. Angle detection component; 3-3. Drive control unit; 4. Edge processing control module; 4-1. Data preprocessing unit; 4-2. Intelligent judgment unit; 4-3. Command output unit; 4-4. Local storage unit; 5. Cloud data management module; 5-1. Data storage unit; 5-2. Trend analysis unit; 5-3. Risk rating unit; 5-4. Data archiving unit; 6. Terminal early warning interaction module; 6-1. Bedside touch terminal; 6-2. Nurse station monitoring host; 6-3. Mobile app. Detailed Implementation

[0021] To better understand the technical solution provided by this invention, the invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0022] As shown in the attached figures, in one embodiment, the nursing bed monitoring system provided by the present invention includes a bed body 1, a multi-source sensing and monitoring module 2, a bed posture adjustment module 3, an edge processing control module 4, a cloud data management module 5, and a terminal early warning and interaction module 6. The multi-source sensing and monitoring module 2 and the bed posture adjustment module 3 are both installed on the main body of the hospital bed 1. The edge processing control module 4 is electrically connected to the multi-source sensing and monitoring module 2 and the bed posture adjustment module 3 respectively. The cloud data management module 5 is wirelessly connected to the edge processing control module 4 for receiving information. The terminal early warning interaction module 6 is bidirectionally connected to the cloud data management module 5 and the edge processing control module 4 for displaying and issuing detection results. The multi-source sensing and monitoring module 2 is used to collect real-time data on patient bed pressure distribution, vital signs, patient body movement and posture, bed posture, and ward environment. The edge processing control module 4 is used to receive various raw data collected by the multi-source sensing and monitoring module 2, complete data filtering, noise reduction, feature extraction and threshold determination, identify pressure ulcer risk, bed fall risk, abnormal vital signs, bed rest time exceeding the limit and abnormal bed posture, and generate posture adjustment instructions and early warning signals. The bed posture adjustment module 3 is used to receive posture adjustment commands from the edge processing control module 4 and adaptively adjust the backrest lift angle, leg lift angle and overall tilt height of the bed body 1. The cloud data management module 5 is used to receive and store the monitoring data, abnormal records, and adjustment logs uploaded by the edge processing control module 4, complete the long-term care data trend analysis, risk level rating, and nursing report generation, and link with the medical record system to complete data archiving. The terminal early warning interaction module 6 is used to receive early warning information from the edge processing control module 4 and the cloud data management module 5, and to provide graded sound and light warnings, pop-up prompts and message pushes according to the risk level. It also supports medical staff to manually control the posture of the hospital bed and view real-time monitoring data and historical records.

[0023] Furthermore, the multi-source sensing and monitoring module 2 includes a pressure sensing unit 2-1, a vital signs sensing unit 2-2, an attitude detection unit 2-3, and an environmental monitoring unit 2-4; The pressure sensing unit 2-1 is a flexible pressure sensing array that is fitted and installed inside the mattress of the main body of the hospital bed 1, covering the pressure areas of the patient's head, back, buttocks and legs. It is used to collect global pressure distribution data and pressure duration data to identify local pressure overload and long-term unchanged body position of the patient. The vital signs sensing unit 2-2 includes a heart rate sensor, a respiratory sensor, and a body temperature sensor; all of which are fixed inside the headboard of the main body of the hospital bed 1, and do not require the patient to wear any devices, thus collecting the patient's real-time heart rate, respiratory rate, and body surface temperature data in a non-contact manner. The posture detection unit 2-3 includes a six-axis gyroscope and an accelerometer, which are installed in the middle of the bed frame of the main body of the hospital bed 1. It is used to collect real-time tilt angle, lifting height data of the hospital bed, as well as patient body movement, turning over, and getting up movement data. The environmental monitoring unit 2-4 includes a temperature and humidity sensor, an air quality sensor, and a noise sensor; it is installed on the headboard of the main body of the hospital bed 1 and is used to monitor the temperature, humidity, air quality, and environmental noise parameters of the ward in real time.

[0024] Furthermore, the pressure sensing unit 2-1 has a sampling frequency of 10-30Hz, which can accurately identify single-point pressure overload, bilateral pressure unevenness, and bed rest without body movement timeout. The pressure detection accuracy error does not exceed 3%, and it can distinguish between the three basic states of the patient: bed rest, sitting up, and getting out of bed.

[0025] Furthermore, the bed posture adjustment module 3 includes multiple sets of electric push rods 3-1, an angle detection component 3-2, and a drive control unit 3-3; the multiple sets of electric push rods 3-1 correspond to the back lifting mechanism, leg lifting mechanism, and overall lifting mechanism of the main body of the hospital bed 1, respectively; the drive control unit 3-3 receives the adjustment command from the edge processing control module 4 and drives the electric push rods 3-1 to extend and retract, accurately adjusting the posture of the hospital bed, with an angle adjustment accuracy of no more than 0.5° and a height adjustment accuracy of no more than 5mm.

[0026] Furthermore, the edge processing control module 4 is internally provided with an edge processing control module 4-1, an intelligent judgment unit 4-2, an intelligent judgment unit 4-3, and a local storage unit 4-4; The edge processing control module 4-1 is used to perform noise reduction filtering, outlier removal, and data normalization on the raw sensor data. The intelligent judgment unit 4-2 incorporates an AI adaptive judgment model and a medical order parameter adaptation database, abandoning the traditional fixed threshold judgment mode. It supports manual import of clinical medical order indicator parameters and can autonomously train and correct the judgment thresholds and review standards of various monitoring indicators based on patient age, symptoms, nursing level, and historical monitoring data. It completes individualized parameter adaptation for core monitoring indicators such as heart rate, respiration, body temperature, bed rest pressure, time out of bed, and time of stillness. The AI ​​adaptive judgment model has complete preliminary data review logic, performs compliance review, quantitative scoring of abnormality, and accurate classification of risk types on real-time collected multi-source data, and triggers corresponding subsequent graded alarms and intelligent intervention strategies based on the preliminary review results, realizing unattended intelligent review and progressive early warning. The intelligent judgment unit 4-3 is used to output attitude adjustment commands and graded early warning commands based on the judgment results; The local storage unit 4-4 is used to temporarily store real-time monitoring data and abnormal records within 24 hours. It can operate independently when the network is disconnected and automatically synchronizes to the cloud data management module 5 when the network is connected.

[0027] Furthermore, the cloud data management module 5 includes a data storage unit 5-1, a trend analysis unit 5-2, a risk rating unit 5-3, and a data archiving unit 5-4; The data storage unit 5-1 is used for long-term storage of patient care monitoring data, bed operation data, early warning records and adjustment logs; The trend analysis unit 5-2 is used to generate patterns of patient position changes, vital sign fluctuation curves, and pressure ulcer risk trends based on historical data. The risk rating unit 5-3 is used to combine real-time data and historical data to rate patient care risk into three levels: low, medium and high, and generate personalized care recommendations. The data archiving unit 5-4 is used to connect to the hospital's HIS and EMR systems to automatically complete the electronic archiving of nursing data and generate daily nursing reports.

[0028] Furthermore, the terminal early warning interaction module 6 includes a bedside touch terminal 6-1, a nurse station monitoring host 6-2, and a mobile app 6-3; The bedside touch terminal 6-1 is installed at the head of the bed body 1 and is used to display monitoring data in real time, provide local sound and light alarms, manually adjust the bed posture, and view local nursing records. The nurse station monitoring host 6-2 is used to centrally display the monitoring status of all beds in the entire hospital or ward, uniformly receive early warning information, and view nursing data in batches; The aforementioned mobile app 6-3 is used by medical staff and family members to remotely view the patient's status, receive abnormal alerts, and realize mobile nursing supervision.

[0029] Furthermore, the monitoring and early warning methods of this system include the following steps: S1. System initialization: Complete self-test of each module, sensor calibration, risk threshold parameter initialization, and bind corresponding patient information and nursing level. S2. The multi-source sensing and monitoring module 2 continuously collects real-time data on patient pressure distribution, vital signs, body posture, bed posture, and ward environment, and transmits it to the edge processing control module 4. S3. The edge processing control module 4 preprocesses the collected data and performs risk identification through the built-in intelligent judgment unit 4-2: when it detects that the patient's local pressure has exceeded the time limit or that the pressure is overloaded, it is judged as a risk of pressure ulcers and generates a body position fine-tuning instruction and early warning prompt; when it detects that the patient gets up, moves around the bedside, or gets out of bed, it is judged as a risk of falling out of bed and triggers a graded early warning; when the vital signs data exceed the normal threshold, it triggers a vital signs abnormality warning; when the duration of bed rest without body movement exceeds the standard, it triggers a turning reminder. S4. The edge processing control module 4, in conjunction with the bed posture adjustment module 3, completes adaptive posture adjustment according to the risk type, and uploads abnormal data and early warning information to the cloud data management module 5. S5. The cloud data management module 5 completes data storage, trend analysis and risk rating, generates nursing suggestions and synchronizes them to the terminal early warning interaction module 6. S6. The terminal early warning interaction module 6 executes different early warning strategies according to the risk level, including low-risk pop-up prompts, medium-risk sound and light warnings, and high-risk emergency message pushes, while also supporting manual intervention by medical staff.

[0030] Furthermore, in step S3, the pressure ulcer risk determination logic is as follows: when the patient's continuous pressure on the same area exceeds a preset threshold and the pressure value is higher than the standard threshold, and there is no significant fluctuation in the data from three consecutive samplings, it is determined to be a high-risk state for pressure ulcers. The system automatically fine-tunes the bed angle to change the pressure point of the patient and pushes a turning care reminder. The bed-leaving risk determination logic is as follows: when the pressure sensing unit 2-1 detects the disappearance of pressure across the entire mattress area and the posture detection unit 2-3 detects the patient's body movement and getting up signal, a bed-leaving warning is immediately triggered, and the warning level is continuously upgraded when no one responds.

[0031] Furthermore, in step S3, the system completes a full preliminary review and subsequent alarm logic based on medical order parameters and AI model, specifically including three stages: indicator adaptation, data preliminary review, and graded alarm. Indicator adaptation process: The system reads the patient-specific medical order indicator range, contraindication parameters, and key nursing indicators entered by medical staff. The AI ​​model automatically replaces the system's default fixed thresholds and generates individualized monitoring and review standards for patients. In the preliminary data review stage, the AI ​​model conducts a dimensional compliance review of real-time stress data, vital sign data, body movement and posture data, and environmental data, distinguishing between four states: normal fluctuations, minor abnormalities, moderate abnormalities, and severe abnormalities. It then generates a preliminary review report, marking abnormal indicators, the degree of deviation, and potential nursing risks. Tiered alarm mechanism: Based on the initial review results, a progressive tiered alarm is triggered. When the initial review indicates normal fluctuations, the system silently records and continuously monitors. When the initial review indicates minor abnormalities, the system retains the record locally and generates a nursing task reminder, without triggering an emergency alarm. When the initial review indicates moderate abnormalities, a bedside audio-visual alarm and a pop-up notification at the nurse station are triggered to remind medical staff to review the information. When the initial review indicates severe abnormalities, an emergency alert is immediately pushed to the medical staff's mobile device, and the entire abnormal process data is recorded simultaneously. The logic for determining pressure ulcer risk is as follows: AI combines the turning interval set by the doctor's orders with the patient's pressure tolerance threshold to determine the state of continuous excessive pressure on the same area, automatically adjusts the bed angle and pushes a turning reminder; the logic for determining the risk of getting out of bed is as follows: AI combines the patient's condition and doctor's orders (postoperative immobilization, bed rest, etc.) to distinguish between the patient's normal sitting up, unintentional slipping, and unauthorized getting out of bed, and triggers corresponding level warnings accordingly.

[0032] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit the technical solutions. Those skilled in the art should understand that any modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A nursing bed monitoring system, characterized in that, It includes the main body of the hospital bed, a multi-source sensing and monitoring module, a bed posture adjustment module, an edge processing control module, a cloud data management module, and a terminal early warning and interaction module; The multi-source sensing and monitoring module and the bed posture adjustment module are both installed on the main body of the hospital bed. The edge processing control module is electrically connected to the multi-source sensing and monitoring module and the bed posture adjustment module respectively. The cloud data management module is wirelessly connected to the edge processing control module for receiving information. The terminal early warning interaction module is bidirectionally connected to the cloud data management module and the edge processing control module for displaying and issuing detection results. The multi-source sensing and monitoring module is used to collect real-time data on patient bed pressure distribution, vital signs, patient body movement and posture, bed posture, and ward environment. The edge processing control module is used to receive various raw data collected by the multi-source sensing and monitoring module, complete data filtering, noise reduction, feature extraction and threshold determination, identify pressure ulcer risk, bed fall risk, abnormal vital signs, bed rest time exceeding the limit and abnormal bed posture, and generate posture adjustment instructions and early warning signals. The bed posture adjustment module is used to receive posture adjustment commands from the edge processing control module and adaptively adjust the backrest lift angle, leg lift angle, and overall tilt height of the bed body. The cloud data management module is used to receive and store the monitoring data, abnormal records, and adjustment logs uploaded by the edge processing control module, complete the long-term care data trend analysis, risk level rating, and nursing report generation, and link with the medical record system to complete data archiving. The terminal early warning interaction module is used to receive early warning information from the edge processing control module and the cloud data management module, and to provide graded sound and light warnings, pop-up prompts and message pushes according to the risk level. It also supports medical staff to manually control the posture of the hospital bed and view real-time monitoring data and historical records.

2. The nursing bed monitoring system according to claim 1, characterized in that, The multi-source sensing and monitoring module includes a pressure sensing unit, a vital signs sensing unit, an attitude detection unit, and an environmental monitoring unit. The pressure sensing unit is a flexible pressure sensing array that is fitted and installed inside the mattress of the main body of the hospital bed, covering the pressure areas of the patient's head, back, buttocks and legs. It is used to collect global pressure distribution data and pressure duration data to identify local pressure overload and long-term unchanged body position of the patient. The vital signs sensing unit includes a heart rate sensor, a respiration sensor, and a body temperature sensor; all of which are fixed inside the headboard of the main body of the hospital bed, eliminating the need for patients to wear devices and enabling non-contact collection of real-time heart rate, respiratory rate, and body surface temperature data. The posture detection unit includes a six-axis gyroscope and an accelerometer, which are installed in the middle of the bed frame of the main body of the hospital bed. It is used to collect real-time tilt angle, lifting height data of the hospital bed, as well as patient body movement, turning over, and getting up movement data. The environmental monitoring unit includes a temperature and humidity sensor, an air quality sensor, and a noise sensor; it is installed on the headboard of the main body of the hospital bed and is used to monitor the temperature, humidity, air quality, and environmental noise parameters of the ward in real time.

3. The nursing bed monitoring system according to claim 2, characterized in that, The sampling frequency of the pressure sensing unit is 10–30 Hz.

4. The nursing bed monitoring system according to claim 1, characterized in that, The bed posture adjustment module includes multiple sets of electric push rods, an angle detection component, and a drive control unit; the multiple sets of electric push rods correspond to the back lifting mechanism, leg lifting mechanism, and overall lifting mechanism of the main body of the hospital bed, respectively; the drive control unit receives the adjustment command from the edge processing control module and drives the electric push rods to extend and retract.

5. The nursing bed monitoring system according to claim 1, characterized in that, The edge processing control module is internally equipped with a data preprocessing unit, an intelligent judgment unit, an instruction output unit, and a local storage unit. The data preprocessing unit is used to perform noise reduction filtering, outlier removal, and data normalization on the raw sensor data. The intelligent judgment unit incorporates an AI adaptive judgment model and a medical order parameter adaptation database, abandoning the traditional fixed threshold judgment mode. It supports the manual import of clinical medical order indicator parameters and can autonomously train and correct the judgment thresholds and review standards of various monitoring indicators based on patient age, symptoms, nursing level, and historical monitoring data. It completes individualized parameter adaptation for core monitoring indicators such as heart rate, respiration, body temperature, bed rest pressure, time out of bed, and duration of static posture. The AI ​​adaptive judgment model has complete preliminary data review logic, performs compliance review, quantitative scoring of abnormality, and accurate classification of risk types on real-time collected multi-source data, and triggers corresponding subsequent graded alarms and intelligent intervention strategies based on the preliminary review results, realizing unattended intelligent review and progressive early warning. The instruction output unit is used to output attitude adjustment instructions and graded early warning instructions based on the judgment results; The local storage unit is used to temporarily store real-time monitoring data and anomaly records within 24 hours. It can operate independently when the network is disconnected and automatically synchronizes to the cloud data management module when the network is connected.

6. The nursing bed monitoring system according to claim 1, characterized in that, The cloud-based data management module includes a data storage unit, a trend analysis unit, a risk rating unit, and a data archiving unit. The data storage unit is used for long-term storage of patient care monitoring data, bed operation data, early warning records, and adjustment logs; The trend analysis unit is used to generate patterns of patient position changes, vital sign fluctuation curves, and pressure ulcer risk trends based on historical data. The risk rating unit is used to combine real-time data and historical data to rate patient care risk into three levels: low, medium, and high, and generate personalized care recommendations. The data archiving unit is used to connect to the hospital's HIS and EMR systems, automatically complete the electronic archiving of nursing data, and generate daily nursing reports.

7. The nursing bed monitoring system according to claim 1, characterized in that, The aforementioned terminal early warning interaction module includes a bedside touch terminal, a nurse station monitoring host, and a mobile app; The bedside touch terminal is installed at the head of the main body of the hospital bed and is used to display monitoring data in real time, provide local sound and light alarms, manually adjust the posture of the hospital bed, and view local nursing records. The aforementioned nurse station monitoring host is used to centrally display the monitoring status of all beds in the entire hospital or ward, uniformly receive early warning information, and view nursing data in batches; The aforementioned mobile app allows medical staff and family members to remotely view patient status, receive abnormal alerts, and achieve mobile nursing supervision.

8. A nursing bed monitoring system according to any one of claims 1-7, characterized in that, The monitoring and early warning method of the system includes the following steps: S1. System initialization: Complete self-test of each module, sensor calibration, risk threshold parameter initialization, and bind corresponding patient information and nursing level. S2. The multi-source sensing and monitoring module continuously collects real-time data on patient pressure distribution, vital signs, body posture, bed posture, and ward environment, and transmits it to the edge processing control module. S3. The edge processing control module preprocesses the collected data and performs risk identification through the built-in intelligent judgment unit: when it detects that the patient's local pressure has exceeded the time limit or that the pressure is overloaded, it is judged as a risk of pressure ulcers, and a positional adjustment instruction and warning prompt are generated; when it detects that the patient gets up, moves around the bedside, or gets out of bed, it is judged as a risk of falling out of bed, and a graded warning is triggered; when the vital signs data exceed the normal threshold, a vital signs abnormality warning is triggered; when the duration of bed rest without body movement exceeds the standard, a turning reminder is triggered. S4. The edge processing control module, in conjunction with the bed posture adjustment module, completes adaptive posture adjustment based on the risk type, and simultaneously uploads abnormal data and early warning information to the cloud data management module. S5. The cloud-based data management module completes data storage, trend analysis, and risk rating, generates nursing suggestions, and synchronizes them to the terminal early warning interaction module. S6. The terminal early warning interaction module executes different early warning strategies according to the risk level: low-risk pop-up prompts, medium-risk sound and light warnings, and high-risk emergency message pushes. It also supports manual intervention by medical staff.

9. A nursing bed monitoring system according to claim 8, characterized in that, In step S3, the pressure ulcer risk determination logic is as follows: when the patient's continuous pressure on the same area exceeds a preset threshold and the pressure value is higher than the standard threshold, and there is no significant fluctuation in the data from three consecutive samplings, it is determined to be a high-risk state for pressure ulcers. The system automatically fine-tunes the bed angle to change the pressure point of the patient and pushes a turning care reminder. The bed-leaving risk determination logic is as follows: when the pressure sensing unit detects the disappearance of pressure across the entire mattress area and the posture detection unit detects the patient's body movement and getting up signal, a bed-leaving warning is immediately triggered, and the warning level is continuously upgraded if no one responds.

10. A nursing bed monitoring system according to claim 8, characterized in that, In step S3, the system completes a full preliminary review and subsequent alarm logic based on medical order parameters and AI model, specifically including three stages: indicator adaptation, data preliminary review, and graded alarm. Indicator adaptation process: The system reads the patient-specific medical order indicator range, contraindication parameters, and key nursing indicators entered by medical staff. The AI ​​model automatically replaces the system's default fixed thresholds and generates individualized monitoring and review standards for patients. In the preliminary data review stage, the AI ​​model conducts a dimensional compliance review of real-time stress data, vital sign data, body movement and posture data, and environmental data, distinguishing between four states: normal fluctuations, minor abnormalities, moderate abnormalities, and severe abnormalities. It then generates a preliminary review report, marking abnormal indicators, the degree of deviation, and potential nursing risks. Tiered alarm process: Based on the initial review results, a progressive tiered alarm is triggered. When the initial review indicates normal fluctuations, the system silently records and continuously monitors. If the initial review indicates a minor abnormality, the system retains the record locally, generates a nursing task reminder, and does not trigger an emergency alarm. When the initial review indicates a moderate abnormality, a bedside audio-visual warning and a pop-up notification at the nurses' station are triggered to remind medical staff to review the case. When the initial review indicates a severe abnormality, an emergency warning is immediately pushed to the medical staff's mobile device, and the entire abnormality process data is recorded simultaneously.

Images