Insulin pump safety reminding system based on remote communication

Abstract

The invention discloses an insulin pump safety reminding system based on remote communication, and the system comprises an insulin pump terminal which is used for collecting equipment operation parameters, and receiving and executing a verified operation instruction; the patient monitoring terminal is bridged with a data transfer interface, a local data encryption cache interface, a graded early warning display interface and a remote man-machine interaction interface of the insulin pump terminal and the hospital comprehensive management platform; the graded authorization user terminal receives early warning information related to the insulin pump terminal, checks the state of a patient and initiates a remote operation request; the hospital comprehensive management platform is in communication connection with each terminal, manages each terminal, performs data aggregation and intelligent analysis, generates early warning information based on a multi-dimensional rule, and performs authority management and auditing and multi-source data fusion application; and the safety communication and control module is used for guaranteeing data transmission safety, operation instruction compliance and system emergency fault-tolerant operation. According to the invention, cross-time-space safety protection and active early warning can be realized, the treatment safety and management efficiency are improved, and the method is suitable for medical and pension institutions at all levels.

Description

Technical Field

[0001] This invention relates to the field of smart healthcare and remote monitoring technology, and in particular to an insulin pump safety reminder system based on remote communication. Background Technology

[0002] As the primary device for continuous subcutaneous insulin infusion, the accurate and reliable operation of insulin pumps is crucial for stabilizing blood glucose levels in diabetic patients. While traditional insulin pumps have local audible and visual alarms (such as low medication level, tubing blockage, and low battery), they have significant shortcomings in practical clinical applications, especially in home settings: Alarm information isolation: The alarm is only triggered by the device itself. If the patient fails to notice it due to hearing impairment, sleep, temporary absence from the device, or being in a noisy environment, the alarm will fail, which may lead to serious hyperglycemia or hypoglycemia events, posing a significant safety hazard.

[0003] Lack of proactive early warning and trend judgment: Existing alarms are mostly passive responses based on instantaneous thresholds (such as when the battery level is below 10%), which cannot make early risk predictions based on historical data and operating trends (such as predicting the risk of insulin depletion or hypoglycemia based on the infusion pattern), thus lacking preventiveness.

[0004] Lack of outpatient monitoring: After a patient is discharged, their insulin pump data becomes an "information silo." The attending physician cannot obtain timely information about the patient's actual treatment adherence and the equipment's operational status, and can only review limited data during follow-up visits, thus losing the opportunity for continuous medical monitoring and timely intervention.

[0005] The lack of remote assistance capabilities: When patients (such as the elderly, children, or visually impaired individuals) encounter device alarms or need to perform complex operations (such as setting temporary baseline rates), family members or medical staff can only provide difficult and inefficient verbal guidance over the phone, which is prone to misunderstandings and operational errors. Currently, there is a lack of technical means to provide authorized personnel with remote status monitoring and assisted operation while ensuring safety and compliance.

[0006] Difficulty in multi-role collaboration: There is a lack of a shared and collaborative information platform among patients, their families, community doctors, hospital specialist nurses, and attending physicians, making it impossible to form an effective care synergy.

[0007] Therefore, the industry urgently needs an integrated system that can achieve real-time data aggregation, intelligent analysis and early warning, multi-terminal collaborative reminders, and support safe remote monitoring, in order to break through the time and space limitations of traditional insulin pump management and build a new model of full-process safety management centered on patients and supported by hospitals. Summary of the Invention

[0008] The purpose of this invention is to provide an insulin pump safety reminder system based on remote communication, which solves the technical problem that traditional insulin pump alarm information cannot reach patients, their families, medical teams, and other relevant parties in a timely and effective manner. Through a multi-terminal collaborative early warning mechanism, it ensures that alarm information is not missed and reduces safety risks.

[0009] To achieve the above objectives, the present invention provides an insulin pump safety reminder system based on remote communication, including an insulin pump terminal, a patient monitoring terminal, a tiered authorization user terminal, a hospital integrated management platform, and a safety communication and control center; The insulin pump terminal is used to collect device operating parameters and upload them via secure near-field communication, and to receive and execute verified operating instructions. The patient monitoring terminal communicates with the insulin pump terminal to bridge the data transfer between the insulin pump terminal and the hospital's integrated management platform, provide local data encryption and caching, display tiered early warnings, and enable remote human-computer interaction. Tiered authorization user terminals are used to receive alerts related to insulin pump terminals, view patient status, and initiate remote operation requests. The hospital's integrated management platform communicates with patient monitoring terminals and hierarchically authorized user terminals to access and manage various terminals, aggregate and intelligently analyze data, generate early warning information based on multi-dimensional rules, manage and audit permissions, and integrate multi-source data applications. Secure communication and control is used to ensure secure data transmission, compliant operation instructions, and fault-tolerant emergency operation of the system.

[0010] Preferably, the insulin pump terminal includes: The real-time data acquisition module is used to collect basic infusion parameters, equipment status parameters, abnormal signals, and identity and time synchronization. Among them, the basic infusion parameters include the current basic infusion rate, the temporary basic rate setting value and effective time, the high-dose infusion volume, the high-dose infusion time and frequency, and the current infusion mode. Device status parameters include the remaining insulin in the reservoir, battery level, and infusion line pressure; abnormal signals include alarm codes, fault types, and device fault status flags; identity and time synchronization includes recording the device's unique identifier (SN), patient-associated ID, and timestamp. The near-field communication module is used to establish a secure pairing connection with the patient monitoring terminal and upload data in a periodic and event-driven manner; the secure pairing connection adopts a two-way authentication mechanism of the device's unique identifier SN code and the patient monitoring terminal's MAC address; The local execution module is used to perform safety checks on received remote operation commands, including pausing infusion, adjusting the base rate, and venting the pipeline. It also uses the RSA-2048 algorithm to verify the digital signature of the operation command. Once the verification is successful, the corresponding operation is executed.

[0011] Preferably, the patient monitoring terminal includes: The data bridging module is used to receive and parse data from the insulin pump terminal. It uses the AES-256 encryption algorithm to locally encrypt and cache the parsed data, and uses the LZ77 compression algorithm to compress the data before transmitting it to the hospital's integrated management platform. The interactive interface module is used to visually display the status of the insulin pump terminal, graded warning information, and provide a remote operation confirmation and assistance request initiation interface. The remote operation confirmation is displayed in the form of a pop-up window, showing the identity of the operation initiator, the specific operation content, and the execution risk warning, and only provides two confirmation options: agree and refuse. The assistance request initiation interface has preset quick request buttons, including changing the pillbox, purging the tubing, parameter consultation, and emergency help. The local early warning module is used to trigger emergency early warnings based on local cached data and preset rules when the network is interrupted or the platform's early warning is delayed, and provides multi-mode reminders.

[0012] Preferably, the insulin pump terminal status includes the remaining insulin level, battery level, current infusion mode, and current basal rate; The tiered early warning information is divided into four levels according to risk level: emergency, high, medium, and low. When a patient experiences three consecutive infusion failures or a tubing pressure value ≥100 kPa, the situation is classified as an emergency. When the remaining drug dosage is ≤10U, the battery charge is ≤10%, or the pipeline pressure is ≥50kPa, it is judged as a high level; When the remaining drug dosage is ≤30U or the battery charge is ≤15%, it is judged as medium level; When the remaining medication is ≤50U, the battery level is ≤25%, the infusion mode is successfully switched, there is a data synchronization delay, or a consumable usage time warning appears, it is judged as a low level.

[0013] Preferred, tiered authorization user terminals include: Family member terminals and caregiver terminals are used to receive alerts from associated patients, view core safety parameters associated with patients, and initiate preset low-risk remote assistance requests. Core safety parameters include remaining insulin levels, battery level, current infusion mode, alert status, and key operation records. Low-risk remote assistance requests include pausing infusion and reminding to replace consumables. The healthcare terminal is used to view the complete treatment data trends of the patients under its care, initiate therapeutic operation suggestions, and manage patient groups. The complete treatment data trends include the basal rate change trend, high-dose infusion distribution, remaining drug volume change curve, battery power change curve, blood glucose correlation trend graph, and historical warning statistics. Therapeutic procedures include recommendations for adjusting basal rates, setting temporary basal rates, calculating high-dose infusions, and switching infusion modes; patient population management includes patient name, gender, age, current alert status, and the time of the most recent data synchronization. The technician terminal is used to view diagnostic information of the patient's insulin pump terminal, including communication status records, firmware version information, fault code history, and sensor calibration records. After authorization, it can initiate non-therapeutic device maintenance operations, including firmware updates, communication module resets, and sensor calibration parameter distribution.

[0014] Preferably, the hospital integrated management platform includes: The access and communication gateway module is used to authenticate and verify the identity of various connected terminals, handle high-concurrency requests, and standardize data. Each terminal must submit qualification documents, including the device serial number and user identity information, when it connects for the first time. After the review is approved, a unique identity identifier and access token are assigned. The data storage and management module is used to structure and store patients' static record data and dynamic treatment data. It employs a dual backup strategy of real-time off-site backup and scheduled full backup for data backup and recovery. Static record data includes basic patient information, medical history information, treatment plan and authorization relationship information; dynamic treatment data includes insulin pump operation data, blood glucose data, early warning records, operation logs and remote assistance records. The intelligent early warning and analysis engine module performs multi-dimensional risk prediction and analysis based on a three-level early warning rule base, and generates graded early warning information using time-series data analysis and risk prediction models. The authorization and security management module is used to manage the access and operation permissions of each terminal based on the role permission matrix, control authorization through a multi-level review process, and record the entire process audit log; The data fusion and application module is used to access multi-source health data, construct a comprehensive patient risk assessment model using the analytic hierarchy process (AHP), and analyze system operation quality indicators.

[0015] Preferably, the three-level early warning rule base includes: Equipment threshold rules are based on the threshold settings of the operating parameters of the insulin pump terminal, including the remaining drug volume and tubing pressure. Behavioral pattern rules identify abnormalities based on the patient's historical infusion behavior patterns, and use the K-means clustering algorithm to construct a normal infusion behavior model for the patient. When the real-time infusion data deviates from the normal model by more than a preset threshold ±25%, an early warning is triggered. Physiological model rules, which integrate insulin pump terminal data, continuous glucose monitoring data and patient behavior data, predict potential health risks through physiological metabolic models; The process of generating tiered early warning information is as follows: The sliding window algorithm is used to analyze continuous time series data. By calculating the mean, variance, and rate of change of the data within the window, the trend of data changes can be identified. A risk prediction model is built based on a Long Short-Term Memory (LSTM) neural network, which integrates operational data from insulin pump terminals, continuous glucose monitoring data, patient behavior data, and patient physiological parameters to predict the probability of patients experiencing hypoglycemia or hyperglycemia. An emergency warning is triggered when the probability of hypoglycemia is ≥70% or the probability of hyperglycemia is ≥80%. A high-level warning is triggered when the probability of hypoglycemia is 50%-70% or the probability of hyperglycemia is 60%-80%. A medium-level warning is triggered when the probability of hypoglycemia is 30%-50% or the probability of hyperglycemia is 40%-60%. A low-level warning is triggered when the probability of hypoglycemia is less than 30% and the probability of hyperglycemia is less than 40%.

[0016] Preferably, the roles in the role-based access control matrix management include patients, family members, nurses, doctors, technicians, and administrators. A role-based access control (RBAC) model is adopted to refine the data access permissions and operation permissions of each role. Authorization is controlled through a multi-level approval process, specifically as follows: S11. The patient initiates an authorization request through the patient monitoring terminal, selecting the authorized user, the scope of permissions, and the authorization validity period; S12. The authorization application is submitted to the hospital's comprehensive management platform, where the patient's attending physician or nurse conducts a second review. The review includes verifying the authenticity of the authorized user's identity and the reasonableness of the scope of authority. S13. After the review is approved, the hospital's integrated management platform activates the corresponding permissions, generates an authorization certificate, and issues it to the authorized user's terminal; if the review fails, the reason for rejection is provided, and the patient must resubmit the application. Record full-process audit logs, including terminal login or logout, data access, early warning triggering or processing, remote operation requests, confirmation or execution, authorization applications, review or revocation, and equipment maintenance operations; the audit log content includes operation time, operation subject, operation type, operation content, operation result, device identifier and IP address, and is stored in encrypted form using blockchain technology.

[0017] Preferably, the Analytic Hierarchy Process (AHP) is used to construct a patient health risk assessment model, which comprehensively scores patients from four dimensions: blood glucose control, treatment adherence, equipment operating status, and complication risk. If the overall score is ≤60, the patient is considered a high-risk patient; If the overall score is 60-80, the patient is classified as a medium-risk patient; If the overall score is ≥80, the patient is considered low-risk. If a patient is at high risk, they will be marked as a key monitoring subject and referred to the medical team for priority follow-up and intervention. The system also statistically analyzed key quality indicators of system operation and care services, including equipment alarm rate, accuracy rate of early warning rule triggering, remote assistance usage rate, success rate of intervention for high-risk patients, and patient follow-up completion rate.

[0018] Preferably, the secure communication and control center includes an encrypted communication module, an instruction security control module, and an emergency support module; The encrypted communication module uses the national standard SM4 encryption algorithm to encrypt the transmitted data end-to-end, and uses the ECDH algorithm with dynamic negotiation mechanism to generate encryption keys. The key transmission process is protected by the asymmetric encryption algorithm RSA-2048. Sensitive personal information of patients is also desensitized. The specific desensitization rules are as follows: the last character of the name is hidden, the middle 8 digits of the ID number are hidden, and the middle 4 digits of the contact information are hidden. The instruction security control module is used to perform digital signature verification, operation timeliness management, and secondary permission verification on remote operation instructions; digital signature verification includes instruction content, initiator identity, and timestamp. The emergency protection module is used to enable local monitoring and early warning mode when the network is interrupted, and to resume data transmission after the network is restored, as well as to monitor and intercept abnormal operations in real time.

[0019] Therefore, the present invention employs the above-mentioned insulin pump safety reminder system based on remote communication, and the beneficial effects are as follows: (1) This invention constructs a multi-terminal synchronous early warning mechanism for patients, their families and medical teams, integrates multiple channels such as APP push, SMS and voice call, and combines early warning upgrade strategies to ensure that alarm information is delivered in a timely manner; at the same time, it builds a shared information platform to achieve collaborative care among all parties, greatly reduce the safety risk of patients not being aware of alarms, and form a management synergy.

[0020] (2) This invention integrates AI algorithms with multi-source data fusion analysis to construct a three-level early warning rule base and an LSTM risk prediction model, which can predict potential risks such as hypoglycemia in advance. This realizes the transformation from "post-event remediation" to "pre-event prevention".

[0021] (3) This invention gathers out-of-hospital insulin pump data and health data in real time through the cloud, fills the gap in out-of-hospital continuous monitoring, and provides precise support for doctors to formulate personalized plans; it constructs a hierarchical authorization and remote assistance mechanism, requiring patients to make final confirmation for all operations, effectively solving the operation difficulties of special groups such as the elderly and improving the universality of the system.

[0022] (4) This invention integrates multiple security mechanisms such as end-to-end encryption and data desensitization, which comply with relevant regulations and are equipped with abnormal operation interception and network interruption fault tolerance functions to ensure system stability. The patient terminal adopts elderly-friendly design such as large font and voice broadcast, and the graded terminal interface is adapted to significantly improve ease of use and adaptability.

[0023] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0024] Figure 1 This is an overall system block diagram of an embodiment of an insulin pump safety reminder system based on remote communication according to the present invention; Figure 2 This is a block diagram of an insulin pump terminal structure according to an embodiment of an insulin pump safety reminder system based on remote communication according to the present invention; Figure 3 This is a flowchart of the multi-level review process control authorization process of an embodiment of an insulin pump safety reminder system based on remote communication according to the present invention; Figure 4 This is a block diagram of the safety communication and control hub mechanism of an insulin pump safety reminder system based on remote communication according to an embodiment of the present invention. Detailed Implementation

[0025] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0027] like Figure 1 As shown, an insulin pump safety reminder system based on remote communication includes an insulin pump terminal, a patient monitoring terminal, a tiered authorization user terminal, a hospital integrated management platform, and a safety communication and control center.

[0028] The insulin pump terminal is used to collect device operating parameters and upload them via secure near-field communication, and to receive and execute verified operating instructions.

[0029] The patient monitoring terminal communicates with the insulin pump terminal to bridge the data transfer between the insulin pump terminal and the hospital's integrated management platform, provide local data encryption and caching, display tiered early warnings, and enable remote human-computer interaction.

[0030] Tiered authorization user terminals are used to receive alerts related to insulin pump terminals, view patient status, and initiate remote operation requests.

[0031] The hospital's integrated management platform communicates with patient monitoring terminals and hierarchically authorized user terminals to access and manage various terminals, aggregate and intelligently analyze data, generate early warning information based on multi-dimensional rules, manage and audit permissions, and integrate multi-source data applications.

[0032] Secure communication and control is used to ensure secure data transmission, compliant operation instructions, and fault-tolerant emergency operation of the system.

[0033] like Figure 2 As shown, the insulin pump terminal includes a real-time data acquisition module, a near-field communication module, and a local execution module. The real-time data acquisition module is used to collect basal infusion parameters, device status parameters, abnormal signals, and identity and time synchronization. Among them, a high-precision flow sensor is used to collect basal infusion parameters, including the current basal infusion rate, temporary basal rate setpoint and effective time, bolus infusion volume, bolus infusion time and frequency, and current infusion mode. The acquisition accuracy is ≤ ±0.01 U / h, and the time synchronization accuracy is ≤ 1s.

[0034] The device integrates high-precision level, voltage, and pressure sensors to collect equipment status parameters, including the remaining insulin level in the reservoir, battery charge, and infusion tubing pressure. Abnormal signals include alarm codes, fault types, and equipment fault status flags. Identity and time synchronization includes recording the device's unique identifier (SN), patient-associated ID, and timestamp. The high-precision level sensor employs optical level detection technology, using infrared light reflection to achieve non-contact detection of the insulin level in the reservoir, avoiding contamination and accuracy degradation caused by direct contact with insulin.

[0035] The near-field communication module uses Bluetooth Low Energy (BLE) technology to establish a secure pairing connection with the patient monitoring terminal. The secure pairing connection uses a two-way authentication mechanism of the device's unique identifier (SN code) and the patient monitoring terminal's MAC address. During the pairing process, a temporary session key is generated based on the AES-128 algorithm. A packet-based transmission mechanism is used, including two transmission modes: periodic upload and event-driven upload. The upload period for periodic upload is set by the patient monitoring terminal. When the insulin pump terminal experiences critical events, including alarms, parameter changes, and malfunctions, the event-driven mode is triggered to upload data.

[0036] The local execution module is used to perform safety checks on received remote operation commands, including pausing infusion, adjusting basal rate, and venting tubing. It uses the RSA-2048 algorithm to verify the digital signature of the operation command. If the verification is successful, the corresponding operation is executed; if the verification fails, the operation is rejected and an error message is fed back. It also retains the traditional audible and visual alarm function, allowing the alarm volume to be set through the patient monitoring terminal and the alarm frequency to be adjusted according to the alarm level.

[0037] The patient monitoring terminal of this invention includes a data bridging module, an interactive interface module, and a local early warning module. The data bridging module establishes a long-term connection with the near-field communication module of the insulin pump terminal to receive and parse data from the insulin pump terminal and convert it into standardized JSON format data. The parsed data is locally encrypted and cached using the AES-256 encryption algorithm. When the network is interrupted, the data is temporarily stored locally, and automatic resume transmission is triggered after the network is restored to ensure no data loss. The encryption key is dynamically generated by the patient terminal and the hospital's integrated management platform through a key negotiation algorithm (ECDH) and is automatically updated every 24 hours.

[0038] Before data transmission, the LZ77 compression algorithm is used to compress the data before transmitting it to the hospital's comprehensive management platform. Data transmission status is monitored, and if transmission fails, it will automatically retry. The number of retries is ≤5, and the retry intervals are 1s, 2s, 4s, 8s, and 16s respectively.

[0039] The interactive interface module is designed based on the usage needs of special groups such as elderly patients. It adopts a simple and user-friendly interactive logic to visually display the status of the insulin pump terminal, graded warning information, and provide an interface for remote operation confirmation and assistance request initiation.

[0040] This invention uses a font size of ≥24 and a high-contrast color scheme, such as black background with white text or white background with black text, to display the status of the insulin pump terminal, including the remaining insulin amount, battery level, current infusion mode, and current basal rate; it supports voice broadcast function, which allows patients to trigger parameter broadcast through voice commands. The broadcast voice speed can be adjusted from 0.8 to 1.2 times to suit the hearing and visual needs of elderly patients.

[0041] In this invention, the graded early warning information is divided into four levels according to risk level: emergency, high, medium, and low. These are distinguished by blue, yellow, orange, and red icons and background colors, respectively, visually displaying the warning type, warning time, risk description, and operational instructions. When a patient experiences three consecutive infusion failures or a tubing pressure value ≥100kPa, it is classified as an emergency. This emergency is indicated by a red icon with a light red background border and 36-point font. It cannot be skipped manually; you must click "Processing" or "Resolved" to confirm.

[0042] When the remaining drug dosage is ≤10U, the battery power is ≤10%, or the pipeline pressure is ≥50kPa, it is judged as a high level, marked with an orange icon and a light orange background border, in 32-point font. It cannot be skipped manually; you must click "Read" to confirm.

[0043] When the remaining drug level is ≤30U or the battery level is ≤15%, it is judged as medium level, and a yellow icon with a light yellow background border and 28-point font is used. It can be manually closed, and if it is not closed within 10 seconds, it will automatically shrink to the top persistent notification.

[0044] When the remaining medication is ≤50U, the battery level is ≤25%, a notification indicates successful infusion mode switching, there is a data synchronization delay, or a consumable usage time warning appears, the alert is classified as a low-level alert. This alert is displayed with a blue icon and a light blue background border, in 24-point font. It can be manually closed; if not closed within 10 seconds, it automatically shrinks to a persistent notification at the top. Warning messages are arranged in reverse chronological order to ensure the latest warnings are displayed first.

[0045] Remote operation confirmation displays the initiator's identity, specific operation details, and execution risk warnings in a pop-up window, offering only two confirmation options: agree and refuse. The assistance request interface includes preset quick request buttons, such as replacing medicine boxes, venting tubing, parameter consultation, and emergency assistance; it also supports voice input of custom request content. After a request is initiated, its status is displayed in real time, including pending response, processing, and resolved.

[0046] The local early warning module is used to trigger emergency early warnings based on local cached data and preset rules when the network is interrupted or the platform's early warning is delayed, and provides multi-mode reminders. The multi-mode reminder unit supports the combined use of three reminder methods: vibration, ringtone, and voice broadcast. The reminder intensity is adjusted according to the early warning level. If the patient does not respond to the early warning within a preset time, the reminder intensity is increased until the patient responds.

[0047] The tiered authorization user terminals in this invention include family member terminals, caregiver terminals, medical staff terminals, and technician terminals. The family member terminals and caregiver terminals are used to receive warning information at three levels (intermediate, advanced, and emergency) associated with the patient, view core safety parameters associated with the patient, and initiate preset low-risk remote assistance requests. Core safety parameters include remaining insulin levels, battery level, current infusion mode, warning status, and key operation records. Low-risk remote assistance only initiates preset low-risk operation requests to pause infusion and remind the user to replace consumables.

[0048] The healthcare terminal is used to view the complete treatment data trends of the patients under its care, initiate therapeutic operation suggestions, and manage patient groups. The complete treatment data trends include the basal rate change trend, high-dose infusion distribution, remaining drug volume change curve, battery power change curve, blood glucose correlation trend graph, and historical early warning statistics.

[0049] Doctors can initiate professional therapeutic action suggestions based on the patient's complete data trends, including suggestions for adjusting basal rates, setting temporary basal rates, calculating high-dose infusions, and switching infusion modes. Patient group management allows doctors to view the status list of patients under their care in batches, including patient name, gender, age, current warning status, and the time of the most recent data synchronization. Patients can be filtered by warning level and patient type to quickly locate high-risk patients, and follow-up tasks can be assigned to nurses through the system's built-in messaging function.

[0050] The technician terminal is used to view diagnostic information of the patient's insulin pump terminal, including communication status records, firmware version information, fault code history, and sensor calibration records, to quickly locate the cause of device failure. After obtaining dual authorization from the patient and the hospital, the technician can initiate non-therapeutic device maintenance operations, including firmware updates, communication module resets, and sensor calibration parameter distribution.

[0051] This invention's hospital integrated management platform includes an access and communication gateway module, a data storage and management module, an intelligent early warning and analysis engine module, an authorization and security management module, and a data fusion and application module. The access and communication gateway module is used to authenticate and verify the identities and permissions of various terminals, such as insulin pump terminals, patient monitoring terminals, and tiered authorized user terminals. Each terminal must submit qualification documents, including device serial numbers and user identity information, upon initial access. After approval, a unique identity identifier and access token are assigned. The module handles high-concurrency user requests and distributes them to multiple application server nodes, each processing requests independently. Simultaneously, a rate-limiting mechanism based on a token bucket algorithm is employed to prevent malicious attacks from causing system paralysis. Furthermore, heterogeneous data from different terminals is converted into a unified data format conforming to the HL7 FHIR standard; and the data is cleaned, including removing duplicate data, correcting outliers, and supplementing missing values.

[0052] The data storage and management module is used to structure and store patients' static file data and dynamic treatment data. This invention uses a relational database to store patients' static file data, including basic patient information, medical history information, treatment plan, and authorization relationship information; it uses a distributed storage array to store dynamic treatment data, including insulin pump operation data, blood glucose data, early warning records, operation logs, and remote assistance records; and it adopts a dual backup strategy of real-time off-site backup and scheduled full backup for data backup and recovery; wherein, real-time off-site backup synchronizes data to an off-site disaster recovery center.

[0053] The intelligent early warning and analysis engine module performs multi-dimensional risk prediction and analysis based on a three-level early warning rule base, and generates tiered early warning information using time-series data analysis and risk prediction models; the three-level early warning rule base includes: The device threshold rules are based on the threshold settings of the operating parameters of the insulin pump terminal, including the remaining drug volume and the tubing pressure value.

[0054] Behavioral pattern rules identify abnormalities based on the patient's historical infusion behavior patterns and construct a normal infusion behavior model of the patient using the K-means clustering algorithm. When the real-time infusion data deviates from the normal model by more than a preset threshold ±25%, an early warning is triggered.

[0055] Physiological model rules, which integrate insulin pump terminal data, continuous glucose monitoring data, and patient behavior data, predict potential health risks through physiological metabolic models.

[0056] The process of generating tiered early warning information is as follows: The sliding window algorithm is used to analyze continuous time series data. By calculating statistical indicators such as the mean, variance, and rate of change of the data within the window, the trend changes of the data are identified.

[0057] A risk prediction model is built based on a long short-term memory neural network (LSTM). It integrates the operating data of the insulin pump terminal, continuous blood glucose monitoring data, patient behavior data, and patient physiological parameters to predict the probability of patients experiencing hypoglycemia or hyperglycemia.

[0058] The Long Short-Term Memory (LSTM) neural network model has an input dimension of 28, 3 hidden layers, 64 neurons in each layer, and outputs a risk probability ranging from 0 to 100%. An emergency warning is triggered when the probability of hypoglycemia is ≥70% or the probability of hyperglycemia is ≥80%.

[0059] A high-level warning is triggered when the probability of hypoglycemia is 50%-70% or the probability of hyperglycemia is 60%-80%.

[0060] A medium-level warning is triggered when the probability of hypoglycemia is 30%-50% or the probability of hyperglycemia is 40%-60%.

[0061] A low-level warning is triggered when the probability of hypoglycemia is less than 30% and the probability of hyperglycemia is less than 40%.

[0062] This invention selects the channel and frequency for pushing early warning information based on the warning level, patient attributes, authorization relationship, and terminal online status. For low-level alerts, they are only pushed to the patient monitoring terminal, using a weak reminder method with pop-up windows and low-volume ringtones.

[0063] For medium-level alerts, notifications are sent to both the patient monitoring terminal and the family member's terminal. The patient monitoring terminal uses a moderate-intensity alert, while the family member's terminal uses a pop-up window and a ringtone alert.

[0064] For high-level alerts, they are pushed to the patient, family members, and medical staff terminals. The patient monitoring terminal uses strong reminders such as pop-up windows, strong vibration, high-volume ringtones, and looped voice broadcasts, while the family members and medical staff terminals use pop-up windows, ringtones, and SMS reminders.

[0065] For emergency alerts, in addition to the high-level alert push channels, a voice call reminder will be added. If no feedback action is detected from the patient monitoring terminal within a preset time for high-level or higher alerts, the alert escalation process will be executed, which will expand the notification scope by adding other emergency contacts and increase the reminder intensity to continuous vibration and maximum volume ringtone.

[0066] The authorization and security management module is used to manage the access and operation permissions of each terminal based on the role-based access control matrix. It controls authorization through a multi-level review process and records the entire process audit log. The roles in the role-based access control matrix include patients, family members, nurses, doctors, technicians and administrators. It adopts the role-based access control (RBAC) model to refine the data access permissions and operation permissions of each role.

[0067] like Figure 3 As shown, authorization is controlled through a multi-level approval process, specifically as follows: S11. The patient initiates an authorization request through the patient monitoring terminal, selecting the authorized user, the scope of permissions, and the authorization validity period.

[0068] S12. The authorization application is submitted to the hospital's integrated management platform, where the patient's attending physician or nurse conducts a second review, which includes verifying the authenticity of the authorized user's identity and the reasonableness of the scope of permissions.

[0069] S13. After the review is approved, the hospital's integrated management platform will activate the corresponding permissions, generate an authorization certificate, and issue it to the authorized user's terminal; if the review is not approved, the reason for rejection will be provided, and the patient will have to resubmit the application.

[0070] Record full-process audit logs, including terminal login or logout, data access, early warning triggering or processing, remote operation requests, confirmation or execution, authorization applications, review or revocation, and equipment maintenance operations; the audit log content includes operation time, operation subject, operation type, operation content, operation result, device identifier and IP address, and is stored in encrypted form using blockchain technology.

[0071] The data fusion and application module is used to access multi-source health data, construct a comprehensive patient risk assessment model using the analytic hierarchy process (AHP), and analyze system operation quality indicators.

[0072] Specifically, data from various health devices is accessed via Bluetooth, Wi-Fi, or 4G / 5G networks and synchronized to the hospital's comprehensive management platform. The data from these health devices includes blood glucose levels, weight, exercise duration and intensity, heart rate, blood pressure, and sleep duration.

[0073] Simultaneously, the Analytic Hierarchy Process (AHP) was used to construct a patient health risk assessment model, comprehensively scoring patients from four dimensions: blood glucose control, treatment adherence, equipment operating status, and complication risk. If the overall score is ≤60, the patient is considered a high-risk patient.

[0074] If the overall score is 60-80, the patient is considered to be at medium risk.

[0075] If the overall score is ≥80, the patient is considered low-risk.

[0076] For high-risk patients, they are marked as key monitoring targets and recommended to the medical team for priority follow-up and intervention. At the same time, a personalized risk assessment report is generated, including risk level, risk point analysis, and intervention suggestions, for medical staff to refer to when developing treatment plans. Key quality indicators of system operation and care services are statistically analyzed, including equipment alarm rate, accuracy rate of early warning rule triggering, remote assistance usage rate, high-risk patient intervention success rate, and patient follow-up completion rate. The statistical results of the indicators are displayed using data visualization charts. Through quality indicator analysis, problems in system operation and weaknesses in care services are identified.

[0077] like Figure 4 As shown, the secure communication and control center includes an encrypted communication module, a command security control module, and an emergency backup module. The encrypted communication module ensures the security of all data transmission and storage within the system, preventing data theft, tampering, or leakage. Specifically, it uses the national standard SM4 encryption algorithm for end-to-end encryption of transmitted data, employs the ECDH algorithm for dynamic negotiation to generate encryption keys, and uses the asymmetric encryption algorithm RSA-2048 for key transmission protection. Furthermore, sensitive personal information of patients is anonymized, with the following rules: the last character of the name is hidden, the middle 8 digits of the ID number are hidden, and the middle 4 digits of the contact information are hidden.

[0078] The instruction security control module is used to perform digital signature verification, operation validity management, and secondary permission verification on remote operation instructions to ensure the compliance and security of operations. Digital signature verification includes the instruction content, the initiator's identity, and a timestamp. This invention uses the RSA-2048 algorithm to digitally sign all downlink instructions, with the signature information including the instruction content, the initiator's identity, and a timestamp. Simultaneously, this invention sets a lifecycle for all remote operation requests, i.e., an operation validity period. If patient confirmation is not obtained within the validity period, the request expires and is resubmitted. Furthermore, a 10-second feedback waiting time is set after instruction execution. If no successful execution feedback is received from the terminal within 10 seconds, the operation is considered a failure, and the platform automatically pushes an operation failure reminder to the initiator and the patient monitoring terminal, recording the reason for the failure.

[0079] Secondary permission verification refers to the hospital's integrated management platform verifying the initiator's permission status and the patient monitoring terminal's online status again before remote operation instructions are sent to the insulin pump terminal for execution. If the initiator's permissions have been revoked, the permission scope does not match, or the patient monitoring terminal is offline, the instruction will be immediately intercepted, execution will be refused, and the reason for the operation failure will be reported back to the initiator.

[0080] The emergency protection module is used to enable local monitoring and early warning mode when the network is interrupted, and to resume data transmission after the network is restored, as well as to monitor and intercept abnormal operations in real time.

[0081] When the network connection between the patient monitoring terminal and the hospital's integrated management platform is interrupted, the network interruption fault tolerance unit enables the patient monitoring terminal to maintain a long Bluetooth connection with the insulin pump terminal, continue to collect and locally encrypt and cache pump data; at the same time, the local early warning module of the patient monitoring terminal is automatically activated, triggering an emergency warning based on the locally cached data; when the network is restored, the patient monitoring terminal automatically initiates a breakpoint resume request to the platform, synchronizing the cached historical data to the platform to ensure that no data is lost and monitoring is uninterrupted.

[0082] The system monitors all operation requests within the system in real time, and uses a rule engine to identify and block abnormal operations, including operations that exceed the scope of permissions, operations with abnormal frequency, and operations suspected of being malicious attacks. After an abnormal operation is blocked, the system immediately pushes an alarm message to the hospital administrator's terminal and records the abnormal operation log.

[0083] The core data flow and workflow of the system of this invention are as follows, taking hypoglycemia risk warning and intervention in a remote monitoring scenario as an example: Step 1: Data Acquisition and Upload: The real-time data acquisition module of the insulin pump terminal continuously collects data such as basic infusion parameters and device status parameters, and periodically uploads the data to the patient monitoring terminal via the near-field communication module (Bluetooth); at the same time, the continuous glucose monitor worn by the patient synchronizes real-time blood glucose data to the patient monitoring terminal.

[0084] Step 2: The patient monitoring terminal parses the data and converts it into a standard format. After being encrypted and stored in a local encrypted cache subunit, the data is uploaded to the hospital's integrated management platform.

[0085] Step 3: The access and communication gateway module of the hospital's integrated management platform receives the data, and after format standardization and cleaning, it is transmitted to the intelligent early warning and analysis engine module. The sliding window algorithm is used to analyze the blood glucose change trend. Based on the LSTM model, the insulin pump data and blood glucose data are fused to predict that the probability of hypoglycemia within the next 60 minutes is ≥85%, triggering an emergency level warning.

[0086] Step 4: Based on the emergency warning level and the patient attributes of elderly patients living alone, the warning information is simultaneously pushed to the patient monitoring terminal, family member terminal, and medical staff terminal. The patient monitoring terminal triggers strong vibration, high-volume ringtone, and looped voice broadcast. The family member and medical staff terminals trigger pop-up windows, ringtones, and SMS reminders, and automatically dial the three parties for voice calls.

[0087] Step 5: If the patient does not respond to the alert due to being asleep, the family member's terminal receives the alert and initiates a remote operation request to "reduce the temporary basal rate to 0U / h". After the request is verified by the hospital's integrated management platform, it is pushed to the patient's monitoring terminal. The patient is awakened by a continuous ringing sound, views the operation details and risk warnings in the remote operation confirmation pop-up window, and clicks "Agree" to confirm the operation.

[0088] Step Six: The hospital's integrated management platform generates an instruction to adjust the temporary basal rate. After being digitally signed by the secure communication and control center, the instruction is sent to the patient monitoring terminal. After verifying the validity of the signature, the patient monitoring terminal sends the instruction to the insulin pump terminal via Bluetooth. After verifying the instruction, the insulin pump terminal executes the basal rate adjustment operation and sends the execution result back to the patient terminal.

[0089] Step 7: The patient monitoring terminal uploads the operation results to the hospital's integrated management platform, and the platform feeds back the results to the family terminal and the medical staff terminal; at the same time, the authorization and security management module records all operations of the entire process, forming an unalterable audit log.

[0090] Step 8: Based on the system feedback and the patient's real-time data, the doctor sends follow-up intervention suggestions (such as carbohydrate supplementation) to the patient and their family through the medical terminal; the nurse conducts telephone follow-ups with the patient according to the follow-up tasks assigned by the platform, confirms the patient's status, and updates the follow-up records.

[0091] Therefore, the present invention adopts the above-mentioned insulin pump safety reminder system based on remote communication, which effectively breaks down the data silos of equipment, realizes intelligent early warning and safety collaboration across time and space, and provides a systematic solution for building a continuous, proactive and safe integrated insulin pump management model inside and outside the hospital.

[0092] 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 them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A remote communication-based insulin pump safety reminder system, characterized in that, This includes insulin pump terminals, patient monitoring terminals, tiered authorization user terminals, hospital integrated management platforms, and secure communication and control centers; The insulin pump terminal is used to collect device operating parameters and upload them via secure near-field communication, and to receive and execute verified operating instructions. The patient monitoring terminal communicates with the insulin pump terminal to bridge the data transfer between the insulin pump terminal and the hospital's integrated management platform, provide local data encryption and caching, display tiered early warnings, and enable remote human-computer interaction. Tiered authorization user terminals are used to receive alerts related to insulin pump terminals, view patient status, and initiate remote operation requests. The hospital's integrated management platform communicates with patient monitoring terminals and hierarchically authorized user terminals to access and manage various terminals, aggregate and intelligently analyze data, generate early warning information based on multi-dimensional rules, manage and audit permissions, and integrate multi-source data applications. Secure communication and control is used to ensure secure data transmission, compliant operation instructions, and fault-tolerant emergency operation of the system.

2. The insulin pump safety reminder system based on remote communication according to claim 1, characterized in that, Insulin pump terminals include: The real-time data acquisition module is used to collect basic infusion parameters, equipment status parameters, abnormal signals, and identity and time synchronization. Among them, the basic infusion parameters include the current basic infusion rate, the temporary basic rate setting value and effective time, the high-dose infusion volume, the high-dose infusion time and frequency, and the current infusion mode. Device status parameters include the remaining insulin in the reservoir, battery level, and infusion line pressure; abnormal signals include alarm codes, fault types, and device fault status flags; identity and time synchronization includes recording the device's unique identifier (SN), patient-associated ID, and timestamp. The near-field communication module is used to establish a secure pairing connection with the patient monitoring terminal and upload data in a periodic and event-driven manner; the secure pairing connection adopts a two-way authentication mechanism of the device's unique identifier SN code and the patient monitoring terminal's MAC address; The local execution module is used to perform safety checks on received remote operation commands, including pausing infusion, adjusting the base rate, and venting the pipeline. It also uses the RSA-2048 algorithm to verify the digital signature of the operation command. Once the verification is successful, the corresponding operation is executed.

3. The insulin pump safety reminder system based on remote communication according to claim 2, characterized in that, The patient monitoring terminal includes: The data bridging module is used to receive and parse data from the insulin pump terminal. It uses the AES-256 encryption algorithm to locally encrypt and cache the parsed data, and uses the LZ77 compression algorithm to compress the data before transmitting it to the hospital's integrated management platform. The interactive interface module is used to visually display the status of the insulin pump terminal, graded warning information, and provide a remote operation confirmation and assistance request initiation interface. The remote operation confirmation is displayed in the form of a pop-up window, showing the identity of the operation initiator, the specific operation content, and the execution risk warning, and only provides two confirmation options: agree and refuse. The assistance request initiation interface has preset quick request buttons, including changing the pillbox, purging the tubing, parameter consultation, and emergency help. The local early warning module is used to trigger emergency early warnings based on local cached data and preset rules when the network is interrupted or the platform's early warning is delayed, and provides multi-mode reminders.

4. The insulin pump safety reminder system based on remote communication according to claim 3, characterized in that, The status of the insulin pump terminal includes the remaining insulin level, battery level, current infusion mode, and current basal rate; The tiered early warning information is divided into four levels according to risk level: emergency, high, medium, and low. When a patient experiences three consecutive infusion failures or a tubing pressure value ≥100 kPa, the situation is classified as an emergency. When the remaining drug dosage is ≤10U, the battery charge is ≤10%, or the pipeline pressure is ≥50kPa, it is judged as a high level; When the remaining drug dosage is ≤30U or the battery charge is ≤15%, it is judged as medium level; When the remaining medication is ≤50U, the battery level is ≤25%, the infusion mode is successfully switched, there is a data synchronization delay, or a consumable usage time warning appears, it is judged as a low level.

5. The insulin pump safety reminder system based on remote communication according to claim 4, characterized in that, Tiered authorization user terminals include: Family member terminals and caregiver terminals are used to receive alerts from associated patients, view core safety parameters associated with patients, and initiate preset low-risk remote assistance requests. Core safety parameters include remaining insulin levels, battery level, current infusion mode, alert status, and key operation records. Low-risk remote assistance requests include pausing infusion and reminding to replace consumables. The healthcare terminal is used to view the complete treatment data trends of the patients under its care, initiate therapeutic operation suggestions, and manage patient groups. The complete treatment data trends include the basal rate change trend, high-dose infusion distribution, remaining drug volume change curve, battery power change curve, blood glucose correlation trend graph, and historical warning statistics. Therapeutic procedures include recommendations for adjusting basal rates, setting temporary basal rates, calculating high-dose infusions, and switching infusion modes; patient population management includes patient name, gender, age, current alert status, and the time of the most recent data synchronization. The technician terminal is used to view diagnostic information of the patient's insulin pump terminal, including communication status records, firmware version information, fault code history, and sensor calibration records. After authorization, it can initiate non-therapeutic device maintenance operations, including firmware updates, communication module resets, and sensor calibration parameter distribution.

6. The insulin pump safety reminder system based on remote communication according to claim 1, characterized in that, The hospital's integrated management platform includes: The access and communication gateway module is used to authenticate and verify the identity of various connected terminals, handle high-concurrency requests, and standardize data. Each terminal must submit qualification documents, including the device serial number and user identity information, when it connects for the first time. After the review is approved, a unique identity identifier and access token are assigned. The data storage and management module is used to structure and store patients' static record data and dynamic treatment data. It employs a dual backup strategy of real-time off-site backup and scheduled full backup for data backup and recovery. Static record data includes basic patient information, medical history information, treatment plan and authorization relationship information; dynamic treatment data includes insulin pump operation data, blood glucose data, early warning records, operation logs and remote assistance records. The intelligent early warning and analysis engine module performs multi-dimensional risk prediction and analysis based on a three-level early warning rule base, and generates graded early warning information using time-series data analysis and risk prediction models. The authorization and security management module is used to manage the access and operation permissions of each terminal based on the role permission matrix, control authorization through a multi-level review process, and record the entire process audit log; The data fusion and application module is used to access multi-source health data, construct a comprehensive patient risk assessment model using the analytic hierarchy process (AHP), and analyze system operation quality indicators.

7. The insulin pump safety reminder system based on remote communication according to claim 6, characterized in that, The Level 3 early warning rule base includes: Equipment threshold rules are based on the threshold settings of the operating parameters of the insulin pump terminal, including the remaining drug volume and tubing pressure. Behavioral pattern rules identify abnormalities based on the patient's historical infusion behavior patterns, and use the K-means clustering algorithm to construct a normal infusion behavior model for the patient. When the real-time infusion data deviates from the normal model by more than a preset threshold ±25%, an early warning is triggered. Physiological model rules, which integrate insulin pump terminal data, continuous glucose monitoring data and patient behavior data, predict potential health risks through physiological metabolic models; The process of generating tiered early warning information is as follows: The sliding window algorithm is used to analyze continuous time series data. By calculating the mean, variance, and rate of change of the data within the window, the trend of data changes can be identified. A risk prediction model is built based on a Long Short-Term Memory (LSTM) neural network, which integrates operational data from insulin pump terminals, continuous glucose monitoring data, patient behavior data, and patient physiological parameters to predict the probability of patients experiencing hypoglycemia or hyperglycemia. An emergency warning is triggered when the probability of hypoglycemia is ≥70% or the probability of hyperglycemia is ≥80%. A high-level warning is triggered when the probability of hypoglycemia is 50%-70% or the probability of hyperglycemia is 60%-80%. A medium-level warning is triggered when the probability of hypoglycemia is 30%-50% or the probability of hyperglycemia is 40%-60%. A low-level warning is triggered when the probability of hypoglycemia is less than 30% and the probability of hyperglycemia is less than 40%.

8. The insulin pump safety reminder system based on remote communication according to claim 6, characterized in that, The roles in the role-based access control matrix include patients, family members, nurses, doctors, technicians, and administrators. It adopts a role-based access control (RBAC) model to refine the data access permissions and operation permissions of each role. Authorization is controlled through a multi-level approval process, specifically as follows: S11. The patient initiates an authorization request through the patient monitoring terminal, selecting the authorized user, the scope of permissions, and the authorization validity period; S12. The authorization application is submitted to the hospital's comprehensive management platform, where the patient's attending physician or nurse conducts a second review. The review includes verifying the authenticity of the authorized user's identity and the reasonableness of the scope of authority. S13. After the review is approved, the hospital's integrated management platform activates the corresponding permissions, generates an authorization certificate, and issues it to the authorized user's terminal; if the review fails, the reason for rejection is provided, and the patient must resubmit the application. Record full-process audit logs, including terminal login or logout, data access, early warning triggering or processing, remote operation requests, confirmation or execution, authorization applications, review or revocation, and equipment maintenance operations; the audit log content includes operation time, operation subject, operation type, operation content, operation result, device identifier and IP address, and is stored in encrypted form using blockchain technology.

9. The insulin pump safety reminder system based on remote communication according to claim 1, characterized in that, A patient health risk assessment model was constructed using the analytic hierarchy process (AHP) to comprehensively score patients across four dimensions: glycemic control, treatment adherence, equipment operational status, and complication risk. If the overall score is ≤60, the patient is considered a high-risk patient; If the overall score is 60-80, the patient is classified as a medium-risk patient; If the overall score is ≥80, the patient is considered low-risk. If a patient is at high risk, they will be marked as a key monitoring subject and referred to the medical team for priority follow-up and intervention. The system also statistically analyzed key quality indicators of system operation and care services, including equipment alarm rate, accuracy rate of early warning rule triggering, remote assistance usage rate, success rate of intervention for high-risk patients, and patient follow-up completion rate.

10. The insulin pump safety reminder system based on remote communication according to claim 1, characterized in that, The secure communication and control center includes an encrypted communication module, a command security control module, and an emergency support module; The encrypted communication module uses the national standard SM4 encryption algorithm to encrypt the transmitted data end-to-end, and uses the ECDH algorithm with dynamic negotiation mechanism to generate encryption keys. The key transmission process is protected by the asymmetric encryption algorithm RSA-2048. Sensitive personal information of patients is also desensitized. The specific desensitization rules are as follows: the last character of the name is hidden, the middle 8 digits of the ID number are hidden, and the middle 4 digits of the contact information are hidden. The instruction security control module is used to perform digital signature verification, operation timeliness management, and secondary permission verification on remote operation instructions. Digital signature verification includes the instruction content, the initiator's identity, and a timestamp; The emergency protection module is used to enable local monitoring and early warning mode when the network is interrupted, and to resume data transmission after the network is restored, as well as to monitor and intercept abnormal operations in real time.

Images