Medical emergency early warning method based on digital twinborn technology

Abstract

The invention relates to the technical field of intelligent medical equipment monitoring, and discloses a medical emergency early warning method based on a digital twinning technology, which comprises the following steps of: 1, constructing a digital twinning model of a plurality of intelligent medical equipment, and forming a virtual image by the digital twinning model according to hardware parameters and use environment data of the equipment; step 2, acquiring operation data of a plurality of intelligent medical devices in real time by using an edge computing device; and step 3, the edge computing device preprocesses the collected operation data. The technical scheme that a multi-device digital twinborn model is constructed, the cooperative working state is fused and analyzed, and the linkage fault extension range is judged through virtual prediction is adopted, the technical effect of accurately recognizing and early warning linkage faults among multiple devices is achieved, and compared with a monitoring scheme relying on single device state feedback in the prior art, the monitoring scheme has the advantage that the monitoring efficiency is improved. The defects of lack of multi-device comprehensive monitoring capability and incapability of early warning of linkage faults are overcome.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical device monitoring technology, specifically a medical emergency early warning method based on digital twin technology. Background Technology

[0002] Due to advancements in medical technology, intelligent medical devices such as pacemakers and insulin pumps are playing an increasingly important role in the daily monitoring and treatment of patients. The safety and stability of these devices are directly related to the patient's life safety.

[0003] However, existing security monitoring systems have revealed the following shortcomings in application:

[0004] Existing monitoring systems mostly rely on the status feedback of single devices and lack the comprehensive monitoring capability for the coordinated operation of multiple devices, making it difficult to provide timely early warnings when linkage failures occur between devices.

[0005] Existing monitoring systems rely on central servers for data analysis. The transmission of massive amounts of data over long distances can cause network delays, resulting in slow early warning response times and making it difficult to meet the real-time requirements of medical emergencies.

[0006] Existing technologies aggregate all raw operational data to a central server for processing, which increases the network bandwidth burden on data transmission and puts enormous computational pressure on the central server, affecting the system's operating efficiency.

[0007] To address these issues, this invention proposes a medical emergency early warning method based on digital twin technology. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a medical emergency early warning method based on digital twin technology to solve the problems mentioned in the background section.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a medical emergency early warning method based on digital twin technology, comprising:

[0010] Step 1: Construct digital twin models of multiple smart medical devices. The digital twin models form virtual images based on the hardware parameters of the devices and the usage environment data.

[0011] Step 2: Use edge computing devices to collect real-time operational data from multiple smart medical devices;

[0012] Step 3: The edge computing device preprocesses the collected operational data and calculates the collaborative working status between devices by fusing and analyzing the operational data.

[0013] Step 4: The edge computing device monitors the risk of linkage failure based on the collaborative working status. When a linkage failure risk is detected, a digital twin model is used to make a virtual prediction to determine the scope of the extended impact of the linkage failure risk.

[0014] Step 5: Based on the expanded impact range, confirm the risk of linkage failure. The edge computing device sends an immediate warning signal and automatically adjusts the device parameters of the smart medical device through the emergency response mechanism.

[0015] Preferably, calculating the collaborative working status between devices includes: establishing a collaborative working status model between the multiple intelligent medical devices, and calculating quantitative indicators of the collaborative working status in real time based on the operating data;

[0016] The virtual prediction using the digital twin model includes: combining the preset interdependencies between the smart medical devices to virtually simulate the risk of the linked failure, so as to determine the scope of the extended impact.

[0017] Preferably, the virtual prediction using the digital twin model includes: combining the preset interdependencies between the intelligent medical devices to virtually simulate the risk of the linked failure, so as to determine the extended impact range.

[0018] Preferably, the monitoring of linkage failure risks includes: monitoring operational or status abnormalities in the collaborative working state, and calculating the possibility that the abnormality will affect other intelligent medical devices.

[0019] Preferably, the automatic adjustment of the device parameters of the intelligent medical device is used to reduce the risk of the linkage failure spreading among multiple intelligent medical devices;

[0020] Sending the real-time warning signal includes: sending the real-time warning signal to the terminal device of medical personnel;

[0021] The multiple intelligent medical devices include pacemakers and insulin pumps.

[0022] Preferably, step one further includes:

[0023] Sub-step A device mechanism model is constructed based on the hardware parameters. The device mechanism model This is used to describe the basic physical characteristics and operating logic of the intelligent medical device;

[0024] Sub-step Based on the aforementioned usage environment data Building a data-driven bias correction model The deviation correction model Used to compensate the device mechanism model Deviations in actual operation;

[0025] Sub-step The device mechanism model is integrated. With the aforementioned deviation correction model To construct the digital twin model, the digital twin model outputs a virtual state. That is, the virtual image, the virtual state The calculation formula is:

[0026] ,

[0027] in, For the current time, It is in a virtual state. For the device mechanism model The running function, For control input signals of intelligent medical devices. For hardware parameters, For the deviation correction model The correction function, This refers to the usage environment data at the current moment;

[0028] Sub-step Set the model fidelity threshold And using the actual operating data of the intelligent medical device Calculate the virtual state Real-time fidelity The real-time fidelity The calculation formula is:

[0029] ,

[0030] in, For intelligent medical devices Actual operating data at any given time For real-time fidelity;

[0031] Sub-step Perform model update determination; if the real-time fidelity... Less than the model fidelity threshold Trigger the deviation correction model Online updates, utilizing the latest usage environment data. Retrain the bias correction model This is to ensure the accuracy of the virtual image.

[0032] Preferably, step two further includes:

[0033] Sub-step Determine the device set of the multiple intelligent medical devices. The set of devices Include A set of devices, and the set of devices. The first in Taiwanese equipment Configure data sampling period ;

[0034] Sub-step The edge computing device according to the data sampling period minimum value Determine the system sampling time The edge computing device at the system sampling time Poll the set of devices , obtain the Taiwanese equipment The original dataset ;

[0035] Sub-step The original dataset Based on the device status The sensor feedback information With the environmental change information constitute;

[0036] Sub-step For the original dataset The sensor feedback information in Perform data validity determination The data validity determination The conditions are:

[0037] ≤ ≤ ,

[0038] , ,

[0039] in, For the first Taiwanese equipment The lower limit of the sensor's measurement range. For the first Taiwanese equipment The upper limit of the sensor's measurement range, For the data to be valid, The data is invalid.

[0040] Sub-step For the original dataset Perform data integrity determination The data integrity determination The conditions are:

[0041] ,

[0042] in, The data field is missing. For data integrity, Data is missing;

[0043] Sub-step When the data validity is determined The result is And the data integrity determination The result is At that time, the edge computing device is the original dataset. Additional synchronization timestamp And will be appended with the synchronization timestamp. The original dataset Store in edge local data cache ;

[0044] Sub-step The edge local data cache area The data set stored in the middle constitutes the running data, which is used for the fusion analysis in step three.

[0045] Preferably, step three further includes:

[0046] Sub-step The edge local data cache Extract the running data, the running data constitutes The intelligent medical device described above was at the system sampling time Synchronized dataset ;

[0047] Sub-step For the synchronized dataset The sensor feedback information in Perform data normalization preprocessing to generate normalized data. The calculation formula for the data normalization preprocessing is as follows:

[0048] ,

[0049] in, For the first The equipment is in Normalized data at any given time For the first The minimum preset range of the sensor on the device. For the first The preset maximum range of the sensor on the device;

[0050] Sub-step Construct a collaborative working state model among the multiple intelligent medical devices. The collaborative working state model At the system sampling time state vector The normalized data from all devices The state vector constitutes The expression is:

[0051] ,

[0052] in, For state vectors, This represents the total number of intelligent medical devices. Transpose of a vector;

[0053] Sub-step Define the baseline state vector of the collaborative working state model. The reference state vector This represents the ideal state when the aforementioned multiple intelligent medical devices work together;

[0054] Sub-step Calculate the state vector With the reference state vector Cooperative deviation between The cooperative deviation This refers to the quantitative indicator of the collaborative working state, namely the collaborative deviation. The calculation formula is:

[0055] ,

[0056] in, For coordination bias, For the first Preset collaboration weights for the devices Reference state vector The first in Quantity;

[0057] Sub-step Set a cooperative deviation threshold The edge computing device is based on the cooperative deviation. With the aforementioned collaborative deviation threshold The comparison determines the current collaborative working state. ;

[0058] Sub-step When the collaborative working state The determination criteria are:

[0059] ,

[0060] Among them, the collaborative working state This provides input for the monitoring and linkage failure risk in step four.

[0061] Preferably, step four further includes:

[0062] Sub-step Receive the collaborative working status determined in step three. And set trigger conditions for linkage fault risk monitoring. ;

[0063] Sub-step The triggering conditions for the linkage fault risk monitoring For: when the collaborative working state When a collaborative anomaly occurs, the edge computing device initiates the linkage fault risk monitoring.

[0064] Sub-step When the linkage fault risk monitoring trigger condition When the condition is met, the edge computing device calculates the collaborative deviation. The device contribution of each of the aforementioned intelligent medical devices The contribution of the device The calculation formula is:

[0065] ,

[0066] in, For the first Preset collaboration weights for the devices For the first Normalized data from the equipment. Reference state vector The first in Quantity;

[0067] Sub-step Determine the contribution of the device. Largest equipment For the source of abnormal equipment The source of the abnormal device index The formula for determining it is:

[0068] ,

[0069] Among them, the source abnormal device This constitutes the aforementioned risk of linkage failure;

[0070] Sub-step The system loads and constructs digital twin models of the multiple smart medical devices, and obtains a preset interdependency matrix among the smart medical devices. The interdependence matrix elements Indicates equipment For equipment The influence coefficient;

[0071] Sub-step Define the source abnormal device abnormal state The abnormal state The source of the abnormal device The normalized data express;

[0072] Sub-step The edge computing device performs the virtual prediction within the digital twin model, utilizing the abnormal state. With the aforementioned interdependence matrix Calculate the risk of the linked failure to other devices. Fault propagation ;

[0073] Sub-step The probability of fault propagation The calculation formula is:

[0074] ,

[0075] in, For the preset fault propagation function, For the abnormal equipment at the source For equipment The influence coefficient mentioned above, To predict the time step;

[0076] Sub-step Set a fault propagation threshold The edge computing device will propagate all the fault probabilities. Greater than the fault propagation threshold equipment The equipment was identified as affected.

[0077] Sub-step The collection of all the affected devices That is, the extended impact range constituting the aforementioned linked failure risk, the extended impact range The determination criteria are:

[0078] ,

[0079] The extended scope of influence The confirmed linkage failure risk is used in step five.

[0080] Preferably, step five further includes:

[0081] Sub-step Receive the expanded influence range determined in step four. The extended range of influence Includes all the affected devices;

[0082] Sub-step The edge computing device is based on the extended influence range. Calculate the comprehensive risk level of the linked failure risk. The aforementioned comprehensive risk level The calculation formula is:

[0083] ,

[0084] in, To expand the scope of influence The Middle Preset criticality weights for affected equipment in Taiwan. For the calculation of the first The probability of fault propagation in the affected equipment described above;

[0085] Sub-step Set risk confirmation thresholds The edge computing device performs the determination of the risk of linkage failure. The determination The conditions are:

[0086] when The risk of linkage failure has been confirmed;

[0087] when ≤ The risk of linkage failure has not been confirmed.

[0088] Sub-step When the determination When the result is confirmed, the edge computing device generates the real-time warning signal. The real-time warning signal Including the aforementioned comprehensive risk level With the aforementioned extended range of influence The edge computing device will transmit the real-time warning signal. Sending data to the terminal devices of medical personnel;

[0089] Sub-step When the determination When the result is confirmed, the edge computing device activates the emergency response mechanism, which is based on the comprehensive risk level. With the aforementioned extended range of influence Load the corresponding emergency adjustment strategy ;

[0090] Sub-step The emergency adjustment strategy For the extended range of influence Each of the affected devices mentioned above Calculate new equipment parameters The new device parameters The calculation formula is: ,

[0091] in, For affected equipment Current device parameters The parameter adjustment function is a preset parameter adjustment function. Based on the aforementioned comprehensive risk level With the aforementioned key weights Calculate the parameter adjustment amount;

[0092] Sub-step The edge computing device extends its influence to the surrounding area. The affected devices in Issue parameter adjustment instructions to apply the new device parameters. This completes the automatic adjustment of the device parameters of the intelligent medical device.

[0093] This invention provides a medical emergency early warning method based on digital twin technology. It has the following beneficial effects:

[0094] 1. This invention adopts a technical solution of constructing a digital twin model of multiple devices and integrating and analyzing their collaborative working status, and using virtual prediction to determine the expansion range of linkage faults. This achieves the technical effect of accurately identifying and warning of linkage faults between multiple devices. Compared with the existing monitoring schemes that rely on the status feedback of a single device, this invention solves the shortcomings of lacking comprehensive monitoring capabilities for multiple devices and being unable to warn of linkage faults.

[0095] 2. This invention adopts a technical solution that deploys data analysis, linkage fault identification, and early warning decision-making on edge computing devices, achieving the technical effect of rapid local response and instant early warning signal transmission on the device. Compared with the existing technology that relies on a central server for data analysis, this invention solves the shortcomings of slow early warning response speed and inability to meet the real-time requirements of medical emergencies caused by network delays in data transmission.

[0096] 3. This invention adopts a technical solution of performing local preprocessing and fusion analysis of raw operating data at the edge computing device, achieving the technical effect of completing the calculation at the data source and reducing the amount of network data transmission. Compared with the existing technology that gathers all raw operating data to the central server for processing, this invention solves the shortcomings of heavy network bandwidth burden and huge computing pressure on the central server, and improves the operating efficiency of the system. Attached Figure Description

[0097] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0098] To enable those skilled in the art to understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0099] The present invention will now be described in detail with reference to the accompanying drawings:

[0100] Example:

[0101] Please see the appendix Figure 1 This invention provides a medical emergency early warning method based on digital twin technology, comprising:

[0102] Step 1: Construct digital twin models of multiple smart medical devices. The digital twin models form virtual images based on the hardware parameters of the devices and the usage environment data.

[0103] Step one further includes:

[0104] Sub-step Build a device mechanism model based on hardware parameters Equipment mechanism model Used to describe the basic physical characteristics and operating logic of intelligent medical devices;

[0105] Sub-step Based on usage environment data Building a data-driven bias correction model Deviation correction model Model for compensating equipment mechanism Deviations in actual operation;

[0106] Sub-step Fusion equipment mechanism model With deviation correction model To construct a digital twin model, the virtual state output by the digital twin model. This refers to a virtual image or virtual state. The calculation formula is:

[0107] ,

[0108] in, For the current time, It is in a virtual state. For the device mechanism model The running function, For control input signals of intelligent medical devices. For hardware parameters, For the deviation correction model The correction function, This refers to the usage environment data at the current moment;

[0109] Sub-step Set the model fidelity threshold Furthermore, utilizing the actual operational data of intelligent medical devices Calculate virtual state Real-time fidelity Real-time fidelity The calculation formula is:

[0110] ,

[0111] in, For intelligent medical devices Actual operating data at any given time For real-time fidelity;

[0112] Sub-step Perform model update determination; if real-time fidelity... Less than the model fidelity threshold Triggering the bias correction model Online updates, utilizing the latest usage environment data Retrain the bias correction model To ensure the accuracy of the virtual image;

[0113] Step 2: Use edge computing devices to collect real-time operational data from multiple smart medical devices. The operational data includes device status, sensor feedback information, and environmental change information of the smart medical devices' environment. The edge computing devices are smart gateways or local processing devices.

[0114] Step two further includes:

[0115] Sub-step Determine the equipment set of multiple intelligent medical devices. Equipment collection Include A set of devices, and a collection of devices. The first in Taiwanese equipment Configure data sampling period ;

[0116] Sub-step Edge computing devices are based on data sampling periods minimum value Determine the system sampling time Edge computing devices at system sampling time Polling device set , obtain the Taiwanese equipment The original dataset ;

[0117] Sub-step Original dataset Based on device status Sensor feedback information Information on environmental change constitute;

[0118] Sub-step For the original dataset Sensor feedback information Perform data validity determination Data validity determination The conditions are:

[0119] ≤ ≤ ,

[0120] , ,

[0121] in, For the first Taiwanese equipment The lower limit of the sensor's measurement range. For the first Taiwanese equipment The upper limit of the sensor's measurement range, For the data to be valid, The data is invalid.

[0122] Sub-step For the original dataset Perform data integrity determination Data integrity determination The conditions are:

[0123] ,

[0124] in, The data field is missing. For data integrity, Data is missing;

[0125] Sub-step When data validity is determined The result is And data integrity determination The result is At that time, the edge computing device is the raw dataset Additional synchronization timestamp And will be accompanied by a synchronization timestamp. The original dataset Store in edge local data cache ;

[0126] Sub-step Edge local data cache The data set stored in the middle constitutes the running data, which is used for the fusion analysis in step three;

[0127] Step 3: The edge computing device preprocesses the collected operational data and calculates the collaborative working status between devices by fusing and analyzing the operational data.

[0128] Step three further includes:

[0129] Sub-step From edge local data cache Extract runtime data from the runtime data; runtime data composition Taiwan intelligent medical device at system sampling time Synchronized dataset ;

[0130] Sub-step For synchronized datasets Sensor feedback information Perform data normalization preprocessing to generate normalized data. The calculation formula for data normalization preprocessing is:

[0131] ,

[0132] in, For the first The equipment is in Normalized data at any given time For the first The minimum preset range of the sensor on the device. For the first The preset maximum range of the sensor on the device;

[0133] Sub-step Construct a collaborative working state model among multiple intelligent medical devices. Collaborative work status model At the system sampling time state vector Normalized data from all devices Composition, state vector The expression is:

[0134] ,

[0135] in, For state vectors, This represents the total number of intelligent medical devices. Transpose of a vector;

[0136] Sub-step Define the baseline state vector of the collaborative working state model. Reference state vector This represents the ideal state when multiple intelligent medical devices work together;

[0137] Sub-step Calculate the state vector With reference state vector Cooperative deviation between Cooperative deviation This refers to the quantitative indicator of collaborative work status, specifically the collaborative deviation. The calculation formula is:

[0138] ,

[0139] in, For coordination bias, For the first Preset collaboration weights for the devices Reference state vector The first in Quantity;

[0140] Sub-step Set a cooperative deviation threshold Edge computing devices based on collaborative deviation With the cooperative deviation threshold The comparison determines the current collaborative work status. ;

[0141] Sub-step When in collaborative working state The determination criteria are:

[0142] ,

[0143] Among them, collaborative work status This provides input for monitoring and linkage failure risks in step four;

[0144] Step 4: The edge computing device monitors the risk of linkage failure based on the collaborative working status. When a linkage failure risk is detected, a digital twin model is used to make a virtual prediction to determine the scope of the extended impact of the linkage failure risk.

[0145] Step four further includes:

[0146] Sub-step Receive the collaborative work status determined in step three. And set trigger conditions for linkage fault risk monitoring. ;

[0147] Sub-step Triggering conditions for linkage fault risk monitoring For: When in collaborative working state In case of collaborative anomalies, edge computing devices initiate linkage fault risk monitoring;

[0148] Sub-step When the linkage fault risk monitoring trigger condition When the conditions are met, the edge computing device calculates the collaborative deviation. The contribution of each intelligent medical device in China Equipment contribution The calculation formula is:

[0149] ,

[0150] in, For the first Preset collaboration weights for the devices For the first Normalized data from the equipment. Reference state vector The first in Quantity;

[0151] Sub-step Determine the contribution of equipment Largest equipment For the source of abnormal equipment Abnormal equipment at the source index The formula for determining it is:

[0152] ,

[0153] Among them, abnormal equipment at the source This constitutes a risk of interconnected failures;

[0154] Sub-step Load the constructed digital twin models of multiple smart medical devices and obtain the preset interdependency matrix between the smart medical devices. Interdependence matrix elements Indicates equipment For equipment The influence coefficient;

[0155] Sub-step Define the source of the abnormal device abnormal state Abnormal state From the source of abnormal equipment Normalized data express;

[0156] Sub-step Edge computing devices perform virtual predictions within digital twin models, utilizing abnormal states. With the interdependency matrix Calculate the risk of linked failures to other equipment Fault propagation ;

[0157] Sub-step Fault propagation probability The calculation formula is:

[0158] ,

[0159] in, For the preset fault propagation function, For the abnormal equipment at the source For equipment Influence coefficient, To predict the time step;

[0160] Sub-step Set a fault propagation threshold Edge computing devices propagate all fault probabilities Greater than the fault propagation threshold equipment The equipment was identified as affected.

[0161] Sub-step The collection of all affected devices This constitutes the extended impact range of the risk of linked failures. The determination criteria are:

[0162] ,

[0163] Expanding the scope of influence Used to confirm the risk of linkage failure in step five;

[0164] Step 5: Based on the expanded impact range, confirm the risk of linkage failure, and the edge computing device sends an immediate warning signal. The edge computing device automatically adjusts the equipment parameters of the smart medical device through the emergency response mechanism.

[0165] Step five further includes:

[0166] Sub-step The extended scope of influence determined in step four. Expand the scope of influence Includes all affected devices;

[0167] Sub-step Edge computing devices expand their influence range. Calculate the comprehensive risk level of the risk of linkage failure. Overall risk level The calculation formula is:

[0168] ,

[0169] in, To expand the scope of influence The Middle Preset criticality weights for affected equipment in Taiwan. For the calculation of the first The probability of fault propagation in affected equipment;

[0170] Sub-step Set risk confirmation thresholds Determining the risk of failure in the execution of confirmation linkage by edge computing devices ,determination The conditions are:

[0171] when The risk of linkage failure has been confirmed;

[0172] when ≤ The risk of linkage failure has not been confirmed.

[0173] Sub-step When judged When the result is confirmed, the edge computing device generates an immediate warning signal. Real-time early warning signal Includes comprehensive risk level and expanding the scope of influence Edge computing devices will provide real-time warning signals. Sending data to the terminal devices of medical personnel;

[0174] Sub-step When judged When the result is confirmed, the edge computing device activates the emergency response mechanism, which is based on the comprehensive risk level. and expanding the scope of influence Load the corresponding emergency adjustment strategy ;

[0175] Sub-step Emergency adjustment strategy To expand the scope of influence Each of the affected devices Calculate new equipment parameters New equipment parameters The calculation formula is: ,

[0176] in, For affected equipment Current device parameters This is a preset parameter adjustment function. Based on comprehensive risk level With key weight Calculate the parameter adjustment amount;

[0177] Sub-step Edge computing devices are expanding their influence. Affected devices Issue parameter adjustment commands to apply the new device parameters. It automatically adjusts the device parameters of intelligent medical equipment.

[0178] By constructing a digital twin model that integrates equipment mechanisms and data-driven corrections, and introducing real-time fidelity verification and online update mechanisms, a virtual image that is highly synchronized with the physical equipment and can dynamically adapt can be created. This provides a simulation and deduction platform for subsequent fault prediction, ensuring the accuracy of virtual prediction results, which is the fundamental basis for achieving accurate early warning.

[0179] By performing unified data collection, validity and integrity checks on edge computing devices, and adding synchronized timestamps, it is ensured that all data entering the analysis phase is of high quality. The reliability of the input is guaranteed at the data source, avoiding misjudgments due to data quality issues, providing a solid data foundation for subsequent multi-device collaborative status analysis, and local caching creates conditions for low-latency processing.

[0180] By normalizing and fusing data from multiple devices on edge computing devices, quantitative indicators of collaborative working status are calculated. This transforms the monitoring of individual device status into a holistic assessment of the collaborative relationship between device systems. Furthermore, by performing calculations at the edge, the data transmission bandwidth and computational pressure on the central server are greatly reduced, thus solving the efficiency bottleneck of traditional architectures.

[0181] When a collaborative anomaly is detected, the source of the anomaly is located by calculating the contribution of each device. A digital twin model and a pre-defined interdependency matrix are used for virtual prediction to estimate the extent of the fault's spread. This achieves an intelligent upgrade from anomaly detection to consequence prediction, enabling the identification of the root cause of the problem and the proactive prediction of the fault's propagation path and impact range. It elevates early warning from simple status alerts to precise risk assessment.

[0182] By comprehensively quantifying the risk level of the predicted expanded impact range, a tiered early warning and automated emergency response are triggered based on this assessment, constructing a complete process from risk confirmation to closed-loop management. This enables rapid and intelligent emergency intervention. On the one hand, it provides medical personnel with key decision-making information including risk level and impact range; on the other hand, by automatically adjusting relevant equipment parameters, it can proactively suppress the occurrence or spread of malfunctions in their early stages, truly achieving the ultimate goal of ensuring patient safety. This is accomplished at the edge, ensuring the immediacy of the emergency response.

[0183] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A medical emergency early warning method based on digital twin technology, characterized in that, The application relates to a method for monitoring the risk of a linkage fault of multiple intelligent medical devices. Step one, constructing a digital twin model of multiple intelligent medical devices, wherein the digital twin model forms a virtual image according to hardware parameters and use environment data of the devices; Step two, collecting running data of the multiple intelligent medical devices in real time by using an edge computing device; Step three, pre-processing the collected running data by the edge computing device, and calculating a cooperative working state among the devices by the edge computing device through fusion analysis of the running data; Step four, monitoring linkage fault risks according to the cooperative working state by the edge computing device, and when detecting a linkage fault risk, performing virtual prediction by using the digital twin model to judge an extension influence range of the linkage fault risk; Step five, confirming the linkage fault risk according to the extension influence range, sending an instant early warning signal by the edge computing device, and automatically adjusting device parameters of the intelligent medical devices by the edge computing device through an emergency response mechanism.

2. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, The calculated cooperative working state among the devices includes: establishing a cooperative working state model among the multiple intelligent medical devices, and calculating quantitative indexes of the cooperative working state in real time according to the running data; The virtual prediction by using the digital twin model includes: combining preset mutual dependency relationships among the intelligent medical devices, virtually deducing the linkage fault risk to determine the extension influence range.

3. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, The virtual prediction by using the digital twin model includes: combining preset mutual dependency relationships among the intelligent medical devices, virtually deducing the linkage fault risk to determine the extension influence range.

4. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, The monitored linkage fault risks include: monitoring operation abnormities or state abnormities in the cooperative working state, and calculating the possibility that the abnormities affect other intelligent medical devices.

5. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, The automatically adjusted device parameters of the intelligent medical devices are used for reducing fault propagation of the linkage fault risk among the multiple intelligent medical devices; The sent instant early warning signal includes: sending the instant early warning signal to a terminal device of medical personnel; The multiple intelligent medical devices include a cardiac pacemaker and an insulin pump.

6. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, In step one, the method further includes: sub-step constructing a device mechanism model based on the hardware parameters the device mechanism model for describing the basic physical characteristics and operation logic of the intelligent medical device sub-step based on the usage environment data constructing a data-driven bias correction model the bias correction model for compensating the equipment mechanism model state bias in actual operation; Sub-step , fusing the equipment mechanism model with the deviation correction model to construct the digital twin model, the virtual state output by the digital twin model is the virtual image, and the calculation formula of the virtual state is: ， in, For the current time, It is in a virtual state. For the device mechanism model The running function, For control input signals of intelligent medical devices. For hardware parameters, For the deviation correction model The correction function, This refers to the usage environment data at the current moment; Sub-step , set a model fidelity threshold , and utilize actual operation data of the intelligent medical device , calculate the virtual state in real time fidelity , the calculation formula of the real-time fidelity is: ， wherein, for intelligent medical devices in real-time running data, is the real-time fidelity; Sub-step , performing a model update decision if the real-time fidelity is less than the model fidelity threshold , triggering an online update of the bias correction model with the latest usage environment data re-training the bias correction model to ensure accuracy of the virtual image.

7. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, In step two, the method further includes: Sub-step , determining a device set of the plurality of intelligent medical devices , the device set contains devices, and is the first device in the device set configuring a data sampling period ;​ Sub-step , the edge computing device determines a system sampling time point according to a minimum value of the data sampling periods , the edge computing device polls the device set at the system sampling time point , and obtains an original data set of the first device ;​​​​​ sub-step , the original data set by the device state , the sensor feedback information with the environmental change information ; sub-step performing a data validity decision on the sensor feedback information in the original data set , the data validity decision being conditional on: ≤ ≤ ， ， ， wherein, is the first station device sensor lower range limit, is the first station device sensor upper range limit, is data valid, is data invalid; sub-step to the original data set performing a data integrity decision the data integrity decision is conditioned on: ， wherein, is data field missing, is data complete, is data missing; Sub-step When the data validity is determined The result is And the data integrity determination The result is At that time, the edge computing device is the original dataset. Additional synchronization timestamp And will be appended with the synchronization timestamp. The original dataset Store in edge local data cache ; sub-step , the edge local data cache area The data set stored in the edge local data cache area constitutes the running data, and the running data is used for the fusion analysis of step three.

8. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, In step three, the method further includes: Sub-step extracting the operational data from the edge local data buffer ;​​​ Sub-step , the sensor feedback information in the synchronization data set Perform data normalization preprocessing to generate normalized data , the calculation formula of the data normalization preprocessing is:​ ， wherein, is a first station device in normalized data at a time instant, is a first preset range minimum value of the station device sensor, is a first preset range maximum value of the station device sensor; sub-step , constructing the collaborative working state model among the plurality of intelligent medical devices , the collaborative working state model at the system sampling time state vector consisting of the normalized data of all devices , the expression of the state vector ​ ， wherein, is a state vector, is the total number of smart medical devices, is the vector transpose; sub-step , defining a reference state vector of the collaborative working state model , the reference state vector represents an ideal state when the plurality of intelligent medical devices collaboratively work Sub-step , calculating the cooperative deviation between the state vector and the reference state vector , the cooperative deviation is the quantitative index of the cooperative working state, and the calculation formula of the cooperative deviation is:​ ， wherein, is a cooperative bias, is a first preset cooperative weight of the station device, is a reference state vector is a first component of the reference state vector; Sub-step Set a cooperative deviation threshold The edge computing device is based on the cooperative deviation. With the aforementioned collaborative deviation threshold The comparison determines the current collaborative working state. ; Sub-step When the cooperative work state The determination condition is: ， wherein the cooperative working state provides input for the monitoring of the risk of a linkage failure of step four.

9. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, In step four, the method further includes: Sub-step receiving the cooperative working state of the step three determination , and setting up linkage fault risk monitoring trigger condition ; Sub-step , the linkage fault risk monitoring trigger condition is: when the cooperative working state is cooperative exception, the edge computing device starts the linkage fault risk monitoring; Sub-step When the linkage fault risk monitoring trigger condition When the condition is met, the edge computing device calculates the collaborative deviation. The device contribution of each of the aforementioned intelligent medical devices The contribution of the device The calculation formula is: ， wherein, is the first preset cooperative weight of the station device, is the first normalization data of the station device, is the first component in the reference state vector ; Sub-step , determining the device contribution degree Maximum device Source abnormal device , the source abnormal device Index The determination formula is: ， The source abnormal device That is, constitutes the linkage fault risk; Sub-step , load the digital twin model of the plurality of intelligent medical devices built, and obtain a preset interdependence matrix among the intelligent medical devices , the interdependence matrix The element of the interdependence matrix represents the influence coefficient of the device on the device . sub-step , defining an exception state of the source exception device , the exception state being represented by the normalized data of the source exception device ;​ Sub-step , the edge computing device performs the virtual prediction in the digital twin model, utilizing the abnormal state With the interdependence matrix , the linkage fault risk of the other devices The propagation probability of the fault ; Sub-step The calculation formula of the fault propagation probability is: ， wherein, is a pre-set fault propagation function, is a source abnormal device to a device the impact coefficient of said device, is a prediction time step; sub-step , set a fault propagation threshold , the edge computing device determines all the devices whose fault propagation probability is greater than the fault propagation threshold as affected devices ; sub-step a set of all said affected devices i.e. the extended impact range constituting said cascading failure risk, said extended impact range is determined by the condition that ， The extension impact range The confirmation linkage failure risk for the step five.

10. The medical emergency early warning method based on digital twin technology according to claim 1, characterized in that, In step five, the method further includes: sub-step receiving the extended impact range of the step four determination the extended impact range comprising all of the affected devices; Sub-step , the edge computing device calculates the comprehensive risk level of the linkage fault risk according to the extended influence range , the comprehensive risk level The calculation formula is:​ ， wherein, to extend the range of influence in the first the affected device's preset criticality weight, to calculate the first the affected device's failure propagation probability; sub-step , setting a risk confirmation threshold , the edge computing device performs the determination of the confirmation of the risk of the cascading failure , the determination is conditional on: When , the risk of a linkage failure has been identified; When ≤ , linkage failure risk not confirmed Sub-step When the result of the determination is confirmed, the edge computing device generates the instant warning signal , the instant warning signal contains the comprehensive risk level and the extended impact range , and the edge computing device sends the instant warning signal to the terminal device of the medical personnel; Sub-step When the result of the determination is confirmed, the edge computing device starts the emergency response mechanism, and the emergency response mechanism loads a corresponding emergency adjustment strategy according to the comprehensive risk level and the extended impact range ; Sub-step The emergency adjustment strategy For the extended range of influence Each of the affected devices mentioned above Calculate new equipment parameters The new device parameters The calculation formula is: , wherein a current device parameter of the affected device, a criticality weight, a preset parameter adjustment function, the parameter adjustment function calculating a parameter adjustment amount according to the comprehensive risk level and the criticality weight and the criticality weight. Sub-step The edge computing device extends its influence to the surrounding area. The affected devices in Issue parameter adjustment instructions to apply the new device parameters. This completes the automatic adjustment of the device parameters of the intelligent medical device.

Images