Medical equipment operation and maintenance system based on artificial intelligence
The AI-based medical equipment operation and maintenance system solves the problems of difficulty in timely detection of potential faults and data recording errors in traditional operation and maintenance models. It enables real-time and accurate monitoring and intelligent management of equipment status, ensuring medical safety and the scientific nature of operation and maintenance decisions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-24
- Publication Date
- 2026-03-31
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional medical equipment operation and maintenance models rely on manual inspections, which makes it difficult to detect potential equipment failures in a timely manner, leads to frequent data recording errors, makes it difficult to achieve real-time and accurate monitoring, and fails to meet the refined and intelligent management needs of modern medical equipment.
An AI-based medical equipment operation and maintenance system is adopted, including a monitoring module, a monitoring and analysis module, an equipment evaluation module, and an operation and maintenance analysis module. Through real-time monitoring, fault prediction, and lifespan prediction, combined with optimization models and evaluation curves, intelligent management of equipment status is achieved.
It enables real-time, continuous, and precise monitoring of medical equipment, reduces the risk of missed fault detection, ensures the integrity and accuracy of data records, and provides reliable support for operation and maintenance decisions.
Smart Images
Figure CN121768616A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical equipment operation and maintenance technology, specifically an artificial intelligence-based medical equipment operation and maintenance system. Background Technology
[0002] In the medical field, the stable operation and efficient maintenance of medical equipment are crucial for ensuring the quality of medical services, improving the patient experience, and ensuring medical safety. Traditional medical equipment operation and maintenance models primarily rely on regular manual inspections, manual recording of equipment operating data, and experience-based fault diagnosis and handling. This approach not only consumes significant human and material resources but also struggles to achieve real-time and accurate monitoring of equipment status. Due to the time intervals inherent in manual inspections, some potential equipment malfunctions may not be detected early enough. By the time obvious symptoms appear, medical services have often already been impacted, potentially leading to medical accidents and endangering patient lives. Furthermore, manual data recording is inefficient and prone to errors and omissions, compromising the completeness and accuracy of equipment operating data and consequently affecting the scientific validity and rationality of maintenance decisions based on this data. With the rapid development of the medical industry, the continuous expansion of hospital scale, and the increasing number and complexity of medical equipment, traditional operation and maintenance models are no longer sufficient to meet the requirements of refined and intelligent management of modern medical equipment. Therefore, an innovative medical equipment operation and maintenance system is urgently needed to change this situation.
[0003] In order to solve the above problems, this invention provides an artificial intelligence-based medical device operation and maintenance system. Summary of the Invention
[0004] To address the problems of the above solutions, this invention provides an artificial intelligence-based medical device operation and maintenance system.
[0005] The objective of this invention can be achieved through the following technical solutions: An artificial intelligence-based medical device operation and maintenance system includes a monitoring module, a monitoring and analysis module, an equipment evaluation module, and an operation and maintenance analysis module. The monitoring module is used to monitor each medical device in real time and obtain the device monitoring data of the medical device.
[0006] The monitoring and analysis module is used to perform real-time monitoring and analysis of medical equipment based on equipment monitoring data, and to obtain real-time fault prediction data and lifespan prediction data of the medical equipment.
[0007] Furthermore, real-time monitoring and analysis of medical equipment includes: A preset model reserve library is provided for storing fault prediction models and lifespan prediction models for various medical devices. Identify information about each medical device, match the corresponding fault prediction model and life prediction model from the model reserve based on the medical device information, and optimize and adjust the matched fault prediction model and life prediction model. The equipment monitoring data is analyzed by optimizing the fault prediction model and the life prediction model to obtain the corresponding fault prediction data and life prediction data.
[0008] Furthermore, the matched fault prediction model and lifetime prediction model are optimized and adjusted, including: Verification data is generated based on historical monitoring data of medical equipment; the verification data is analyzed using a fault prediction model and a lifespan prediction model to obtain fault prediction data and lifespan prediction data. Based on the verification data, analyze the fault prediction data and life prediction data to determine whether the fault prediction model and life prediction model need to be optimized and adjusted. When it is determined that no adjustment is needed to the fault prediction model and the lifetime prediction model, no optimization adjustment is performed; When it is determined that adjustments to the fault prediction model and / or life prediction model are necessary, the corresponding fault prediction model and / or life prediction model shall be optimized and adjusted based on the validation data.
[0009] Furthermore, the failure prediction data and lifespan prediction data are analyzed based on the verification data, including: Establish a validation and evaluation model. The expression for the validation and evaluation model is as follows: ; In the formula: s represents the input data, including validation data, failure prediction data, and lifetime prediction data; the output data is the validation evaluation value YG(s), which is 1 or 0. By analyzing the validation data, failure prediction data, and life prediction data through the validation evaluation model, the corresponding validation evaluation values are obtained. When the verification evaluation value is 1, it is determined that the failure prediction model and / or lifetime prediction model need to be adjusted. When the verification evaluation value is 0, it is determined that no adjustment is needed to the fault prediction model and the lifetime prediction model.
[0010] The equipment evaluation module is used to evaluate medical equipment in real time based on fault prediction data and lifespan prediction data, and obtain an equipment evaluation curve. The horizontal axis of the equipment evaluation curve is time, and the vertical axis is the equipment score.
[0011] Furthermore, real-time evaluation of medical devices is performed based on failure prediction data and lifespan prediction data, including: Identify the corresponding prediction time based on the fault prediction data, evaluate the fault score corresponding to the fault prediction data at the corresponding prediction time, and label the fault score with the corresponding prediction time label. Identify the corresponding prediction time based on the life prediction data, identify the predicted life corresponding to the life prediction data; convert the predicted life into the corresponding life score; and label the life score with the corresponding prediction time label. The equipment score corresponding to the predicted time is calculated based on the scoring formula, which is: PG = 100 - GZ - SZ; In the formula: PG is the equipment score; GZ and SZ are the fault score and lifespan score corresponding to the corresponding prediction time, respectively; Based on the equipment scores corresponding to each prediction time, a corresponding equipment evaluation curve is generated.
[0012] Furthermore, the fault score corresponding to the fault prediction data is evaluated, including: A preset functional rating table for the medical device is used to statistically analyze the functional failure values corresponding to different effects on the different functions of the medical device. Based on fault prediction data, identify corresponding equipment faults in real time; When no equipment fault is identified, the fault score for the corresponding predicted time is 0; When a device malfunction is detected, the functional impact data corresponding to the device malfunction is determined. Based on the functional impact data, the corresponding functional malfunction value is matched from the functional scoring table. The functional malfunction values corresponding to the device malfunction are accumulated to obtain a single malfunction score for the device malfunction. The single malfunction scores corresponding to each device malfunction are accumulated to obtain a malfunction score for the predicted time.
[0013] The operation and maintenance analysis module is used to perform operation and maintenance analysis, obtain various operation and maintenance methods of medical equipment, simulate and analyze the operation and maintenance methods based on equipment evaluation curves, obtain the operation and maintenance time of the operation and maintenance methods, and display the operation and maintenance time and operation and maintenance methods to the relevant staff.
[0014] Furthermore, based on equipment evaluation curves, simulation analysis of operation and maintenance methods is conducted, including: Based on the operation and maintenance method, conduct operation and maintenance simulation to obtain the operation and maintenance simulation curve for the corresponding time period; The operation and maintenance simulation curve is compared with the equipment evaluation curve to obtain the corresponding operation and maintenance evaluation results; the operation and maintenance time of the corresponding operation and maintenance method is determined based on the operation and maintenance evaluation results.
[0015] Furthermore, the operation and maintenance simulation curves are compared with the equipment evaluation curves, including: The operation and maintenance simulation curve and the equipment evaluation curve are fitted separately to obtain the operation and maintenance simulation function and the equipment evaluation function. The operation and maintenance simulation function and the equipment evaluation function are labeled as MP(t) and SP(t) respectively, where t is time. Determine the comparison duration, and based on the comparison duration and the operation and maintenance simulation curve, determine the comparison period, and mark the comparison period as [t1, t2]; The difference between the operation and maintenance simulation curve and the equipment evaluation curve is calculated using a comparison formula, which is: ; In the formula: BP is the comparison difference; When the difference between the comparison values is greater than the threshold X1, the evaluation of the operation and maintenance simulation meets the qualification requirements. When the difference between the comparison values is not greater than the threshold X1, the evaluation of the operation and maintenance simulation does not meet the qualification requirements.
[0016] Compared with the prior art, the beneficial effects of the present invention are: The AI-based medical equipment operation and maintenance system proposed in this invention effectively overcomes the limitations of traditional operation and maintenance models, achieving an intelligent upgrade of medical equipment operation and maintenance. Through the deep application of AI technology, the system can monitor the equipment's operating status in real time, continuously, and accurately, significantly reducing the risk of missed fault detection due to manual inspection intervals. It ensures that potential problems are identified and warned of at an early stage, effectively preventing equipment failures from impacting medical services and truly safeguarding medical safety and patient lives. Simultaneously, the system's automated data collection and processing mechanism completely eliminates the inefficient traditional manual recording mode, significantly improving data recording efficiency and fundamentally eliminating human error and data omissions. This ensures the integrity and accuracy of equipment operating data, providing solid and reliable data support for operation and maintenance decisions. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a block diagram illustrating the principle of the present invention. Detailed Implementation
[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.
[0020] like Figure 1 As shown, an artificial intelligence-based medical device operation and maintenance system includes a monitoring module, a monitoring and analysis module, an equipment evaluation module, and an operation and maintenance analysis module. The monitoring module is used to perform real-time monitoring of various medical devices that require operational monitoring based on existing Internet of Things (IoT) technology, and to obtain the device monitoring data of the corresponding medical devices.
[0021] For example, by using Internet of Things (IoT) technology, medical devices can be connected to sensors, monitoring systems, etc., to achieve real-time collection of device operating status, performance parameters, environmental data, etc.; the data collection range is wide, including but not limited to key indicators such as device temperature, pressure, vibration, power consumption, etc.
[0022] The monitoring and analysis module is used to perform real-time monitoring and analysis of medical equipment based on equipment monitoring data, and to obtain real-time fault prediction data and lifespan prediction data of medical equipment.
[0023] In one embodiment, both fault prediction and lifespan prediction can be performed using existing mature fault prediction technologies and equipment lifespan prediction technologies.
[0024] For example, fault prediction: Data Acquisition: Collect equipment operation data, including sensor data (such as temperature, pressure, vibration, current, etc.), equipment status data (such as power on / off status, operating mode, etc.), process parameter data (such as equipment load, running time, etc.), and log records (such as error codes, alarm information, maintenance records, etc.); ensure the accuracy, completeness, and real-time nature of the data, which can be acquired in real time through sensors or extracted from the equipment control system and log files.
[0025] Data preprocessing: data cleaning, data transformation, data normalization; Feature engineering: Extracting features that have a significant impact on fault prediction from the raw data, such as time-domain features (mean, variance, peak value, etc.), frequency-domain features (extracted through Fourier transform), time-frequency domain features (such as short-time Fourier transform, STFT), and statistical features (such as kurtosis, skewness, etc.); using statistical feature-based methods (such as mutual information, chi-square test) or machine learning algorithms (such as L1 regularization) to select the most influential features, reduce data dimensionality, and avoid overfitting.
[0026] Model selection: Select a suitable fault prediction model based on data characteristics and prediction needs. Common models include neural networks, support vector machines, decision trees, and random forests.
[0027] Model training: The selected model is trained using preprocessed data, and model parameters are adjusted to improve performance; for example, in neural networks, weights and biases are adjusted through backpropagation to minimize the loss function; cross-validation methods (such as K-fold cross-validation) are used to evaluate the model's generalization ability and avoid overfitting.
[0028] Model evaluation: The trained model is evaluated using test data. Evaluation metrics include accuracy, recall, F1 score, area under the ROC curve (AUC), etc. The model's prediction results are analyzed to identify its strengths and weaknesses, providing a basis for model optimization.
[0029] Model deployment: Deploy the trained model into the actual production environment to achieve real-time prediction of equipment failures.
[0030] Life expectancy prediction: Data collection: Collect data related to equipment lifespan, including historical fault records, operating parameters (such as temperature, pressure, load, etc.), environmental conditions (such as temperature, humidity, vibration, etc.), and maintenance records (such as maintenance logs, fault repair lists, component replacement records, etc.).
[0031] Feature engineering: Extract useful features for life prediction from the collected data, such as equipment usage frequency, load conditions, cumulative operating time, historical failure count, maintenance frequency, etc.; combine physical model knowledge to construct fault sensitivity indicators, such as "vibration energy ratio" or "response delay accumulation", to reveal the degradation law of equipment; use feature selection algorithms (such as principal component analysis PCA, L1 regularization) to screen out the features with the greatest impact on life prediction and reduce data dimensionality.
[0032] Model selection: Choose an appropriate lifetime prediction model based on the nature of the problem and the characteristics of the data; common models include: Physics-based degradation models: These models are built by analyzing the degradation patterns and historical data of equipment during use to predict its remaining service life. For example, a life distribution model is established for key components (such as life fitting for key components like pumps, valves, and sensors), and the remaining service life (RUL) is estimated based on actual operating conditions.
[0033] Data-driven statistical models: Utilizing historical failure data and normal operation data, machine learning algorithms (such as support vector machines, neural networks, random forests, etc.) are used to mine failure patterns and predict the remaining lifespan of equipment.
[0034] Hybrid models combine the advantages of physical models and data-driven methods, using physical constraints to improve the generalization of data-driven models and enhance prediction accuracy.
[0035] Model training: The selected model is trained using the collected data, and the model parameters are adjusted to optimize the prediction performance; for example, in neural networks, weights and biases are adjusted through the backpropagation algorithm to minimize the loss function; in physics-based degenerate models, model parameters are adjusted by fitting historical data; cross-validation is used to evaluate the model's generalization ability and avoid overfitting.
[0036] Model validation: The trained model is validated using an independent dataset to evaluate its predictive accuracy and reliability. Validation metrics may include the error in remaining lifetime prediction and the confidence level of the prediction interval. The model's prediction results are analyzed to identify its strengths and weaknesses, providing a basis for model optimization.
[0037] Model optimization: Adjust and optimize the model based on the validation results, such as adjusting model parameters, improving feature engineering methods, and trying different model algorithms to improve the accuracy and reliability of predictions; customize the model by considering the actual use of the equipment and environmental conditions.
[0038] Model deployment: Deploying the trained model into real-world applications for real-time or periodic lifetime prediction.
[0039] In one embodiment, in order to improve the efficiency of resource utilization, operators, platform providers, etc. can pre-set a model reserve library. The model reserve library is used to store fault prediction models and life prediction models of various medical devices, which are established by the platform provider or obtained from other channels. Identify information about each medical device, match the corresponding fault prediction model and life prediction model from the model reserve based on the medical device information, and optimize and adjust the matched fault prediction model and life prediction model, that is, optimize and adjust it using actual medical device information to make it more suitable for this medical device. The equipment monitoring data is analyzed by optimizing the fault prediction model and the life prediction model to obtain the corresponding fault prediction data and life prediction data.
[0040] In one embodiment, the matched fault prediction model and life prediction model are optimized and adjusted based on existing methods, such as directly using relevant historical data to set optimization data, or the optimization is carried out by platform staff.
[0041] In one embodiment, optimizing and adjusting the matched fault prediction model and lifetime prediction model includes: Verification data is generated based on historical monitoring data of medical devices. The verification data includes the corresponding device monitoring data, as well as standard fault diagnosis results and standard lifespan diagnosis results. It can be set by the user or the platform, or it can be generated intelligently. By analyzing the corresponding verification data through fault prediction models and life prediction models, the corresponding fault prediction data and life prediction data are obtained. Based on the verification data, analyze the fault prediction data and life prediction data to determine whether the fault prediction model and life prediction model need to be optimized and adjusted. When it is determined that no adjustment is needed to the fault prediction model and the lifetime prediction model, no optimization adjustment is performed; When it is determined that adjustments to the fault prediction model and / or life prediction model are necessary, the corresponding fault prediction model and / or life prediction model shall be optimized and adjusted based on the validation data.
[0042] In one embodiment, fault prediction data and lifespan prediction data are analyzed based on verification data, and the analysis and judgment are made based on existing technologies.
[0043] In one embodiment, analyzing failure prediction data and lifetime prediction data based on verification data includes: Establish a validation and evaluation model. The expression for the validation and evaluation model is as follows: ; In the formula: s represents the input data, including validation data, fault prediction data, and life prediction data; s represents abnormal data, indicating that at least one of the fault prediction data and life prediction data does not meet the requirements of the validation data, such as the deviation exceeding the preset value, etc. The platform presets the corresponding validation requirements; the corresponding training set is labeled using the corresponding historical data for training; the output data is the validation evaluation value YG(s), which is 1 or 0. By analyzing the validation data, failure prediction data, and life prediction data through the validation evaluation model, the corresponding validation evaluation values are obtained. When the verification evaluation value is 1, it is determined that the failure prediction model and / or lifetime prediction model need to be adjusted. When the verification evaluation value is 0, it is determined that no adjustment is needed to the fault prediction model and the lifetime prediction model.
[0044] The equipment evaluation module is used to evaluate medical equipment in real time based on fault prediction data and lifespan prediction data, and obtain an equipment evaluation curve. The horizontal axis of the equipment evaluation curve is time, and the vertical axis is the equipment score.
[0045] In one embodiment, real-time evaluation of a medical device based on fault prediction data and lifespan prediction data includes: Identify the corresponding prediction time based on the fault prediction data, evaluate the fault score corresponding to the fault prediction data, and label the fault score with the corresponding prediction time label. Based on lifespan prediction data, the corresponding prediction time is identified, and the predicted lifespan corresponding to the lifespan prediction data is identified. The predicted lifespan is converted into a corresponding lifespan score. The lifespan score is set according to the magnitude of the adverse impact on the medical device at different lifespans. The greater the adverse impact, the higher the lifespan score. The platform can preset lifespan scores corresponding to different lifespans based on the available lifespans. Furthermore, the medical device can be simulated based on technologies such as digital twins. The simulation data can be used to adjust the lifespan conversion method and the fault score evaluation method. For example, if the score of the medical device in its ideal state is 100, the device state under different remaining lifespans and fault conditions can be simulated. The corresponding score is determined based on the difference between the device state and the ideal state. Then, the scoring method of fault score and lifespan score can be optimized in reverse based on the score to improve the scoring accuracy. That is, the functional fault values of different functions under different impact levels and the lifespan scores corresponding to different predicted lifespans are adjusted. The corresponding prediction time label is marked on the lifespan score. The equipment score corresponding to the predicted time is calculated based on the scoring formula, which is: PG = 100 - GZ - SZ; In the formula: PG is the equipment score; GZ and SZ are the fault score and lifespan score corresponding to the corresponding prediction time, respectively; Based on the equipment scores corresponding to each prediction time, a corresponding equipment evaluation curve is generated.
[0046] In one embodiment, evaluating the fault score corresponding to the fault prediction data includes: The platform acquires various equipment faults of medical devices, and based on the impact of these faults, it presets corresponding fault scores for each fault. The scores are set according to the impact on the use of the medical devices; the greater the impact, the higher the fault score. These scores are then integrated into a fault score table. The corresponding fault score is matched in real time based on the fault prediction data.
[0047] In one embodiment, the fault score corresponding to the fault prediction data can also be evaluated based on existing methods, such as using intelligent models built based on machine learning, deep learning algorithms, etc. for analysis.
[0048] In one embodiment, evaluating the fault score corresponding to the fault prediction data includes: Identify the various functions of medical devices, such as medical functions, data acquisition functions, protective functions, and mobility functions; preset corresponding functional fault values for different functions affected by different factors, which are set by the platform and integrated into a functional scoring table; that is, the functional scoring table is used to statistically analyze the functional fault values corresponding to different functions of medical devices affected by different factors.
[0049] Based on fault prediction data, identify corresponding equipment faults in real time; When no device malfunction is identified, the malfunction score is 0; When a device malfunction is detected, the functions affected by the malfunction and the degree of functional impact are determined and summarized into functional impact data for the corresponding functions. Based on the functional impact data, the corresponding functional malfunction values are matched from the functional scoring table. The functional malfunction values corresponding to the device malfunction are accumulated to obtain the individual malfunction score of the device malfunction, that is, the sum of the functional malfunction values of each function. The individual malfunction scores corresponding to each device malfunction are accumulated to obtain the malfunction score for the corresponding predicted time. If there is only one device malfunction, the individual malfunction score of that device malfunction is the malfunction score.
[0050] The operation and maintenance analysis module is used to perform operation and maintenance analysis, obtain various operation and maintenance methods of medical equipment, simulate and analyze the operation and maintenance methods based on equipment evaluation curves, determine the operation and maintenance time of the corresponding operation and maintenance methods, and display the operation and maintenance time and operation and maintenance methods to the relevant staff; subsequently, the staff can perform operation and maintenance on the medical equipment according to the operation and maintenance time and operation and maintenance methods.
[0051] In one embodiment, the simulation analysis of the operation and maintenance method based on the equipment evaluation curve includes: The simulation involves applying appropriate maintenance methods to medical equipment at a given time, simulating the equipment monitoring data after maintenance, and then determining the changes in the equipment evaluation curve after maintenance to form a maintenance simulation curve. In other words, the maintenance simulation curve is generated according to the method of generating equipment monitoring data and equipment evaluation curves after simulated maintenance; that is, maintenance simulation is performed according to the maintenance method to obtain the maintenance simulation curve for maintenance at the given time. The operation and maintenance simulation curve is compared with the equipment evaluation curve to determine whether the operation and maintenance of the corresponding operation and maintenance method meets the qualification requirements, that is, whether the operation and maintenance is qualified according to the corresponding operation and maintenance method at the corresponding time, and obtain the corresponding operation and maintenance evaluation result; the operation and maintenance time of the corresponding operation and maintenance method is determined based on the operation and maintenance evaluation result.
[0052] For example, if the operation and maintenance result is that the operation and maintenance simulation meets the qualification requirements, the corresponding simulation time is marked as the operation and maintenance time; if the operation and maintenance result is that the operation and maintenance simulation does not meet the qualification requirements and no operation and maintenance is required, then the operation and maintenance time is zero, indicating that no operation and maintenance is required.
[0053] In one embodiment, operation and maintenance simulation is performed according to the operation and maintenance method, including: A simulation model of medical equipment is established based on digital twins and other methods. Operation and maintenance simulation is carried out based on the operation and maintenance method, and the corresponding operation and maintenance simulation curve is determined according to the operation and maintenance method at the corresponding time.
[0054] In one embodiment, operation and maintenance simulation can be performed based on the operation and maintenance method, or it can be performed based on other existing methods.
[0055] The simulation time for different maintenance methods needs to be set according to the actual situation of the medical equipment. For example, high-risk equipment, such as CT scanners, MRI machines, and ventilators, may cause diagnostic delays or treatment interruptions due to malfunctions, requiring high-frequency maintenance (such as daily inspections and quarterly in-depth maintenance); low-risk equipment, such as blood pressure monitors and thermometers, can have a reduced maintenance frequency, but a rapid response mechanism needs to be established. The specific settings can be configured by the user as needed, or preset by the platform.
[0056] In one embodiment, the operation and maintenance simulation curve is compared with the equipment evaluation curve. Based on existing technologies, comparative analysis can be conducted to determine whether the corresponding operation and maintenance is necessary. For example, the judgment can be made based on the difference between the two subsequent maximum equipment scores. If the difference is greater than a preset value, it is considered to meet the qualification requirements. Alternatively, an intelligent analysis model can be established based on machine learning, deep learning algorithms, etc., for analysis.
[0057] In one embodiment, comparing the operation and maintenance simulation curve with the equipment evaluation curve includes: The operation and maintenance simulation curve and the equipment evaluation curve are fitted separately to obtain the corresponding operation and maintenance simulation function and equipment evaluation function. The operation and maintenance simulation function and the equipment evaluation function are labeled as MP(t) and SP(t) respectively, where t is time. Determine the comparison duration, i.e., the time interval for comparison. Start timing from the start of the operation and maintenance simulation. The comparison period can be set according to the different medical equipment and the differences in operation and maintenance goals, such as 1 hour, 4 hours, 8 hours, 12 hours, etc. For example, the step-by-step expansion method: initially select a short period of time (such as 1 hour) for rapid verification. If the result is significant, it is adopted directly; if it is not significant, the period is extended. Gradually expand the period of time in increments of 2 hours (1→3→5 hours) until the data difference is stable.
[0058] The comparison period is determined based on the comparison duration and the operation and maintenance simulation curve, that is, the time from the start of the simulation to the time corresponding to the addition of the comparison duration; the comparison period is marked as [t1, t2]; The difference between the operation and maintenance simulation curve and the equipment evaluation curve is calculated using a comparison formula, which is: ; In the formula: BP is the comparison difference; When the difference between the comparison values is greater than the threshold X1, the evaluation of the operation and maintenance simulation meets the qualification requirements. When the difference between the comparison values is not greater than the threshold X1, the evaluation of the operation and maintenance simulation does not meet the qualification requirements. The threshold X1 is set by the platform and is mainly based on the value and medical importance of the medical device, because the maintenance cost of some medical devices is relatively cost-effective compared to their lifespan and the impact of failure.
[0059] In one embodiment, the operation and maintenance cost of the operation and maintenance method can be added to the comparison formula to improve the accuracy of the assessment by comprehensively considering the operation and maintenance cost; for example, the comparison formula is: ; In the formula: BP is the comparison difference; CB is the operation and maintenance cost.
[0060] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.
[0061] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. 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 be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.
Claims
1. An artificial intelligence-based medical equipment operation and maintenance system, characterized in that, The monitoring module, the monitoring analysis module, the equipment evaluation module, and the operation and maintenance analysis module are included. The monitoring module is configured to monitor each medical device in real time to obtain device monitoring data of the medical device. The monitoring analysis module is configured to perform real-time monitoring analysis on the medical device based on the device monitoring data to obtain real-time failure prediction data and life prediction data of the medical device. The equipment evaluation module is configured to perform real-time evaluation on the medical device based on the failure prediction data and the life prediction data to obtain an equipment evaluation curve, wherein the horizontal axis of the equipment evaluation curve is time, and the vertical axis is an equipment score. The operation and maintenance analysis module is configured to perform operation and maintenance analysis, obtain various operation and maintenance modes of the medical device, simulate and analyze the operation and maintenance modes based on the equipment evaluation curve, and obtain operation and maintenance time of the operation and maintenance modes.
2. The medical equipment operation and maintenance system based on artificial intelligence according to claim 1, characterized in that, Real-time monitoring and analysis of medical devices include: A preset model repository is provided, and the model repository is configured to store failure prediction models and life prediction models of various medical devices. Each medical device information is identified, and the corresponding failure prediction model and life prediction model are matched from the model repository based on the medical device information. The matched failure prediction model and life prediction model are optimized and adjusted.
3. The medical equipment operation and maintenance system based on artificial intelligence according to claim 2, characterized in that, The device monitoring data is analyzed by using the optimized failure prediction model and life prediction model to obtain corresponding failure prediction data and life prediction data. Optimizing and adjusting the matched failure prediction model and life prediction model includes: Validation data is generated based on historical monitoring data of the medical device. The failure prediction data and the life prediction data are analyzed based on the validation data to determine whether the failure prediction model and the life prediction model need to be optimized and adjusted. When it is determined that the failure prediction model and the life prediction model do not need to be adjusted, no optimization and adjustment are performed.
4. The medical equipment operation and maintenance system based on artificial intelligence according to claim 3, characterized in that, When it is determined that the failure prediction model and / or the life prediction model need to be adjusted, the corresponding failure prediction model and / or the life prediction model are optimized and adjusted based on the validation data. Analyzing the failure prediction data and the life prediction data based on the validation data includes: ； A validation evaluation model is established, and the expression of the validation evaluation model is: In the formula, s is input data including validation data, failure prediction data, and life prediction data; and output data is a validation evaluation value YG(s), which is 1 or 0. The validation data, the failure prediction data, and the life prediction data are analyzed by using the validation evaluation model to obtain corresponding validation evaluation values. When the validation evaluation value is 1, it is determined that the failure prediction model and / or the life prediction model need to be adjusted.
5. The medical equipment operation and maintenance system based on artificial intelligence according to claim 1, characterized in that, When the validation evaluation value is 0, it is determined that the failure prediction model and the life prediction model do not need to be adjusted. Real-time evaluation of the medical device based on the failure prediction data and the life prediction data includes: A corresponding prediction time is identified based on the failure prediction data, and a failure score corresponding to the failure prediction data of the corresponding prediction time is evaluated. The corresponding prediction time label is marked for the failure score. According to the life prediction data, a corresponding prediction time is identified, and a prediction life corresponding to the life prediction data is identified; the prediction life is converted into a corresponding life score; and the life score is marked with a corresponding prediction time label; According to the scoring formula, a device score corresponding to the corresponding prediction time is calculated, and the scoring formula is: PG=100-GZ-SZ; In the formula: PG is the device score; GZ and SZ are the failure score and the life score corresponding to the corresponding prediction time, respectively; According to the device score corresponding to each prediction time, a corresponding device evaluation curve is generated.
6. The medical equipment operation and maintenance system based on artificial intelligence according to claim 5, characterized in that, The failure score corresponding to the failure prediction data is evaluated, including: A function score table of the medical device is preset, and the function score table is used to count the function failure value corresponding to the different influences on different functions of the medical device; According to the failure prediction data, a corresponding device failure is identified in real time; When no device failure is identified, the failure score of the corresponding prediction time is 0; When a device failure is identified, the function impact data corresponding to the device failure is determined, the corresponding function failure value is matched from the function score table according to the function impact data, the function failure value corresponding to the device failure is accumulated to obtain the single failure score of the device failure; and the single failure scores corresponding to each device failure are accumulated to obtain the failure score of the prediction time.
7. The medical equipment operation and maintenance system based on artificial intelligence according to claim 1, characterized in that, Based on the device evaluation curve, a simulation analysis of the operation and maintenance mode is performed, including: According to the operation and maintenance mode, an operation and maintenance simulation is performed to obtain an operation and maintenance simulation curve for operation and maintenance at a corresponding time; The operation and maintenance simulation curve and the device evaluation curve are compared to obtain a corresponding operation and maintenance evaluation result; and the operation and maintenance time of the corresponding operation and maintenance mode is determined according to the operation and maintenance evaluation result.
8. The medical equipment operation and maintenance system based on artificial intelligence according to claim 7, characterized in that, The comparison between the operation and maintenance simulation curve and the device evaluation curve includes: The operation and maintenance simulation curve and the device evaluation curve are fitted respectively to obtain an operation and maintenance simulation function and a device evaluation function, and the operation and maintenance simulation function and the device evaluation function are marked as MP(t) and SP(t), respectively, t being time; A comparison duration is determined, a comparison period is determined according to the comparison duration and the operation and maintenance simulation curve, and the comparison period is marked as [t1, t2]; According to the comparison formula, a comparison difference value between the operation and maintenance simulation curve and the device evaluation curve is calculated, and the comparison formula is: ； In the formula: BP is the comparison difference value; When the comparison difference value is greater than a threshold X1, it is evaluated that the operation and maintenance simulation meets the qualified requirement; When the comparison difference value is not greater than the threshold X1, it is evaluated that the operation and maintenance simulation does not meet the qualified requirement.