Medical monitoring data transmission method, monitoring method and application method

Abstract

The invention belongs to the technical field of data transmission, and particularly relates to a medical monitoring data transmission method, a monitoring method and an application method.The data transmission link is subjected to standardized conversion, and the method comprises the steps that connection modes of different devices are adapted through a physical interface or a wireless communication protocol, and a communication link is established; key parameters in the data flow are analyzed in real time through a lightweight neural network model to obtain an analysis result, potential semantic features of the data flow are automatically identified through multi-modal signal analysis to obtain an identification result, and a temporary analysis rule is dynamically generated based on the analysis result and the identification result; and the disassembled independent parameters are recombined according to a unified standard format. According to the method, equipment hardware does not need to be transformed, continuous data transmission of various equipment is ensured by automatically establishing a standardized data stream transmission channel, the workload of manually configuring analysis rules can be reduced through implementation, the risk of wrong analysis caused by waveform distortion can be reduced through multi-modal signal analysis, and finally, the private format difference of the equipment is eliminated.

Description

Technical Field

[0001] The present invention belongs to the technical field of data transmission, and particularly relates to a method for transmitting medical monitoring data, a monitoring method, and an application method. Background Art

[0002] The core process of medical monitoring data transmission includes: real-time collection of patients' physiological data by devices, and after operations such as cleaning, encryption, and formatting, transmitting it to the cloud or server using a reliable protocol, storing it, and then performing analysis, visualization, and abnormal warning. At the same time, data security is ensured through backup, permission control, and system maintenance.

[0003] However, in the above process, the heterogeneity of medical devices may lead to difficulties in data integration. Specifically, since different types of monitoring devices, such as electrocardiogram monitors, ventilators, etc., often use private data formats or dedicated communication protocols, additional development of adaptation interfaces is required when transmitting to a unified platform. This not only increases the system complexity but also may cause the loss or misjudgment of key indicators due to data parsing errors, such as arrhythmia waveforms, thus affecting the timeliness of clinical decisions. The above is particularly prominent in scenarios of cross-department or multi-institution collaboration, which may lead to risks of delayed diagnosis or misdiagnosis. Summary of the Invention

[0004] The present invention provides a method for transmitting medical monitoring data, a monitoring method, and an application method, which can effectively solve the problems in the background art.

[0005] To achieve the above object, the technical solution of the present invention is as follows: A method for transmitting medical monitoring data, which performs standardized conversion in the data transmission link, including: Adapting the connection methods of different devices through a physical interface or a wireless communication protocol, and establishing a communication link for data stream transmission; Real-time analyzing key parameters in the data stream through a lightweight neural network model to obtain an analysis result, automatically identifying potential semantic features of the data stream through multi-modal signal analysis to obtain an identification result, and dynamically generating a temporary parsing rule based on the analysis result and the identification result to disassemble the data stream into independent parameters; Recombining each of the independent parameters according to a unified standard format, and pushing them to the target system through a standardized interface.

[0006] Further, automatically identifying potential semantic features of the data stream through multi-modal signal analysis includes: Cutting the data stream into adaptive time windows; Distinguishing the data modality of the current time window, extracting corresponding features based on the distinguishing result, and combining with the context to obtain a multi-dimensional feature tensor; Based on the multi-dimensional feature tensor, using a rule engine to identify the semantic features of the data stream.

[0007] Further, based on the multi-dimensional feature tensor, use a rule engine to identify the semantic features of the data stream, including: Use the parsing result as the dynamic input for the semantic reasoning of the rule engine; Use predefined rules as the static input for the semantic reasoning of the rule engine; The rule engine performs a structured analysis on the data stream based on the multi-dimensional feature tensor to obtain an analysis result, and matches the dynamic input and the static input based on the analysis result to identify the semantic features.

[0008] Further, the lightweight network model performs the following tasks: Receive the data stream, extract local temporal features in the data stream through convolution operations, model temporal dependencies through a gated recurrent unit, and capture long-term temporal dependencies through a temporal convolutional network, and output temporal features; compress the temporal features into a vector of fixed length to obtain a pooling feature; map the pooling feature to output parameters to generate structured key parameters; Among them, the kernel size and stride of the convolution operation are optimized for the key waveform features in the data stream, and / or when compressing the temporal features, neurons or channels that contribute the most to the prediction of the key parameters are preferentially retained.

[0009] Further, the kernel size and stride of the convolution operation are optimized for the key waveform features in the data stream, including: Identify key waveform features in the data stream through a data visualization tool; Calculate the initial value of the convolution kernel according to the width and sampling rate of the key waveform features; Calculate the initial value of the stride according to the interval and sampling rate of the key waveform features; Use a grid search method to automatically search for the optimal convolution kernel size and stride based on the initial values within a preset parameter range.

[0010] Further, dynamically generate temporary parsing rules, including: Initialize the basic parsing rule template according to the device type; Modify the field mapping relationship in the basic parsing rule template according to the recognition result; Optimize the parameter calculation logic of the modified rule according to the parsing result and generate the temporary parsing rule.

[0011] Further, during the establishment of the communication link, automatically identify the device type based on the device unique identifier and / or communication characteristics, and complete authentication and permission management.

[0012] Further, it also includes calculating a confidence score for each parsing result when executing the temporary parsing rule; When the confidence score is lower than a set threshold, artificial confirmation and feedback are triggered.

[0013] A medical monitoring method uses the above-mentioned medical monitoring data transmission method to transmit monitoring data.

[0014] An application method of the above-mentioned medical monitoring data transmission method synchronously transmits monitoring data of at least two of an electrocardiogram monitor, a ventilator, a blood oxygen monitor, a blood pressure monitor, an electroencephalogram monitor, a blood glucose monitor, a body temperature monitor, an infusion pump, and an anesthesia machine to the same target system.

[0015] Through the technical solution of the present invention, the following technical effects can be achieved: In the present invention, aiming at the data integration problem caused by device heterogeneity in the process of medical monitoring data transmission, by introducing a data standardization conversion mechanism, the data from different devices, such as electrocardiogram monitors, ventilators, etc., can be uniformly parsed, disassembled, and recombined, so as to ensure that key information is not lost, the risk of misjudgment is minimized, and the timeliness of clinical decision-making is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of the medical monitoring data transmission method; Figure 2 It is a flowchart of automatically identifying potential semantic features of data streams through multimodal signal analysis; Figure 3 It is a flowchart of identifying semantic features of data streams using a rule engine based on a multi-dimensional feature tensor; Figure 4 It is a flowchart of optimizing the convolution kernel size and stride of the convolution operation for key waveform features in the data stream. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the description of this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0020] Embodiment 1 A medical monitoring data transmission method, which performs standardized conversion in the data transmission link, such as Figure 1 shown, includes: Adapting the connection methods of different devices through a physical interface or a wireless communication protocol, and establishing a communication link for data stream transmission; Real-time analyzing key parameters in the data stream through a lightweight neural network model to obtain an analysis result, automatically identifying potential semantic features of the data stream through multimodal signal analysis to obtain an identification result, and dynamically generating a temporary parsing rule based on the analysis result and the identification result to disassemble the data stream into independent parameters; Recombining each independent parameter according to a unified standard format and pushing it to the target system through a standardized interface.

[0021] In this invention, aiming at the data integration problem caused by device heterogeneity in the medical monitoring data transmission process, by introducing a data standardization conversion mechanism, it realizes the unified parsing, disassembly and recombination of data from different devices, such as electrocardiogram monitors, ventilators, etc., so as to ensure that key information is not lost, the risk of misjudgment is minimized, and the timeliness of clinical decision-making is improved.

[0022] In the implementation process, a physical interface or a wireless communication protocol is used to connect different monitoring devices to build a stable communication link. This step ensures that even if there are differences in the connection methods between devices, the data collected by each device can be uniformly transmitted to the platform. Through the established data link, the real-time transmission of patients' physiological data is realized. To achieve better implementation effects, the data is preferably pre-cleaned, encrypted and formatted during the transmission process to ensure data integrity and security during the transmission process.

[0023] In the data transmission process, a lightweight neural network model is used to perform real-time parsing on the data stream, which can efficiently extract key parameters in the data, such as heart rate, blood oxygen saturation, etc., while ensuring low latency and being suitable for real-time monitoring scenarios. For various signals that may exist in the data stream, such as waveforms, noise information, etc., multi-modal signal analysis technology is used to automatically identify their potential semantic features. This step helps to detect abnormal waveforms, such as arrhythmia, and other clinically important implicit information, ensuring more comprehensive and accurate data parsing. The application of the lightweight neural network model enables real-time parsing of key physiological parameters. At the same time, multi-modal signal analysis effectively identifies the potential semantic information in the data stream, thus avoiding the loss or misjudgment of important indicators caused by parsing errors and improving the timeliness and accuracy of clinical decisions.

[0024] Based on the parsing results and semantic recognition results, the system dynamically generates temporary parsing rules adapted to the current data characteristics. This adaptive mechanism enables the system to handle situations where the data formats between devices are not unified or the data parsing difficulty is high. The mechanism of dynamically generating temporary parsing rules endows the system with a high degree of adaptability. No matter what kind of device or data format changes, it can quickly adjust the parsing strategy, thus ensuring that the system has good scalability and long-term adaptability. Using the generated temporary parsing rules, the original data stream is disassembled into independent parameter units to ensure that each parameter can be accurately captured and subsequently processed, thereby avoiding the omission or misjudgment of key indicators caused by improper data integration. Finally, the disassembled independent parameters are recombined according to a predefined standard format. The standardized data has good interoperability, facilitating unified processing and analysis by each backend system.

[0025] As a preference of the above embodiment, as Figure 2 shown, automatically identifying the potential semantic features of the data stream through multi-modal signal analysis includes: Cutting the data stream into adaptive time windows, which can be specifically cut according to the signal characteristics and dynamic changes of the data stream. The data within each time window has relatively stable characteristics, which helps with subsequent analysis. The adaptive window can automatically adjust the window length according to the actual fluctuations of the data, enabling effective capture of signals with different rhythms or change rates; Distinguishing the data modality of the current time window, extracting corresponding features based on the distinguishing result, and obtaining a multi-dimensional feature tensor in combination with the context. In specific implementation, the signal type to which the data in the current window belongs can be identified, such as: electrocardiogram waveform, respiratory signal, noise, etc.; based on the distinguishing result, corresponding features are specifically extracted, such as waveform morphology, spectral features, statistical indicators, etc., and the context information of the previous and subsequent windows is taken into account to construct a multi-dimensional feature tensor. This tensor can describe the comprehensive characteristics of the data from time domain, frequency domain, and even spatial dimensions; Based on the multi-dimensional feature tensor, the semantic features of the data stream are identified using a rule engine. In this step, an analysis method based on the rule engine is introduced. As a specific method in implementation, it can match and analyze the features in the tensor according to the preset rules, combined with clinical knowledge and experience, so as to automatically identify potential semantic features in the data stream, such as abnormal waveforms, trend changes, unexpected events, etc. The rule engine provides a transparent and interpretable way to convert complex feature data into clinically meaningful semantic information.

[0026] Through the organic combination of the adaptive time window, multi-modal feature extraction, and rule engine recognition, this optimization scheme can capture potential semantic features in medical monitoring data more accurately and comprehensively, effectively improving the accuracy of data analysis and the real-time warning ability.

[0027] As a preference of the above embodiment, as Figure 3 shown, based on the multi-dimensional feature tensor, using the rule engine to identify the semantic features of the data stream includes: Taking the parsing result as the dynamic input for the semantic reasoning of the rule engine; Taking the predefined rules as the static input for the semantic reasoning of the rule engine; The rule engine performs a structured analysis on the data stream based on the multi-dimensional feature tensor to obtain an analysis result, and based on the analysis result, matches the dynamic input and the static input to identify semantic features. In the matching process, the semantic features are identified by comparing the real-time data with the preset rules, making the description easier to understand.

[0028] In the above preferred scheme, the dynamic input only contains the key parameters obtained by parsing, and the data volume is small, which helps the system to quickly process and respond to changes, ensuring the timely capture of abnormal events. Combining with the structured analysis result, the rule engine can quickly match the predefined rules, reducing the unnecessary computational burden and improving the overall processing efficiency. By only extracting the key parameters, the system will focus on the most clinically significant information, avoiding interference caused by excessive signal details, reducing data redundancy, making the rule matching and semantic reasoning process clearer, and being easy to maintain and upgrade.

[0029] In the implementation process, the comprehensive judgment basis provided by the structured analysis enables the system to still operate stably and accurately judge the abnormalities of key parameters even in the case of large data noise or short-term fluctuations. The system can be flexibly adapted by adjusting the predefined rules according to different clinical scenarios or device characteristics without having to redesign the data parsing part. The predefined rules in the static input provide a clear basis for the system's judgment, facilitating clinical personnel to understand and trust the system output. The structured analysis and rule matching process make each abnormal judgment well-founded, enhancing the verifiability and clinical practicality of the system.

[0030] This optimization scheme uses the key parameters obtained from real-time parsing as dynamic inputs, combines with predefined static rules, and realizes the structured analysis and matching of multi-dimensional feature tensors by the rule engine. This scheme focuses on the extraction and analysis of core information, simplifies the data processing flow, improves real-time performance and accuracy, and at the same time has good robustness, adaptability and interpretability.

[0031] As a preference of the above embodiment, the lightweight network model performs the following operations: Receives a data stream, extracts local temporal features in the data stream through convolution operations, models temporal dependencies through a gated recurrent unit, and captures long-term temporal dependencies through a temporal convolutional network, and outputs temporal features; compresses the temporal features into a vector of fixed length to obtain pooled features; maps the pooled features to output parameters to generate structured key parameters; Among them, the kernel size and stride of the convolution operation are optimized for the key waveform features in the data stream, and / or when compressing the temporal features, neurons or channels that contribute the most to the prediction of key parameters are preferentially retained.

[0032] In the above optimization scheme, by optimizing the kernel size and stride for key waveform features, the convolution operation can more accurately capture local temporal information in the data stream, ensuring that important details are not missed; compressing the temporal features into a vector of fixed length not only reduces the data dimension but also retains key temporal information, making the subsequent process of mapping output parameters more efficient and stable. Preferentially retaining neurons or channels that contribute the most to the prediction of key parameters during the feature compression process effectively filters redundant information and improves the accuracy and robustness of the model in key metric extraction.

[0033] By using the above lightweight network model, the accurate extraction of local temporal features, the comprehensive modeling of temporal dependencies, the efficient compression of key features, and the retention of important information are realized, which not only ensures real-time performance and low computational burden but also significantly improves the prediction accuracy and robustness of key parameters.

[0034] As a preference of the above embodiment, as Figure 4 shown, the kernel size and stride of the convolution operation are optimized for the key waveform features in the data stream, including: Identifying key waveform features in the data stream through a data visualization tool; Calculating the initial value of the convolution kernel according to the width and sampling rate of the key waveform features; Calculating the initial value of the stride according to the interval and sampling rate of the key waveform features; Using a grid search method to automatically search for the optimal kernel size and stride based on the initial values within a preset parameter range.

[0035] In the above optimization scheme, key waveform features are identified through data visualization, enabling parameter calculation based on the actual data situation, ensuring that the convolutional kernel and stride can accurately cover and slide over the key waveform regions, thereby improving the extraction quality of local temporal features; the grid search method is used to automatically find the optimal parameters within a preset range, improving the adaptability of the system and enabling flexible adjustment under different devices or data patterns, enhancing the robustness of the model.

[0036] After implementation, through a reasonable convolutional kernel size and stride, it is possible to reduce the computational amount while ensuring the accuracy of feature capture, reducing the overall computational burden of the network, which helps to achieve real-time monitoring and rapid response, meeting the real-time requirements of the medical monitoring system.

[0037] As an optimization of the above embodiment, dynamically generate a temporary parsing rule, including: Initialize the basic parsing rule template according to the device type; Modify the field mapping relationship in the basic parsing rule template according to the recognition result; Optimize the parameter calculation logic after modification according to the parsing result, and generate a temporary parsing rule.

[0038] Since the data formats, data structures, and signal characteristics generated by different devices may vary, a more adaptable rule template is preselected according to the device type to ensure a reasonable starting point for subsequent parsing work, providing a customized rule framework for different devices, enabling the parsing work to adapt to diverse data inputs, and reducing possible errors or deviations of general rules when processing data of special devices.

[0039] During implementation, the recognition result is used to modify the field mapping relationship in the basic parsing rule template, adjusting the correspondence between data fields and actual signal characteristics to ensure that key indicators can be accurately extracted during data parsing. Dynamically correcting and optimizing the field mapping relationship can improve the parsing accuracy, making the meaning of each field in the data more consistent with the actual measured value. After modifying the field mapping, based on the latest parsing result, further optimize the parameter calculation logic. Specifically, it may include adjusting the calculation formula, weight distribution, or logical judgment conditions in the algorithm, and finally generating a set of temporary parsing rules suitable for the current data stream. The optimized parameter calculation logic can better adapt to actual data fluctuations, reflect changes in data characteristics in real time, and the generated temporary parsing rules are more flexible and accurate, and can quickly respond to device status changes and data anomalies.

[0040] To ensure better implementation effects, during the establishment of the communication link, automatically identify the device type based on the device unique identification code and / or communication characteristics, and complete authentication and permission management.

[0041] In specific implementation, when the device initiates a connection request, the system first automatically extracts the unique identification code of the device, such as the device ID, MAC address, etc., and / or relevant communication characteristics, such as connection protocol, signal strength, firmware version, etc.; To achieve the above technical objectives, the system can pre-maintain a set of device type databases or rule bases, and automatically identify by comparing the extracted information to ensure that subsequent processing can perform customized operations based on device characteristics.

[0042] After the device type is confirmed, the system will perform identity verification based on the device identification code and / or other authentication credentials, such as digital certificates, preset passwords, tokens, etc. Through this authentication mechanism, only legal and authorized devices are allowed to access the communication link, preventing unauthorized devices from entering the system. After successful authentication, according to the device type and predefined permission policies, the system automatically assigns corresponding access permissions to the device. As a specific implementation method, different device types can be given different permission levels, such as read-only, write, or management permissions, to ensure the secure use of data and resources.

[0043] As an optimization of the above embodiment, the medical monitoring data transmission method further includes calculating a confidence score for each parsing result when executing the temporary parsing rule; when the confidence score is lower than the set threshold, triggering manual confirmation and feedback.

[0044] By triggering the manual confirmation mechanism, it is possible to avoid wrong decisions or false alarms in low-confidence scenarios, effectively reducing the security risks brought by automated parsing errors. When the score is lower than the pre-set threshold, it indicates that there is a large uncertainty in the current automatic parsing or it may deviate from the expected. During the implementation process, once low confidence is detected, the system can automatically pause the current parsing process and present the parsing result, relevant data, and current rule information to the manual operator through the human-machine collaboration interface.

[0045] Through professionals' confirmation, supplementation, or correction of the low-confidence parsing results, and updating the adjusted information into the temporary parsing rule through the system feedback mechanism, this process can be either an immediate intervention for a single parsing result or an improvement of the rule template, thereby improving the accuracy of subsequent parsing.

[0046] Embodiment 2 The medical monitoring method uses the medical monitoring data transmission method described in Embodiment 1 to transmit monitoring data.

[0047] In this embodiment, through the use of the above transmission method, the security and integrity of the data transmission process are ensured, and the real-time performance and the adaptive ability of the system are also significantly improved, thereby providing accurate and timely monitoring data support for clinical decision-making, reducing the risk of misdiagnosis, and optimizing the utilization of medical resources.

[0048] Embodiment 3 The application method of the medical monitoring data transmission method described in Embodiment 1 synchronously transmits the monitoring data of at least two of an electrocardiograph, a ventilator, a blood oxygen monitor, a blood pressure monitor, an electroencephalogram monitor, a blood glucose monitor, a body temperature monitor, an infusion pump, and an anesthetic machine to the same target system.

[0049] Through this embodiment, at the same time point, the monitoring data collected by multiple devices can be integrated and presented in real time. The synchronous transmission enables the data from multiple devices such as electrocardiogram, respiration, blood oxygen, and blood pressure to be gathered in real time, providing an all-round monitoring of the patient's physiological state and forming a global and dynamic patient physiological state system. There may be differences in the design, data format, and communication protocol of various monitoring devices. In the application of the method, through the adaptation and synchronization mechanism of the unified platform, the data collected by different devices is converted into a unified format and kept consistent in time to ensure the comparability and integration of the data.

[0050] The fusion of multi-type data helps clinicians grasp the patient's health status as a whole, timely detect the associated changes between various indicators, improve the accuracy and timeliness of diagnosis. The unified data platform enables different departments or medical institutions to share real-time monitoring data, promotes cross-departmental and cross-institutional collaborative diagnosis and treatment, reduces the risk of misdiagnosis, and improves the overall medical service level.

[0051] The above has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for transmitting medical monitoring data, characterized in that: Standardized conversion is performed during data transmission, including: Adapt the connection methods of different devices through physical interfaces or wireless communication protocols, and establish communication links for data stream transmission; The key parameters in the data stream are analyzed in real time by a lightweight neural network model to obtain analysis results, potential semantic features of the data stream are automatically identified by multimodal signal analysis to obtain recognition results, and temporary analysis rules are dynamically generated based on the analysis results and recognition results to decompose the data stream into independent parameters; The independent parameters are reorganized in a unified standard format and pushed to the target system through a standardized interface.

2. The medical monitoring data transmission method according to claim 1, characterized in that: Automatically identify latent semantic features of data streams through multimodal signal analysis, including: Cutting the data stream into adaptive time windows; Distinguish the data mode of the current time window, extract corresponding features based on the differentiation results, and combine them with the context to obtain a multi-dimensional feature tensor; Based on the multi-dimensional feature tensor, a rule engine is used to identify semantic features of the data stream.

3. The medical monitoring data transmission method according to claim 2, characterized in that: Based on the multidimensional feature tensor, using a rule engine to identify semantic features of the data stream includes: Using the parsing result as dynamic input for semantic reasoning of the rule engine; Using the predefined rules as static input for semantic reasoning of the rule engine; The rule engine performs structured analysis on the data stream based on the multi-dimensional feature tensor to obtain an analysis result, and matches the dynamic input with the static input based on the analysis result to identify the semantic feature.

4. The medical monitoring data transmission method according to any one of claims 1 to 3, characterized in that: The lightweight network model performs the following tasks: Receive the data stream, extract local temporal features in the data stream through a convolution operation, model temporal dependencies through a gated recurrent unit, and capture long temporal dependencies through a temporal convolutional network, and output temporal features; compress the temporal features into a vector of a fixed length to obtain pooling features; map the pooling features to output parameters to generate structured key parameters; The convolution kernel size and step size of the convolution operation are optimized for key waveform features in the data stream, and / or, when compressing the timing features, neurons or channels that contribute most to the prediction of key parameters are preferentially retained.

5. The medical monitoring data transmission method according to claim 4, characterized in that: The convolution kernel size and step size of the convolution operation are optimized for key waveform features in the data stream, including: Identify key waveform features in data streams through data visualization tools; Calculate the initial value of the convolution kernel based on the width and sampling rate of the key waveform features; Calculate the initial value of the step size based on the interval and sampling rate of the key waveform features; A grid search method is used to automatically search for the optimal convolution kernel size and step size based on the initial value within a preset parameter range.

6. The medical monitoring data transmission method according to claim 1, characterized in that: Dynamically generate temporary parsing rules, including: Initialize the basic parsing rule template according to the device type; Modify the field mapping relationship in the basic parsing rule template according to the recognition result; The modified parameter calculation logic is optimized according to the analysis result, and the temporary analysis rule is generated.

7. The medical monitoring data transmission method according to claim 1, characterized in that: During the process of establishing a communication link, the device type is automatically identified based on the device's unique identification code and / or communication characteristics, and authentication and permission management are completed.

8. The medical monitoring data transmission method according to claim 1, characterized in that: Also included is calculating a confidence score for each parsing result when executing the temporary parsing rule; When the confidence score is lower than a set threshold, manual confirmation and feedback are triggered.

9. A medical monitoring method, characterized in that: The monitoring data is transmitted using the medical monitoring data transmission method as described in any one of claims 1 to 8.

10. The application method of the medical monitoring data transmission method according to any one of claims 1 to 8, characterized in that: For at least two of the electrocardiogram monitor, ventilator, blood oxygen monitor, blood pressure monitor, electroencephalogram monitor, blood glucose monitor, temperature monitor, infusion pump, and anesthesia machine, monitoring data are synchronously transmitted to the same target system.

Images