Method for automatic conversion and standardization of heterogeneous data formats for monitors / anaesthesia machines

Abstract

The application discloses a heterogeneous data format automatic conversion and standardization processing method for monitors / anaesthetic machines, and relates to the technical field of medical data processing.The method comprises the following steps: S1, constructing a hierarchical surgery whole-process scene label system based on multi-center time sequence data and clinical value dynamic calibration correlation strength, combing the core clinical stages of surgery and setting exclusive scene labels, associating corresponding clinical operations, core monitoring parameters and data precision requirements for each label, and establishing a stage correlation mapping relationship between labels to support cross-stage dynamic flow; the core technology combination of the scene label system construction-real-time scene recognition-dynamic priority adjustment-unified standardization conversion solves the pain points of the insufficient adaptation of existing general conversion technology, and the hierarchical scene label system constructed by the application covers the induction period, maintenance period and recovery period of anesthesia, and can be associated with clinical operations, core parameters and precision requirements.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, specifically to a method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines. Background Technology

[0002] Patient monitors are core devices for real-time monitoring of patients' vital signs during clinical surgery, providing medical staff with real-time feedback on the patient's physiological status. Anesthesia machines are crucial devices for precise drug administration and maintaining the patient's respiratory function during surgery; their operating parameters are directly related to the anesthetic effect and patient safety. As indispensable medical devices in the surgical setting, the data generated by both are important bases for medical staff to make clinical decisions, predict risks, and evaluate surgical outcomes. They also serve as the core foundation for medical data traceability, clinical research, and multi-center collaborative research, playing a vital role in ensuring surgical safety and improving the quality of medical care.

[0003] With numerous manufacturers and diverse models of patient monitors and anesthesia machines, significant differences exist in the parameter definition rules, data storage formats, and data transmission protocols used by different manufacturers during equipment development. This results in heterogeneous data output from different devices, a phenomenon known as heterogeneous data from patient monitors and anesthesia machines. As medical informatization advances, the integration and utilization of heterogeneous data has become an industry necessity. Automatic conversion and standardization of heterogeneous data formats are crucial for breaking down data barriers between devices and achieving data interoperability and sharing. This not only ensures the real-time and effective use of data in clinical settings but also provides a unified data foundation for long-term medical data storage, cross-institutional collaboration, and research data analysis.

[0004] However, existing heterogeneous data conversion technologies are mostly general-purpose solutions that focus only on data format standardization and do not take into account the clinical context of the entire surgical process. The surgical process includes multiple core stages such as anesthesia induction, maintenance, and recovery. The clinical operation priorities and monitoring data requirements of different stages are significantly different. Existing technologies do not consider the impact of these stage differences on data parsing priority and accuracy requirements, resulting in standardized data that cannot accurately match actual clinical needs. In some critical stages, core monitoring data may not be parsed in a timely manner, while non-critical data may be redundant in routine stages. This not only affects the efficiency of clinical operations but may also lead to decision-making biases by medical staff due to insufficient data adaptation, making it difficult to meet the dynamically changing clinical data needs throughout the surgical process. Therefore, developing an automatic conversion and standardization method for heterogeneous data formats for monitors / anesthesia machines is of great significance. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an automatic conversion and standardization method for heterogeneous data formats for monitors / anesthesia machines. It addresses the pain points of insufficient adaptability of existing general conversion technologies through a core technology combination of scene labeling system construction, real-time scene recognition, dynamic priority adjustment, and unified standardized conversion. The invention constructs a hierarchical scene labeling system covering the anesthesia induction, maintenance, and recovery phases, linking clinical operations, core parameters, and accuracy requirements. Combined with real-time feature extraction and scene matching algorithms, it achieves precise binding of data processing with clinical stages. Multi-protocol compatible acquisition interfaces break down device barriers, a unified standardized data model ensures interoperability, emergency scene optimization and custom configuration enhance adaptability flexibility, and dual verification and traceability mechanisms ensure data reliability.

[0006] To solve the above-mentioned technical problems, this invention provides the following technical solution: a method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines, the method comprising the following steps: S1. Construct a hierarchical surgical process scenario labeling system based on the correlation strength of multi-center time-series data and dynamic calibration of clinical value. Identify the core clinical stages of surgery and set exclusive scenario labels. Associate each label with corresponding clinical operations, core monitoring parameters and data accuracy requirements. Establish a stage association mapping relationship between labels that supports dynamic flow across stages. S2. Through a multi-protocol compatible acquisition mechanism that integrates dynamic protocol identification and adaptation algorithms, heterogeneous raw data from monitors and anesthesia machines are collected in real time. Equipment operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend mutation characteristics are extracted. Combined with the time-stamped structured clinical operation execution information of the surgical anesthesia record system, the current surgical clinical stage is determined and the corresponding scene label is matched through a scene matching algorithm that uses multi-dimensional feature weighted fusion. S3. Based on the matched scene tags, call the preset parsing rules, dynamically adjust the parsing priority of heterogeneous data based on the clinical importance of parameters, scene relevance and real-time computing resource utilization, and establish a priority dynamic feedback adjustment mechanism that includes parsing delay feedback and resource scheduling linkage. S4. According to the accuracy requirements corresponding to the scene label, the parsed data is converted into a unified standardized data model containing cross-organization data interoperability identifiers and scene semantic metadata based on the preset field association rules that support heterogeneous parameter semantic mapping, so as to achieve accurate mapping between heterogeneous parameters and standard model fields. S5. Based on clinical needs, standardized data is output through a multi-type adaptive interface that supports real-time push and batch export. Before output, data is verified using dual rules of field integrity verification and data logic consistency verification. A data anomaly traceability link is established that is deeply bound to scene tags and parsing priority records to complete the dynamic adaptation of heterogeneous data to clinical scenarios.

[0007] Furthermore, step S1, in constructing the surgical process scenario labeling system, includes the following steps: By reviewing the clinical surgical anesthesia practice process, the core clinical stages in the entire surgical process are identified, which include the anesthesia induction period, maintenance period, and recovery period. Each core clinical phase is assigned a unique scenario label. The scenario label adopts a character encoding format that includes a phase type identifier, a priority prefix, and a multi-center adaptation identifier. Different phase labels are not repeated. For each scenario label, frequently occurring clinical procedures in that phase of clinical practice are collected, and the core monitoring parameters corresponding to each procedure are identified. The correlation strength between scenario labels, clinical procedures, and core parameters is dynamically calibrated using a formula that integrates multi-center time-series decay characteristics and data dispersion correction. The formula is as follows: ,in, This represents the dynamic correlation strength between the current scene label and the corresponding clinical operation / core parameter. The initial baseline value for the strength of association was determined based on consensus among clinical experts. The dynamic time decay factor, based on the statistical distribution of surgical duration at multiple centers, was determined through time-series analysis of multi-center surgical anesthesia data, combined with statistical analysis of the average duration of each clinical stage. The scenario stability coefficient is calculated by combining the variance of parameter changes and trend consistency over five consecutive data collection periods. This represents the number of associated combinations of clinical procedures and core parameters under this scenario label. For the first The actual frequency of group-related combinations in multicenter clinical data over the past three years. For the first The clinical value coefficient of the group association combination was determined by a team of clinical anesthesiology experts using a three-round assessment based on the operational risk level and the importance of parameter warnings. For the first The dispersion of the data corresponding to the group association combination was calculated based on the historical monitoring data of nearly 100,000 similar surgeries. Based on clinical treatment guidelines, the data accuracy requirements of core monitoring parameters at this stage are determined, and a mapping table is formed that supports the version synchronization of clinical treatment guidelines and the overlay of multi-center custom rules, which is associated with clinical operations, core parameters, and accuracy requirements.

[0008] Furthermore, step S2 includes the following steps when performing heterogeneous data acquisition and scene recognition: The heterogeneous raw data output from different models of monitors and anesthesia machines is acquired in real time through a universal data acquisition interface compatible with devices from multiple manufacturers. The heterogeneous raw data includes physiological parameter data monitored by the equipment, equipment operating status data, and records of parameter trend changes. Extract equipment operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend mutation characteristics from the collected data, and connect with the surgical anesthesia record system to obtain structured clinical operation execution information with timestamps; The extracted features and clinical operation execution information are compared with the contents of the scene label association mapping table, using a formula. Calculate the current scene matching degree, where The number of clinical procedures associated with this scenario label. For the first The weight of each clinical procedure For current clinical procedure execution information and the first The degree of matching of related clinical procedures. This represents the number of parameter features associated with the scene label. For the first The weights of the feature parameters, For the currently extracted parameter features and the first Similarity of features of associated parameters, and Based on the consensus of clinical experts, the Delphi method was used to determine the clinical stage of each operation and parameter in the corresponding scenario, and the clinical relevance was determined by combining the frequency of occurrence and clinical relevance of each clinical operation and parameter. The current clinical stage was determined by the calculated scenario matching degree and the dynamic preset matching threshold for adaptive scenario switching, and then the corresponding scenario label was matched.

[0009] Furthermore, step S3 includes the following steps when performing dynamic adaptation and parsing of rules: Based on the list of core monitoring parameters corresponding to the scene labels, using the formula Calculate the priority coefficients for each parameter, where For the first Priority coefficients of each parameter, For the first Clinical dependence coefficient of each parameter, For the first Clinical dependence of each parameter For the first The scene correlation coefficient of each parameter, For the first The degree of correlation between each parameter and the current scene label. and The importance of parameters was determined through multi-center clinical data statistical analysis, combined with the supporting role of parameters in clinical operation under different surgical scenarios and the data processing resource consumption, and classified into three levels of importance: high, medium and low according to priority coefficients. Set the parsing priority sequence according to the importance level, and assign the parsing tasks corresponding to the core monitoring parameters to the high priority queue, and non-core parameters to the low priority queue. During the parsing process, parsing tasks are executed in order of priority queues. Parsing tasks in the high-priority queue occupy no less than 70% of the computing resources, and the resource allocation ratio is adjusted in real time based on parsing delay feedback. Parsing tasks in the low-priority queue are executed when computing resources are idle, and a dynamic priority adjustment feedback link is established, which includes dynamic updates of priority coefficients and optimization of queue scheduling.

[0010] Furthermore, the general data acquisition interface in step S2 supports multiple data transmission protocols, including HL7 protocol, DICOM protocol and custom transmission protocols of various device manufacturers. During the data acquisition process, the dynamic protocol identification and adaptation algorithm of the interface first verifies the device protocol type and then automatically adapts and is compatible to complete the compatible acquisition of heterogeneous raw data. The acquired data is stored in real time to a temporary data buffer. The buffer adopts a circular overwrite storage mechanism, and the buffer capacity is dynamically configured according to the device data output rate and parameter importance level.

[0011] Furthermore, the unified standardized data model in step S4 is constructed based on medical data-related standards and industry-standard data specifications. The data model includes basic fields such as parameter name, data type, unit, collection timestamp, and device identifier. It also reserves three types of extended fields: cross-institutional collaboration identifier field, scenario semantic metadata field, and parameter priority identifier field. During the format conversion process, a field mapping table that supports both semantic fuzzy matching and precise mapping modes maps the heterogeneous parameter names and data structures of different devices to the corresponding fields of the standardized data model. The field mapping table is version-managed and supports manual updates and multi-center rule synchronization.

[0012] Furthermore, in step S4, the storage optimization in the conventional monitoring scenario is achieved through a data compression algorithm. The compression algorithm selects the corresponding compression method according to the data type and parameter priority level. For numerical data, the LZ77 lossless compression algorithm is used, and for text data, the static dictionary encoding compression algorithm is used. The compressed standardized data is stored in the medical data center. During the storage process, a four-dimensional data index of scene tag - device identifier - collection timestamp - priority level is established. The index is automatically updated every 24 hours and supports incremental update optimization.

[0013] Furthermore, the criteria for determining an emergency intervention scenario in step S4 are: the core monitoring parameters exceed the clinical safety threshold set based on clinical treatment guidelines, a clinical emergency operation trigger signal is received transmitted through a dedicated signal channel, or the parameter trend change characteristics meet the emergency warning conditions. When an emergency intervention scenario is determined, the data processing thread scheduling mechanism is optimized to raise the priority of the data processing thread to the highest level, suspend non-core data processing processes, reduce unnecessary data preprocessing steps, retain only the core parsing and format conversion process, and push emergency data to the clinical warning terminal in real time.

[0014] Furthermore, the core monitoring parameters and data accuracy requirements associated with the scene tags in step S1 can be customized by medical staff according to the specific surgical type, patient condition characteristics, and multi-center collaboration needs. The configuration process is achieved through a visual interactive interface that includes four steps: parameter selection, accuracy setting, priority adjustment, and confirmation submission. The configuration information is stored in a distributed multi-center synchronous configuration database. When calling the parsing rules in step S3, the customized configuration information in the configuration database is read first, and then the data parsing priority and accuracy adaptation strategy are adjusted in combination with the preset parsing rules.

[0015] Furthermore, the data verification and backtracking steps in step S5 are specifically as follows: standardized data is verified using dual rules of field integrity verification and data logic consistency verification; a four-level data anomaly tracing association link is established, including scene matching records, priority adjustment logs, parsing process details, and original data fragments; and standardized data is associated and stored with original collected data, scene tag matching records, parsing process logs, and resource scheduling records to ensure that data anomalies can be accurately traced back to the data source, processing stage, and resource allocation.

[0016] Compared with existing technologies, this method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines has the following advantages: This invention addresses the pain points of insufficient adaptability of existing general conversion technologies by combining core technologies such as scene tagging system construction, real-time scene recognition, dynamic priority adjustment, and unified standardized conversion. It constructs a hierarchical scene tagging system covering the anesthesia induction, maintenance, and recovery phases, linking clinical operations, core parameters, and accuracy requirements. Combined with real-time feature extraction and scene matching algorithms, it achieves precise binding of data processing with clinical stages. Through dynamic priority formulas, computational resources are allocated, with core parameter parsing prioritizing over 70% of resources, resolving issues of core data latency and non-critical data redundancy. Simultaneously, multi-protocol compatible acquisition interfaces break down device barriers, a unified standardized data model ensures interoperability, emergency scene optimization and custom configuration enhance adaptability flexibility, and dual verification and traceability mechanisms ensure data reliability.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 A flowchart illustrating the automatic conversion and standardization process for heterogeneous data formats used in patient monitors / anesthesia machines; Figure 2 A flowchart illustrating the automatic conversion and standardization process for heterogeneous data formats used in patient monitors / anesthesia machines; Figure 3 This is a flowchart for heterogeneous data acquisition and scene recognition. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, will describe the specific implementation methods, structure, features, and effects of the present invention.

[0021] The invention patent provides an automatic conversion and standardization method for heterogeneous data formats for monitors / anesthesia machines. This method addresses the problem that the heterogeneous data formats generated by monitors and anesthesia machines from different manufacturers and models are not uniform, and that existing general conversion technologies are not adapted to the differences in various clinical stages of surgery. It provides a complete solution for automatic conversion and standardization of heterogeneous data, with the core technology system built around "scene binding - precise acquisition - dynamic analysis - standard conversion - reliable output".

[0022] See Figure 1 and Figure 2 The core of the technical solution of this invention consists of the following steps: The solution first constructs a hierarchical surgical procedure scenario labeling system, identifying three core clinical stages: anesthesia induction, maintenance, and recovery. Each stage is assigned a unique character-coded scenario label. Each label is associated with high-frequency clinical procedures, core monitoring parameters, and data accuracy requirements for that stage, forming a mapping table that can be updated with clinical treatment guidelines. Simultaneously, it allows medical staff to customize core parameters and accuracy requirements through a visual interface, based on the surgical type and patient condition.

[0023] The data acquisition process employs a universal interface compatible with multiple protocols, including HL7, DICOM, and vendor-defined protocols. It collects heterogeneous raw data in real time, such as physiological parameters and device operating status, and temporarily stores the collected data in a dynamically adaptable, cyclically overwritten buffer. By extracting device operation records and parameter change characteristics, it interfaces with the clinical operation information of the surgical anesthesia recording system and compares it with an associated mapping table to match the current clinical stage and corresponding scenario label.

[0024] During the parsing process, the data parsing priority is dynamically adjusted based on the scenario labels. The data is divided into high, medium, and low queues according to the clinical importance and scenario relevance of the parameters. Core parameter parsing tasks occupy no less than 70% of the computing resources, while non-core parameters are executed when resources are idle. A dynamic feedback adjustment mechanism is established.

[0025] The format conversion stage employs a unified, standardized data model, including basic fields such as parameter names and data types, as well as extended fields. A version-managed field mapping table ensures accurate mapping between heterogeneous parameters and standard fields. In normal scenarios, storage is optimized and indexed based on data type using corresponding compression algorithms. In emergency intervention scenarios, the processing thread priority is increased, unnecessary steps are streamlined, and only core processes are retained.

[0026] Ultimately, standardized data is output through multi-type adaptation interfaces. Before output, the data undergoes dual verification of field integrity and logical consistency. At the same time, a data anomaly tracing link is established, linking the original data, scenario matching records, and parsing logs to ensure data reliability. This enables dynamic adaptation of heterogeneous data to clinical scenarios, improving the efficiency of clinical decision-making and the value of data utilization.

[0027] Example 1 This embodiment is applied to a routine surgical anesthesia treatment scenario in the operating room of a tertiary hospital. This operating room is equipped with multiple monitors and anesthesia machines from different manufacturers and of various models. During surgery, these devices continuously generate crucial information such as patient physiological parameter data and their own operational status data. Due to significant differences in parameter definition rules, data storage formats, and data transmission protocols among different devices, the output data exhibits significant heterogeneity. The entire surgical process includes core stages such as anesthesia induction, maintenance, and recovery, each with different clinical operational focuses and data requirements. Existing general-purpose data conversion technologies cannot adapt to these stage differences, often resulting in problems such as untimely parsing of core data and redundancy of non-critical data, affecting the efficiency of clinical decision-making by medical staff and surgical safety. Therefore, this invention employs an automatic conversion and standardization method for heterogeneous data formats from monitors / anesthesia machines to achieve accurate conversion and standardization of heterogeneous data.

[0028] See Figure 1 and Figure 2The specific implementation process of this embodiment is as follows: A comprehensive surgical procedure scenario labeling system was constructed. First, by reviewing the hospital's clinical surgical anesthesia practices, the core clinical phases of the entire surgical procedure were identified: induction, maintenance, and recovery. A unique scenario label was assigned to each core clinical phase. These labels used a character-coded format including a phase type identifier, a priority prefix, and a multi-center adaptation identifier to ensure that labels for different phases were not duplicated. For each scenario label, frequently occurring clinical procedures within that phase were collected, and the corresponding core monitoring parameters for each procedure were identified and implemented using formulas. Dynamically calibrate the correlation strength. Among them, The initial baseline value for the strength of association was determined based on consensus among clinical experts. The time decay factor was determined through time-series analysis of multicenter surgical anesthesia data combined with statistical analysis of the average duration of each clinical stage. The scene stability coefficient is calculated by the variance of parameter changes over five consecutive data acquisition periods. This represents the number of associated combinations of clinical procedures and core parameters under this scenario label. For the first The actual frequency of group-related combinations in multicenter clinical data over the past three years. For the first The clinical value coefficient of the group association combination was determined by a team of clinical anesthesiology experts using a three-round assessment based on the operational risk level and the importance of parameter warnings. For the first The dispersion of the data corresponding to the group association combination was calculated based on the historical monitoring data of nearly 100,000 similar surgeries.

[0029] By combining clinical treatment guidelines to determine the data accuracy requirements of core monitoring parameters at each stage, a mapping table is formed that supports the synchronization of clinical treatment guidelines versions and the overlay of multi-center customized rules, linking scenario tags with clinical operations, core parameters, and accuracy requirements. This mapping table can be updated and adjusted according to the hospital's clinical needs. Medical staff can also customize and configure core monitoring parameters and data accuracy requirements based on specific surgical types, patient condition characteristics, and multi-center collaboration needs through a visual interactive interface that includes four steps: parameter selection, accuracy setting, priority adjustment, and confirmation submission. The configuration information is stored in a distributed multi-center synchronized configuration database.

[0030] Heterogeneous data acquisition and scene recognition are performed. A universal data acquisition interface compatible with multiple manufacturers' devices is used to acquire heterogeneous raw data from different models of monitors and anesthesia machines in the operating room in real time. This interface supports HL7, DICOM, and custom transmission protocols from various device manufacturers. A dynamic protocol identification and adaptation algorithm is used to first verify the device protocol type and then automatically adapt for compatibility, completing the compatible acquisition of heterogeneous raw data. The acquired data is stored in real time in a circular overlay temporary buffer (the buffer capacity is dynamically configured based on the device data output rate and parameter importance level). After acquisition, device operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend abrupt changes are extracted. This data is then integrated with the surgical anesthesia recording system to obtain structured clinical operation execution information with timestamps.

[0031] The extracted features and clinical operation execution information are compared with the association mapping table. In the specific implementation of this embodiment, the formula is used... Calculate the current scene matching degree. The number of clinical procedures associated with this scenario label. For the first The weight of each clinical procedure For current clinical procedure execution information and the first The degree of matching of related clinical procedures. This represents the number of parameter features associated with the scene label. For the first The weights of the feature parameters, For the currently extracted parameter features and the first Similarity of features of associated parameters. and Based on clinical expert consensus, the Delphi method was used to determine the frequency and clinical relevance of various clinical operations and parameter characteristics in corresponding scenarios. When the calculated matching degree reaches a preset threshold, the current clinical stage is determined and the corresponding scenario label is matched.

[0032] Dynamically adapt parsing rules. When calling the parsing rules in step S3, the custom configuration information in the configuration database is read first, and then the parsing priority and accuracy adaptation strategy are adjusted in conjunction with the preset parsing rules. Based on the list of core monitoring parameters corresponding to the matched scene tags, in the specific implementation of this embodiment, the formula is used... Calculate the priority coefficients for each parameter. For the first Priority coefficients of each parameter For the first Clinical dependence coefficient of each parameter, For the first Clinical dependence of each parameter For the first The scene correlation coefficient of each parameter, For the first The degree of correlation between each parameter and the current scene label. and The parameters were determined through statistical analysis of multi-center clinical data, taking into account their supporting role in clinical operations and the resource consumption of data processing under different surgical scenarios.

[0033] The importance of monitoring parameters is categorized into high, medium, and low levels based on priority coefficients. Tasks corresponding to core monitoring parameters are assigned to high-priority queues, while non-core parameters are assigned to low-priority queues. During the parsing process, tasks are executed in the order of the priority queues. High-priority queue tasks consume no less than 70% of the computing resources, and the resource allocation ratio is adjusted in real-time based on parsing delay feedback. Low-priority queue tasks are executed when computing resources are idle. This establishes a dynamic priority adjustment feedback loop that includes dynamic updates of priority coefficients and queue scheduling optimization.

[0034] Perform format conversion and standardization. Based on the precision requirements corresponding to the matched scene tags, the parsed data is converted into a unified standardized data model. This model includes basic fields such as parameter name, data type, unit, collection timestamp, and device identifier. It also reserves three types of extended fields: cross-organizational collaboration identifier, scene semantic metadata field, and parameter priority identifier field. During the format conversion process, a field mapping table supporting both semantic fuzzy matching and precise mapping modes uniformly maps heterogeneous parameter names and data structures from different devices to the corresponding fields in the standardized data model. The field mapping table is version-managed and supports manual updates.

[0035] In routine monitoring scenarios, numerical data uses the LZ77 lossless compression algorithm, while text data uses a static dictionary encoding compression algorithm. The compressed, standardized data is stored in a medical data center. During storage, a four-dimensional data index is established, consisting of scenario tags, device identifiers, collection timestamps, and priority levels. This index is automatically updated every 24 hours and supports incremental updates for optimization. If core monitoring parameters exceed clinical safety thresholds set based on clinical treatment guidelines, a clinical emergency operation trigger signal is received via a dedicated signal channel, or a sudden change in parameter trends meets emergency warning conditions, an emergency intervention scenario is identified. In this case, the data processing thread scheduling mechanism is optimized, prioritizing the data processing threads to the highest level, suspending non-core data processing processes, reducing unnecessary data preprocessing steps, retaining only core parsing and format conversion processes, and pushing emergency data to the clinical warning terminal in real time.

[0036] Standardized data is output and, based on clinical needs, utilizes adaptive multi-type interfaces supporting real-time push and batch export to deliver standardized data to healthcare workers' work terminals, hospital information systems, and medical data centers. Before output, standardized data undergoes dual verification using field integrity checks and data logic consistency checks. Simultaneously, a four-level data anomaly tracing link is established, including scenario matching records, priority adjustment logs, parsing process details, and original data fragments. Standardized data is associated and stored with original collected data, scenario tag matching records, parsing process logs, and resource scheduling records to ensure accurate tracing of data anomalies back to the data source, processing stage, and resource allocation.

[0037] After a three-month clinical trial at the hospital, the application of this method reduced the core parameter parsing delay from 2.5 seconds to 0.4 seconds, reduced non-critical data redundancy by 65%, achieved 100% equipment compatibility, and improved clinical decision response speed by 35%, effectively avoiding decision-making biases caused by data compatibility issues.

[0038] In summary, this embodiment, by fully applying the heterogeneous data format automatic conversion and standardization processing method of the present invention, achieves deep binding between heterogeneous data conversion rules, standardized parameters, and clinical scenario semantics. At different core stages of the entire surgical process, core monitoring data can be accurately and timely parsed, effectively reducing non-critical data redundancy, improving clinical operation and decision-making efficiency, and providing high-quality standardized data support for hospital medical data traceability, clinical research, and multi-center collaborative research.

[0039] Example 2 This embodiment is applied to a multi-center clinical anesthesia research collaboration scenario. The participating medical institutions are equipped with monitors and anesthesia machines of different brands and models, generating a large amount of heterogeneous data during surgical anesthesia procedures. Due to inconsistencies in equipment parameter definitions, data formats, and transmission protocols across institutions, and the need for joint analysis of multi-center, end-to-end surgical data, existing general conversion technologies cannot meet the requirements for standardized data integration across institutions. Furthermore, while surgical procedures share commonalities, subtle differences exist, and the emphasis on core monitoring parameters varies slightly, resulting in low data integration efficiency and inconsistent research data quality. Therefore, this invention employs an automatic conversion and standardization method for heterogeneous data formats from monitors / anesthesia machines, optimizing the implementation process to achieve efficient integration and standardization of heterogeneous data across institutions.

[0040] See Figure 1 and Figure 2 The specific implementation process of this embodiment is as follows: A multi-center adapted surgical workflow scenario labeling system was constructed. Clinical experts and researchers from multiple centers jointly identified common processes, defining anesthesia induction, maintenance, and recovery phases as unified core clinical stages. A general basic scenario labeling system was developed, with scenario labels using a unified character encoding format including stage type identifiers, priority prefixes, and multi-center adaptation identifiers to ensure consistent identification across institutions. To address differences between institutions, a custom configuration function was retained, allowing medical staff to add extended labels or adjust core parameters and precision requirements through a visual interactive interface that includes parameter selection, precision setting, priority adjustment, and confirmation submission. Configuration information is stored in a distributed, multi-center synchronized configuration database. During the construction process, formulas were used... The correlation strength is dynamically calibrated to form a unified, institution-adapted correlation mapping table that supports the synchronization of clinical treatment guidelines and the overlay of multi-center custom rules. The definitions of each parameter are consistent with those in Example 1, ensuring the uniformity of multi-center data correlation logic.

[0041] Multi-center heterogeneous data acquisition and scene recognition were conducted. Distributed data acquisition nodes were uniformly deployed across multiple centers, each equipped with a universal data acquisition interface. This interface supports HL7, DICOM, and custom transmission protocols from various equipment manufacturers. A dynamic protocol identification and adaptation algorithm was used to first verify the device protocol type and then automatically adapt for compatibility, enabling real-time acquisition of heterogeneous raw data across hospitals. The data was stored in a distributed, circular, overlay-style temporary buffer (the buffer capacity was dynamically configured based on the data output rate and parameter importance level of each institution's equipment). Equipment operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend abrupt changes were extracted. This data was then integrated with the surgical anesthesia record systems of various institutions to obtain time-stamped structured clinical operation execution information.

[0042] See Figure 3 The extracted features and clinical operation execution information are compared with a multi-center unified scene label association mapping table. In the specific implementation of this embodiment, the formula is used... Calculate the current scenario matching degree, determine the current clinical stage based on the matching degree, and match the corresponding scenario label to ensure the consistency of scenario identification in data from various centers.

[0043] Dynamically adapt to multi-center parsing rules. Based on the parsing priority adjustment logic of the aforementioned embodiments, a multi-center data processing resource pool scheduling mechanism is added. When calling the parsing rules in step S3, the custom configuration information of each center in the distributed configuration database is first read, and then combined with the multi-center general preset parsing rules, using formulas... Calculate the priority coefficient of each parameter, determine the importance level of the parameters, and divide the parsing priority queue.

[0044] Meanwhile, based on the real-time data processing resource occupancy rate of each center, computing resources are dynamically allocated by the resource pool to prioritize the parallel processing of high-priority parsing tasks across centers (high-priority queue parsing tasks occupy no less than 70% of computing resources). For parsing tasks of non-core parameters, an idle-time distributed computing mode is adopted to balance the processing pressure of each center, avoid parsing delays caused by resource overload of a single center, and establish a dynamic priority adjustment feedback link including dynamic updates of priority coefficients and optimization of queue scheduling.

[0045] Performs format conversion and standardization processing for multi-center adaptation. Based on a unified standardized data model, it includes basic fields such as parameter name, data type, unit, collection timestamp, and device identifier. It also reserves three types of extended fields: cross-institutional collaboration identifier, scenario semantic metadata field, and parameter priority identifier field, to meet the needs of multi-center collaborative data traceability and interoperability. Through a field mapping table jointly developed by multiple centers, supporting both semantic fuzzy matching and precise mapping modes, heterogeneous parameters are uniformly mapped to standard fields. The field mapping table is version-managed and supports manual updates and multi-center rule synchronization.

[0046] In standard scenarios, a distributed storage plus local backup strategy is adopted. Storage is optimized using either the LZ77 lossless compression algorithm (for numerical data) or the static dictionary encoding compression algorithm (for text data) based on data type. A four-dimensional data index—scenario tag, device identifier, collection timestamp, and priority level—is established, automatically updated every 24 hours and supporting incremental updates. If core monitoring parameters exceed clinical safety thresholds set based on clinical treatment guidelines, a clinical emergency operation trigger signal is received via a dedicated signal channel, or the parameter trend abruptly changes to meet emergency warning conditions, an emergency intervention scenario is identified. In this case, the data processing thread is prioritized to the highest level, non-core data processing processes are suspended, unnecessary preprocessing steps are streamlined, and only the core parsing and format conversion processes are retained. Simultaneously, cross-center warning information is triggered, and emergency data is pushed to the relevant institution's clinical warning terminal.

[0047] Output standardized data from multiple centers. Based on the needs of clinical diagnosis and treatment and scientific research collaboration, standardized data is output to local clinical terminals of various institutions, multi-center scientific research collaboration platforms, and cross-institutional data sharing centers through adaptive multi-type interfaces that support real-time push and batch export, establishing a hierarchical data access control mechanism. Before output, standardized data is verified using dual rules of field integrity check and data logic consistency check. Simultaneously, a four-level data anomaly tracing and association link is established, including scenario matching records, priority adjustment logs, parsing process details, and original data fragments. Standardized data is associated and stored with original collected data, scenario tag matching records, parsing process logs, and resource scheduling records to ensure that in the event of data anomalies, accurate cross-institutional tracing back to the data source, processing stage, and resource allocation is possible.

[0048] After joint testing and verification by six collaborating hospitals, the method achieved 100% cross-institutional equipment compatibility, improved data integration efficiency by 50%, achieved 99.8% accuracy in standardized data, reduced data transmission latency in emergency scenarios to 0.08 seconds, and shortened the multi-center research data integration cycle from 15 days to 3 days, effectively supporting cross-institutional clinical decision-making and research collaboration.

[0049] In summary, this embodiment, based on the aforementioned embodiments, optimizes the implementation process for multi-center collaborative scenarios, achieving deep adaptation of heterogeneous data conversion rules to multi-center clinical scenarios and research needs. By combining a unified basic labeling system with institutional custom configurations, it balances the uniformity and specificity of cross-institutional data; the distributed acquisition and resource pool scheduling mechanism improves the efficiency and stability of multi-center data processing; and the expansion of standardized data models and cross-center storage and output strategies break down data barriers between institutions. This implementation process ensures accurate analysis and timely sharing of core monitoring data from multiple centers, providing a high-quality, unified-format data source for multi-center collaborative research, and significantly improving the efficiency and quality of multi-center collaboration.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines, characterized in that, The method includes the following steps: S1. Construct a hierarchical surgical process scenario labeling system based on the correlation strength of multi-center time-series data and dynamic calibration of clinical value. Identify the core clinical stages of surgery and set exclusive scenario labels. Associate each label with corresponding clinical operations, core monitoring parameters and data accuracy requirements. Establish a stage association mapping relationship between labels that supports dynamic flow across stages. S2. Through a multi-protocol compatible acquisition mechanism that integrates dynamic protocol identification and adaptation algorithms, heterogeneous raw data from monitors and anesthesia machines are collected in real time. Equipment operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend mutation characteristics are extracted. Combined with the time-stamped structured clinical operation execution information of the surgical anesthesia record system, the current surgical clinical stage is determined and the corresponding scene label is matched through a scene matching algorithm that uses multi-dimensional feature weighted fusion. S3. Based on the matched scene tags, call the preset parsing rules, dynamically adjust the parsing priority of heterogeneous data based on the clinical importance of parameters, scene relevance and real-time computing resource utilization, and establish a priority dynamic feedback adjustment mechanism that includes parsing delay feedback and resource scheduling linkage. S4. According to the accuracy requirements corresponding to the scene label, the parsed data is converted into a unified standardized data model containing cross-organization data interoperability identifiers and scene semantic metadata based on the preset field association rules that support heterogeneous parameter semantic mapping, so as to achieve accurate mapping between heterogeneous parameters and standard model fields. S5. Based on clinical needs, standardized data is output through a multi-type adaptive interface that supports real-time push and batch export. Before output, data is verified using dual rules of field integrity verification and data logic consistency verification. A data anomaly traceability link is established that is deeply bound to scene tags and parsing priority records to complete the dynamic adaptation of heterogeneous data to clinical scenarios.

2. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, Step S1, in constructing the surgical process scenario labeling system, includes the following steps: By reviewing the clinical surgical anesthesia practice process, the core clinical stages in the entire surgical process are identified, which include the anesthesia induction period, maintenance period, and recovery period. Each core clinical phase is assigned a unique scenario label. The scenario label adopts a character encoding format that includes a phase type identifier, a priority prefix, and a multi-center adaptation identifier. Different phase labels are not repeated. For each scenario label, frequently occurring clinical procedures in that phase of clinical practice are collected, and the core monitoring parameters corresponding to each procedure are identified. The correlation strength between scenario labels, clinical procedures, and core parameters is dynamically calibrated using a formula that integrates multi-center time-series decay characteristics and data dispersion correction. The formula is as follows: ,in, This represents the dynamic correlation strength between the current scene label and the corresponding clinical operation / core parameter. The initial baseline value for the strength of association was determined based on consensus among clinical experts. This is a dynamic time decay factor based on the statistical distribution of surgical phase durations in multiple centers. The scenario stability coefficient is calculated by combining the variance of parameter changes and trend consistency over five consecutive data collection periods. This represents the number of associated combinations of clinical procedures and core parameters under this scenario label. For the first The actual frequency of group-related combinations in multicenter clinical data over the past three years. For the first Clinical value coefficient of group-related combinations This is the first statistical analysis based on stratified sampling data from nearly 100,000 similar surgeries. The dispersion of data corresponding to group association combinations; Based on clinical treatment guidelines, the data accuracy requirements of core monitoring parameters at this stage are determined, and a mapping table is formed that supports the version synchronization of clinical treatment guidelines and the overlay of multi-center custom rules, which is associated with clinical operations, core parameters, and accuracy requirements.

3. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, Step S2, when performing heterogeneous data acquisition and scene recognition, includes the following steps: The heterogeneous raw data output from different models of monitors and anesthesia machines is acquired in real time through a universal data acquisition interface compatible with devices from multiple manufacturers. The heterogeneous raw data includes physiological parameter data monitored by the equipment, equipment operating status data, and records of parameter trend changes. Extract equipment operation records, parameter change amplitude, change frequency, data transmission stability, and parameter trend mutation characteristics from the collected data, and connect with the surgical anesthesia record system to obtain structured clinical operation execution information with timestamps; The extracted features and clinical operation execution information are compared with the contents of the scene label association mapping table, using a formula. Calculate the current scene matching degree, where The number of clinical procedures associated with this scenario label. The first step is to optimize the accuracy of multi-center scene recognition by reverse engineering. The weight of each clinical procedure For current clinical procedure execution information and the first The degree of matching of related clinical procedures. This represents the number of parameter features associated with the scene label. The first step is to optimize the accuracy of multi-center scene recognition by reverse engineering. The weights of the feature parameters, For the currently extracted parameter features and the first The similarity of the associated parameter features is used to determine the current clinical stage based on the calculated scene matching degree and the dynamically preset matching threshold for adaptive scene switching, and then the corresponding scene label is matched.

4. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, Step S3, when performing dynamic adaptation and parsing rules, includes the following steps: Based on the list of core monitoring parameters corresponding to the scene labels, using the formula Calculate the priority coefficients for each parameter, where For the first Priority coefficients of each parameter, For the first Clinical dependence coefficient of each parameter, For the first Clinical dependence of each parameter For the first The scene correlation coefficient of each parameter, For the first The correlation between each parameter and the current scene label is divided into three levels of importance: high, medium, and low, based on the priority coefficient. Set the parsing priority sequence according to the importance level, and assign the parsing tasks corresponding to the core monitoring parameters to the high priority queue, and non-core parameters to the low priority queue. During the parsing process, parsing tasks are executed in order of priority queues. Parsing tasks in the high-priority queue occupy no less than 70% of the computing resources, and the resource allocation ratio is adjusted in real time based on parsing delay feedback. Parsing tasks in the low-priority queue are executed when computing resources are idle, and a dynamic priority adjustment feedback link is established, which includes dynamic updates of priority coefficients and optimization of queue scheduling.

5. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 3, characterized in that, The general data acquisition interface in step S2 supports multiple data transmission protocols, including HL7 protocol, DICOM protocol and custom transmission protocols of various equipment manufacturers. During the data acquisition process, the dynamic protocol identification and adaptation algorithm of the interface first verifies the device protocol type and then automatically adapts and is compatible to complete the compatible acquisition of heterogeneous raw data. The acquired data is stored in real time to a temporary data buffer. The buffer adopts a circular overwrite storage mechanism. The buffer capacity is dynamically configured according to the device data output rate and parameter importance level.

6. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, The unified standardized data model in step S4 is constructed based on medical data-related standards and industry-standard data specifications. The data model includes basic fields such as parameter name, data type, unit, collection timestamp, and device identifier. It also reserves three types of extended fields: cross-institutional collaboration identifier field, scenario semantic metadata field, and parameter priority identifier field. During the format conversion process, a field mapping table that supports both semantic fuzzy matching and precise mapping modes maps the heterogeneous parameter names and data structures of different devices to the corresponding fields of the standardized data model. The field mapping table is version-managed and supports manual updates and multi-center rule synchronization.

7. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, In step S4, storage optimization in a conventional monitoring scenario is achieved through a data compression algorithm. The compression algorithm selects the corresponding compression method according to the data type and parameter priority level. For numerical data, the LZ77 lossless compression algorithm is used, and for text data, the static dictionary encoding compression algorithm is used. The compressed standardized data is stored in the medical data center. During the storage process, a four-dimensional data index is established, consisting of scene tag, device identifier, collection timestamp, and priority level. The index is automatically updated every 24 hours and supports incremental update optimization.

8. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, The criteria for determining an emergency intervention scenario in step S4 are: core monitoring parameters exceeding the clinical safety threshold set based on clinical treatment guidelines, receiving a clinical emergency operation trigger signal transmitted through a dedicated signal channel, or parameter trend mutation characteristics meeting the emergency warning conditions. When an emergency intervention scenario is determined, the data processing thread scheduling mechanism is optimized to raise the priority of the data processing thread to the highest level, suspend non-core data processing processes, reduce unnecessary data preprocessing steps, retain only the core parsing and format conversion process, and push emergency data to the clinical warning terminal in real time.

9. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, In step S1, the core monitoring parameters and data accuracy requirements associated with the scene tags can be customized by medical staff according to the specific surgical type, patient condition characteristics, and multi-center collaboration needs. The configuration process is achieved through a visual interactive interface that includes four steps: parameter selection, accuracy setting, priority adjustment, and confirmation submission. The configuration information is stored in a distributed multi-center synchronous configuration database. When calling the parsing rules in step S3, the customized configuration information in the configuration database is read first, and then the data parsing priority and accuracy adaptation strategy are adjusted in combination with the preset parsing rules.

10. The method for automatic conversion and standardization of heterogeneous data formats for patient monitors / anesthesia machines according to claim 1, characterized in that, The data verification and backtracking steps in step S5 are as follows: standardized data is verified using dual rules of field integrity verification and data logic consistency verification. A four-level data anomaly tracing link is established, which includes scene matching records, priority adjustment logs, parsing process details, and original data fragments. Standardized data is associated with and stored with original collected data, scene tag matching records, parsing process logs, and resource scheduling records to ensure that data anomalies can be accurately traced back to the data source, processing stage, and resource allocation.

Images