Cardiology department nursing monitoring data abnormity identification and alarm method

Abstract

The invention discloses a cardiology department nursing monitoring data abnormity identification alarm method, and relates to the technical field of nursing monitoring data identification, and the method comprises the steps: building a monitoring data topological framework based on the monitoring data type of a multifunctional nursing monitoring device; collecting monitoring data of the multifunctional nursing monitoring equipment in a preset monitoring window according to a preset collection frequency, and generating a window monitoring data topological framework sequence in combination with the monitoring data topological framework; obtaining an enhanced window monitoring data topological framework; matching the enhanced window monitoring data topological framework with a monitoring data exception prototype library to obtain a matched monitoring data exception prototype; and obtaining first alarm information based on the early warning level matched with the abnormal prototype of the monitoring data. According to the invention, the technical problems of low monitoring data abnormity identification accuracy and lagging of alarm in the prior art are solved. The technical effect of improving the abnormal recognition reliability of the monitoring data is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of nursing monitoring data recognition, and in particular to a method for abnormal recognition and alarm of cardiovascular medicine nursing monitoring data. Background Art

[0002] Cardiovascular medicine nursing monitoring devices can collect patients' physiological parameters in real time, such as electrocardiogram signals, blood pressure, blood oxygen saturation, etc., and evaluate patients' health status through data analysis. Currently, multifunctional nursing monitoring devices are widely used in hospitals to continuously monitor patients' vital signs and remind nursing staff and doctors to take timely interventions through an abnormal alarm function. However, the existing abnormal recognition and alarm methods still have certain limitations.

[0003] Currently, abnormal alarms are mainly based on single data thresholds or simple rule matching. For example, when parameters such as heart rate and blood pressure exceed the set range, the system will trigger an alarm, mainly suffering from high false alarm rates and a lack of comprehensive analysis. Summary of the Invention

[0004] The present application provides a method for abnormal recognition and alarm of cardiovascular medicine nursing monitoring data, which is used to solve the technical problems of low accuracy in abnormal recognition of monitoring data and lag in alarm in the prior art.

[0005] In view of the above problems, the present application provides a method for abnormal recognition and alarm of cardiovascular medicine nursing monitoring data, and the method includes:

[0006] Building a monitoring data topology framework based on the types of monitoring data of the multifunctional nursing monitoring device; collecting the monitoring data of the multifunctional nursing monitoring device at a preset collection frequency within a preset monitoring window, and combining with the monitoring data topology framework to generate a window monitoring data topology framework sequence; performing window monitoring data enhancement on the window monitoring data topology framework sequence in chronological order from front to back to obtain an enhanced window monitoring data topology framework; matching the enhanced window monitoring data topology framework with a monitoring data abnormal prototype library to obtain a matching monitoring data abnormal prototype; and obtaining a first alarm message based on the warning level of the matching monitoring data abnormal prototype.

[0007] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0008] This application constructs a monitoring data topology framework based on the types of monitoring data of a multifunctional nursing monitoring device, and then collects the monitoring data of the multifunctional nursing monitoring device at a preset collection frequency within a preset monitoring window. Combining with the monitoring data topology framework, a sequence of window monitoring data topology frameworks is generated. Then, the sequence of window monitoring data topology frameworks is enhanced in the order of time from front to back to obtain an enhanced window monitoring data topology framework. Then, the enhanced window monitoring data topology framework is matched with a monitoring data anomaly prototype library to obtain a matching monitoring data anomaly prototype; based on the warning level of the matching monitoring data anomaly prototype, a first alarm message is obtained. It achieves the technical effects of improving the reliability of monitoring data anomaly recognition and shortening the anomaly recognition response time. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] FIG. Figure 1 is a schematic flowchart of a method for identifying and alarming abnormal monitoring data of cardiology nursing provided by an embodiment of the present invention.

[0010] FIG. Figure 2 is a schematic flowchart of constructing a monitoring data topology framework in a method for identifying and alarming abnormal monitoring data of cardiology nursing provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0011] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0012] It should be noted that the terms "include" and "have" are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0013] Embodiment, as shown in FIG. Figure 1 The present application provides a method for identifying and alarming abnormal monitoring data of cardiology nursing, wherein the method includes:

[0014] S1: Construct a monitoring data topology framework based on the types of monitoring data of a multifunctional nursing monitoring device;

[0015] Further, as shown in FIG. Figure 2 When constructing a monitoring data topology framework based on the types of monitoring data of a multifunctional nursing monitoring device, step S1 of the embodiment of the present application further includes:

[0016] Obtain a set of historical nursing anomaly monitoring logs;

[0017] Index the set of historical nursing anomaly monitoring logs by the anomaly monitoring data type, and obtain a set of historical nursing anomaly monitoring data type groups;

[0018] Extract the top K monitoring data types with the highest occurrence frequencies in the set of historical nursing anomaly monitoring data type groups as K central topology nodes, where K is a positive integer;

[0019] Build the monitoring data topology framework based on the K central topology nodes and the set of historical nursing anomaly monitoring data type groups.

[0020] Further, when building the monitoring data topology framework based on the K central topology nodes and the set of historical nursing anomaly monitoring data type groups, step S1 of this application embodiment further includes:

[0021] Index the set of historical nursing anomaly monitoring data type groups by the K central topology nodes, and obtain a set of historical nursing anomaly monitoring data types associated with the K nodes;

[0022] Identify the weights of the sets of historical nursing anomaly monitoring data types associated with the K nodes respectively, and obtain a set of weights associated with the K nodes;

[0023] Respectively use the historical nursing anomaly monitoring data types associated with the K nodes in the set of historical nursing anomaly monitoring data types associated with the K nodes as associated nodes to obtain a set of K associated nodes, and connect the set of K associated nodes with the corresponding central topology nodes to obtain a set of connection edges of the K associated nodes;

[0024] Use the set of weights associated with the K nodes to map and identify the set of connection edges of the K associated nodes, and obtain a set of marked connection edges of the K associated nodes;

[0025] Build the monitoring data topology framework based on the set of marked connection edges of the K associated nodes, the K central topology nodes, and the set of K associated nodes.

[0026] In a possible embodiment, the monitoring data topology framework refers to constructing a structured topology network for various nursing monitoring data (such as heart rate, blood pressure, blood oxygen saturation, etc.) according to the association relationships between the data, so as to perform anomaly identification and analysis. The historical nursing anomaly monitoring log set refers to the records of nursing monitoring data collected in the past, which contain the abnormal data that has occurred and its corresponding nursing situations. The multi-functional nursing monitoring device is used to perform nursing monitoring from multiple different dimensions, such as electrocardiogram monitoring, blood pressure monitoring, pulse monitoring, etc. The central topology node refers to the most core and representative monitoring data types in the nursing monitoring data network, and these data types appear more frequently in the historical anomaly logs. The associated node refers to the monitoring data type that has a strong association relationship with the central topology node.

[0027] In an embodiment, the anomaly monitoring data type refers to the type of physiological parameters marked as abnormal in the monitoring data, such as abnormal heart rate, blood pressure fluctuation, low blood oxygen, etc. Indexed by the anomaly monitoring data type, the historical nursing anomaly monitoring log set is retrieved to determine the anomaly monitoring data types that appear simultaneously in different historical nursing anomalies, and a historical nursing anomaly monitoring data type group is formed, thereby obtaining the set of historical nursing anomaly monitoring data type groups. Exemplarily, the historical nursing anomaly monitoring data type group includes low blood pressure + high heart rate, arrhythmia + blood oxygen drop, etc.

[0028] Traverse and extract the top K monitoring data types with higher overall occurrence frequencies in the set of historical nursing anomaly monitoring data type groups, and use them as the central topology nodes. For example, in a large amount of historical data, if the abnormal data of the three types of low blood pressure, arrhythmia, and blood oxygen drop appear most frequently, they will be selected as the central topology nodes. After obtaining the K central topology nodes, respectively taking the K central topology nodes as the clustering objects, the union of the historical nursing anomaly monitoring data types that appear simultaneously with the central topology nodes in the set of historical nursing anomaly monitoring data type groups is obtained to obtain the set of historical nursing anomaly monitoring data types associated with the K nodes.

[0029] Furthermore, according to the number of times the historical nursing anomaly monitoring data type associated with the node appears simultaneously with the central topology node, the association degree between it and the central topology node is determined. After analysis, the set of K node association weights is obtained. Among them, the greater the node association weight, the greater the association degree between the corresponding historical nursing anomaly monitoring data type associated with the node and the central topology node. Exemplarily, taking blood pressure anomaly as the central topology node, when "low blood pressure" and "rapid heart rate" appear simultaneously in 70% of the abnormal cases, while "low blood pressure" and "blood oxygen drop" only appear simultaneously in 40% of the cases, the weight of the associated node corresponding to the former "rapid heart rate" is higher.

[0030] Connect K associated nodes to the corresponding central topology nodes to form a set of K associated node connection edges. Moreover, through a set of K node association weights, different weight values are assigned to the connection edges, so that the data topology structure can more accurately reflect the association strength between data types. Thus, the goal of mapping and identifying the set of K associated node connection edges is achieved, and the set of K identified associated node connection edges is obtained. Exemplarily, numerical values can be used to mark the strength of the edges (such as 0.7 indicating high association and 0.4 indicating low association), or colors can be used to distinguish association relationships of different strengths.

[0031] That is to say, extract the K monitoring data types with the highest occurrence frequencies as the central topology nodes, such as "hypotension", "high heart rate", "oxygen desaturation", etc. Then, using the central topology nodes as indexes, retrieve their associated data types in the set of historical nursing abnormal monitoring data type groups to form a set of K associated nodes. For example, "hypotension" may be associated with "tachycardia", "oxygen desaturation", "shortness of breath", etc. Next, calculate the weight relationship between the central topology nodes and their associated nodes, and count the probability of different abnormal data types occurring simultaneously to form a set of K node association weights. And generate a set of identified associated node connection edges based on the weights, making the relationships between the nodes clear and quantifiable. For example, the association weight between "hypotension" and "tachycardia" is 0.7, while the association weight between "hypotension" and "oxygen desaturation" is 0.5. Subsequently, connect the central topology nodes to the associated nodes to form a monitoring data topology structure.

[0032] Furthermore, step S1 of the embodiment of the present application further includes:

[0033] Extract a first set of node-associated historical nursing abnormal monitoring data types from the set of K node-associated historical nursing abnormal monitoring data types;

[0034] Respectively divide the occurrence frequency of any one first node-associated historical nursing abnormal monitoring data type in the first set of node-associated historical nursing abnormal monitoring data types by the total occurrence frequency of the data types in the first set of node-associated historical nursing abnormal monitoring data types to obtain a first set of node association weights;

[0035] Perform weight identification on the set of K node-associated historical nursing abnormal monitoring data types to obtain a set of K node association weights.

[0036] In a possible embodiment, the set of historical nursing anomaly monitoring data types associated with the first node is a subset extracted from the set of historical nursing anomaly monitoring data types associated with K nodes, and it contains the anomaly monitoring data types associated with a specific node (the first node). The occurrence frequency of the historical nursing anomaly monitoring data types associated with the first node refers to the number of times a specific anomaly monitoring data type appears in all the sets of historical nursing anomaly monitoring data types. The set of weights associated with the first node refers to the set of weights calculated based on the occurrence frequencies of the various data types in the set of historical nursing anomaly monitoring data types associated with the first node, and the weights are based on the relative occurrence frequencies of each data type in the set. The set of weights associated with K nodes refers to the set of weights calculated and obtained according to the occurrence frequencies of the data types in the set of historical nursing anomaly monitoring data types associated with K nodes and their association relationships.

[0037] In a possible embodiment, for each data type in the set of historical nursing anomaly monitoring data types associated with the first node, the weight of each data type is obtained by calculating the ratio of its occurrence frequency in this set to the total frequency of all data types. For example, if "too fast heart rate" appears 100 times among all the data associated with "low blood pressure", and the total occurrence frequency of these data types is 200 times, then the weight of "too fast heart rate" is 100 / 200 = 0.5. This process is repeated for each associated data type, and finally, the set of weights associated with the first node is obtained, for example, containing data type weights such as 0.5, 0.3, etc. Based on the same calculation principle, similar weight calculations are performed on all the data types in the set of historical nursing anomaly monitoring data types associated with K nodes to obtain the set of weights associated with K nodes. By calculating the ratio of the occurrence frequency of each monitoring data type to the total frequency, the weight relationship between each node is determined, thereby accurately establishing the association strength between the monitoring data types, achieving the technical effect of providing more accurate weight support for the subsequent construction of the topological framework.

[0038] S2: Collect the monitoring data of the multifunctional nursing monitoring device at a preset collection frequency within a preset monitoring window, and combine it with the monitoring data topological framework to generate a window monitoring data topological framework sequence;

[0039] In a possible embodiment, the preset monitoring window is a time period or a time window for data collection preset by those skilled in the art, and it is used to limit the scope of data collection. This time period can be seconds, minutes, hours, etc., and is set according to the monitoring requirements. The preset collection frequency refers to the frequency or time interval of data collection within the monitoring window. For example, it can be set to collect data once per second, once per minute, or once per hour. The window monitoring data topological framework sequence refers to the time sequence of the topological frameworks formed by the continuously collected monitoring data over time within a monitoring window. It contains the data topological relationships at different moments.

[0040] Preferably, within a preset monitoring window, data of the multifunctional nursing monitoring device, such as heart rate, blood pressure, blood oxygen, etc., are continuously collected according to a preset collection frequency (such as per second, per minute). These data are organized by monitoring type and combined with the previously constructed monitoring data topology framework to determine the association relationships between data types.

[0041] For example, within a 5-minute window, the collected data may include "normal heart rate" and "slightly low blood oxygen". By combining with the topology framework, it can be determined which data types appear simultaneously within this window and which data types are associated with each other, so as to fill the collected data into the monitoring data topology framework to generate a window monitoring data topology framework sequence. Among them, the window monitoring data topology framework sequence reflects the change situation of the monitoring data of the multifunctional nursing monitoring device within the preset monitoring window.

[0042] By collecting data at a frequency within a preset monitoring window and combining with the topology framework, a dynamic monitoring data topology framework sequence is generated to better reflect the association situation between data types under time change, providing a basis for subsequent data enhancement and anomaly recognition.

[0043] S3: Perform window monitoring data enhancement on the window monitoring data topology framework sequence in the order from front to back in time to obtain an enhanced window monitoring data topology framework;

[0044] In a possible embodiment, the enhanced window monitoring data topology framework reflects the situation of the device monitoring data after enhancement within the preset monitoring window. Among multiple monitoring windows, as time goes by, different window monitoring data topology frameworks will be generated. To enhance these data, the data of the previous window can be fused or extended with the data of the current window in chronological order to ensure the continuity and consistency of the data in time. For example, in the previous monitoring window, it may be detected that there is a strong association between "low blood pressure" and "high heart rate", while in the current window, it may be detected that the relationship between "low blood pressure" and "shortness of breath" is enhanced. Then, by combining the data of the front and back windows, these changes can be captured, so as to perform data enhancement on the monitoring data of the current window to make it better reflect the actual situation of the monitoring data. After enhancement, the topology framework may show a stronger node association degree, thus improving the sensitivity of anomaly recognition and alarm.

[0045] Furthermore, performing window monitoring data enhancement on the window monitoring data topology framework sequence in the order from front to back in time to obtain an enhanced window monitoring data topology framework, step S3 of the embodiment of the present application further includes:

[0046] Extract the first window monitoring data topology framework and the second window monitoring data topology framework from the window monitoring data topology framework sequence in chronological order from front to back;

[0047] Perform topological node mapping approximation degree recognition on the first window monitoring data topology framework and the second window monitoring data topology framework to obtain the first topological node mapping approximation degree set;

[0048] Perform normalization processing on the first topological node mapping approximation degree set, and construct the first topological node mapping matrix according to the normalization processing result;

[0049] Perform convolution analysis on the first topological node mapping matrix and the second window monitoring data topology framework to obtain the first-stage enhanced window monitoring data topology framework;

[0050] Perform window monitoring data enhancement on the remaining window monitoring data topology frameworks in the window monitoring data topology framework sequence based on the first-stage enhanced window monitoring data topology framework to obtain the enhanced window monitoring data topology framework.

[0051] Specifically, the first topological node mapping approximation degree set reflects the similarity degree between different topological nodes in the first window monitoring data topology framework and the second window monitoring data topology framework. The first topological node mapping matrix reflects the data association degree between different topological nodes between the first window monitoring data topology framework and the second window monitoring data topology framework. The first-stage enhanced window monitoring data topology framework reflects the enhanced framework after the second window monitoring data topology framework integrates the data information reflected by the first window monitoring data topology framework.

[0052] Preferably, extract the first window monitoring data topology framework and the second window monitoring data topology framework from the entire window monitoring data topology framework sequence in chronological order for calculating the change trend of the topological structure. Use the cosine similarity calculation formula to quantitatively calculate the similarity degree between the topological nodes at the same position in the first window monitoring data topology framework and the second window monitoring data topology framework, so as to obtain the first topological node mapping approximation degree set.

[0053] Preferably, use the softmax formula to perform normalization processing on the first topological node mapping approximation degree set, normalize the approximation degree to between 0 and 1, and fill the data after normalization into an initially empty matrix to obtain the first topological node mapping matrix. Furthermore, use the convolution analysis network layer to perform convolution operations on the first topological node mapping matrix and the second window monitoring data topology framework to obtain the first-stage enhanced window monitoring data topology framework.

[0054] In one embodiment, a plurality of sample first topological node mapping matrices and a plurality of sample second window monitoring data topological frameworks are obtained as training data, and the framework constructed based on the convolutional neural network is supervised and trained using the training data until the training converges, obtaining the trained convolutional analysis network layer. Preferably, based on the same principle as obtaining the first-stage enhanced window monitoring data topological framework, the remaining window monitoring data topological frameworks in the window monitoring data topological framework sequence are subjected to window monitoring data enhancement to obtain the enhanced window monitoring data topological framework. By performing topological mapping on adjacent frameworks, calculating their similarity, and using normalization and convolutional analysis methods for data enhancement, a more stable and reliable monitoring data topological framework can be obtained. This helps improve the accuracy of anomaly recognition and enhances the processing ability for complex nursing monitoring data.

[0055] S4: Match the enhanced window monitoring data topological framework with the monitoring data anomaly prototype library to obtain a matching monitoring data anomaly prototype;

[0056] S5: Obtain a first alarm message based on the warning level of the matching monitoring data anomaly prototype.

[0057] Furthermore, when matching the enhanced window monitoring data topological framework with the monitoring data anomaly prototype library to obtain a matching monitoring data anomaly prototype, step S4 of the embodiment of the present application further includes:

[0058] Using a probability matching function, traverse and calculate the prototype matching probability coefficients between the enhanced window monitoring data topological framework and the monitoring data prototypes in the monitoring data anomaly prototype library to obtain a set of prototype matching probability coefficients;

[0059] Extract the monitoring data anomaly prototype corresponding to the maximum value in the set of prototype matching probability coefficients as the matching monitoring data anomaly prototype.

[0060] Furthermore, the probability matching function is:

[0061] l(y = c∣x i ) = exp(-d(x i ,p c ));

[0062] Wherein, l(y = c∣x i ) is the prototype matching probability coefficient of the enhanced window monitoring data topological framework belonging to the monitoring data prototype, y is the event that the enhanced window monitoring data topological framework belongs to the monitoring data prototype, x i is the enhanced window monitoring data topological framework, p c is the monitoring data prototype, and d(x i ,p c ) is the distance metric function.

[0063] Further, the distance metric function is:

[0064] x ij To enhance the data of the j-th topology node in the window monitoring data topology framework, p cj Is the data of the j-th topology node in the monitoring data prototype, n i Is the total number of topology nodes in the window monitoring data topology framework.

[0065] In a possible embodiment, the monitoring data anomaly prototype library stores known anomaly data patterns. Each anomaly prototype represents the data topology framework corresponding to a specific nursing monitoring data anomaly situation, such as arrhythmia, hypotensive shock, etc., and is accompanied by a corresponding warning level. The prototype matching probability coefficient represents the matching degree between the enhanced window monitoring data topology framework and an anomaly prototype in the monitoring data anomaly prototype library. Its value range is usually between 0 and 1. The higher the value, the higher the matching degree. The warning level is the risk level corresponding to the matched anomaly prototype, usually divided into three levels: low, medium, and high, for different levels of nursing response. Exemplarily, the risk levels include: low-level warning (only display a reminder at the nursing workstation, no immediate intervention is required), medium-level warning (notify the nurse for additional monitoring and recommend taking nursing measures), high-level warning (immediately trigger an emergency alarm, notify the doctor, and initiate an emergency plan).

[0066] Preferably, a probability matching function is used to traverse and calculate the matching probability between the enhanced window monitoring data topology framework and each anomaly prototype in the monitoring data anomaly prototype library. For example, if the data pattern in the current topology framework (such as hypotension + high heart rate + shortness of breath) has a high matching probability with the "hypotensive shock" prototype in the database, it may be determined that the patient is in a state of hypotensive shock. During the calculation process, all matching probability values are stored in the prototype matching probability coefficient set for subsequent screening.

[0067] Furthermore, select the monitoring data anomaly prototype corresponding to the maximum value from the prototype matching probability coefficient set, that is, the most likely anomaly situation. For example, if the matching coefficient of an anomaly prototype is 0.92 (close to 1), then this prototype is considered the most likely anomaly situation corresponding to the current data. Thus, accurate identification of anomaly situations is achieved, rather than false alarms or missed reports. For example, if the data pattern in the current topology framework (such as hypotension + high heart rate + shortness of breath) has a high matching probability with the "hypotensive shock" prototype in the database, it may be determined that the patient is in a state of hypotensive shock.

[0068] Each abnormal prototype is associated with a warning level (such as low, medium, high) in the monitoring data abnormal prototype library. For example, if "severe hypotensive shock" is matched, it may trigger a "high"-level warning, while if "mild arrhythmia" is matched, it may trigger a "low"-level warning. Furthermore, according to the warning level of the matched monitoring data abnormal prototype, the system generates a first alarm message and sends it to the nursing staff or the medical system.

[0069] By matching the enhanced window monitoring data topology framework with known abnormal patterns, potential abnormal conditions of the patient are identified, and corresponding-level alarms are triggered based on the matching results. The technical effects of improving the automation level of nursing monitoring, reducing the workload of manual analysis, and simultaneously enhancing the timeliness and accuracy of nursing response are achieved.

[0070] In summary, the embodiments of the present application at least have the following technical effects:

[0071] By introducing the monitoring data topology framework, the present application achieves the goal of enhancing data correlation, and through topology node mapping approximation recognition and convolutional analysis for data enhancement, it can more effectively capture the temporal changes between data, achieving the goal of improving the accuracy of anomaly detection. Furthermore, according to the matched abnormal prototype and its warning level (low, medium, high), graded alarm information is provided, achieving the technical effect of improving the reliability of anomaly recognition and alarm.

[0072] It should be noted that the above sequence of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above specific embodiments of this specification have been described. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0073] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0074] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, changes, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A cardiology nursing monitoring data abnormality identification and alarm method, characterized in that: The method comprises: Build a monitoring data topology framework based on the monitoring data types of multifunctional nursing monitoring equipment; Collecting the monitoring data of the multifunctional nursing monitoring device in a preset monitoring window according to a preset collection frequency, and generating a window monitoring data topology framework sequence in combination with the monitoring data topology framework; Performing window monitoring data enhancement on the window monitoring data topology frame sequence in a time-ordered order to obtain an enhanced window monitoring data topology frame; Matching the enhanced window monitoring data topology framework with the monitoring data anomaly prototype library to obtain matching monitoring data anomaly prototypes; Based on the warning level of the matching monitoring data abnormal prototype, a first alarm information is obtained.

2. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 1, characterized in that: Performing window monitoring data enhancement on the window monitoring data topology framework sequence in a time-ordered order to obtain an enhanced window monitoring data topology framework includes: Extracting a first window monitoring data topology frame and a second window monitoring data topology frame from the window monitoring data topology frame sequence in a time-ordered order; Perform topological node mapping approximation identification on the first window monitoring data topological framework and the second window monitoring data topological framework to obtain a first topological node mapping approximation set; Performing normalization processing on the first topological node mapping approximation set, and constructing a first topological node mapping matrix according to the normalization processing result; Perform convolution analysis on the first topological node mapping matrix and the second window monitoring data topological framework to obtain a first-stage enhanced window monitoring data topological framework; Based on the first-stage enhanced window monitoring data topology framework, window monitoring data enhancement is performed on the remaining window monitoring data topology frameworks in the window monitoring data topology framework sequence to obtain the enhanced window monitoring data topology framework.

3. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 1, characterized in that: The enhanced window monitoring data topology framework is matched with the monitoring data anomaly prototype library to obtain a matching monitoring data anomaly prototype, including: Using a probability matching function, traversing and calculating the prototype matching probability coefficients between the enhanced window monitoring data topology framework and the monitoring data prototypes in the monitoring data anomaly prototype library, to obtain a prototype matching probability coefficient set; The monitoring data anomaly prototype corresponding to the maximum value in the prototype matching probability coefficient set is extracted as the matching monitoring data anomaly prototype.

4. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 3, characterized in that: The probability matching function is: l(y=c∣x i )=exp(-d(x i ,p c )); Where l(y=c|x i ) is the prototype matching probability coefficient of the enhanced window monitoring data topology framework belonging to the monitoring data prototype, y is the event that the enhanced window monitoring data topology framework belongs to the monitoring data prototype, x i To enhance the window monitoring data topology framework, p c is the monitoring data prototype, d(x i ,p c ) is the distance metric function.

5. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 4, characterized in that: The distance metric function is: x ij To enhance the data of the jth topological node in the window monitoring data topological framework, p cj is the data of the jth topological node in the monitoring data prototype, n i The total number of topology nodes in the data topology framework for the enhanced window monitoring.

6. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 1, characterized in that: The monitoring data topology framework is built based on the monitoring data types of multifunctional nursing monitoring equipment, including: Get the historical nursing abnormality monitoring log collection; Using the abnormal monitoring data type as an index, searching the historical nursing abnormal monitoring log set to obtain a historical nursing abnormal monitoring data type group set; Extract K monitoring data types with the top K occurrence frequencies in the historical nursing abnormal monitoring data type group set as K central topological nodes, where K is a positive integer; The monitoring data topology framework is built based on the K central topology nodes and the historical nursing abnormality monitoring data type group set.

7. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 6, characterized in that: The monitoring data topology framework is constructed based on the K central topology nodes and the historical nursing abnormal monitoring data type group set, including: Using the K central topological nodes as indexes, searching the historical nursing abnormality monitoring data type group set to obtain the K node-associated historical nursing abnormality monitoring data type set; Performing weight identification on the K node-associated historical nursing abnormality monitoring data type sets respectively to obtain K node-associated weight sets; The node-associated historical nursing abnormality monitoring data types in the K node-associated historical nursing abnormality monitoring data type sets are respectively used as associated nodes to obtain K associated node sets, and the K associated node sets are connected to the corresponding central topological nodes to obtain K associated node connection edge sets; Using the K node association weight sets, mapping and identifying the K association node connection edge sets, to obtain K identified association node connection edge sets; The monitoring data topology framework is constructed based on the K identification-associated node connection edge sets, the K central topological nodes and the K associated node sets.

8. A cardiology nursing monitoring data abnormality identification and alarm method as claimed in claim 7, characterized in that: include: Extracting a first node-associated historical nursing anomaly monitoring data type set from the K node-associated historical nursing anomaly monitoring data type sets; The occurrence frequency of any first node associated historical nursing abnormality monitoring data type in the first node associated historical nursing abnormality monitoring data type set is respectively compared with the total occurrence frequency of the data type in the first node associated historical nursing abnormality monitoring data type set to obtain a first node associated weight set; Weight identification is performed on the K node-associated historical nursing abnormality monitoring data type set to obtain a K node-associated weight set.

Images