Nutrition pump medication path verification method and system based on misconnection identification mechanism

Abstract

The invention relates to the technical field of medical information management, in particular to a nutrition pump medication path verification method and system based on a misconnection recognition mechanism, and the method comprises the following steps: S1, collecting operation behavior data and path state data of key periods before and after the infusion of a nutrition pump is started, and constructing a path behavior initial map for describing abnormal features; s2, identifying a misconnection behavior unit based on the path behavior initial map, extracting an abnormal channel response and unexpected liquid medicine combination, constructing a misconnection identification rule set, and generating a misconnection risk prediction model; and S3, dynamically verifying the current infusion path by using the prediction model, and if the current infusion path is matched with a rule and high risk is judged, triggering a medication path alarm signal. According to the method, high-precision identification and real-time early warning of nutrition pump medication path misconnection behaviors are achieved by constructing the path behavior map, extracting the misconnection features and combining the dynamic risk scores, and the safety and the intelligent level of clinical infusion are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of medical information management technology, and in particular to a method and system for verifying the medication route of a nutrition pump based on a misconnection identification mechanism. Background Technology

[0002] With the increasing demand for clinical nutritional therapy, nutrition pumps are being used more and more widely in scenarios involving the parallel infusion of multiple medications and multiple pathways. To ensure patient safety, accurate connection of infusion pathways and consistency of medication labeling have become critical links. However, in actual operation, especially under conditions of night duty, heavy nursing workload, or complex equipment interfaces, nutrition pump pathways are prone to problems such as misconnection, incorrect connection, and mixed connection, which directly affect the medication effect and bring safety risks. Therefore, it is of great significance to establish a pathway verification mechanism that can identify the risk of misconnection during infusion and respond in a timely manner.

[0003] Existing technologies mainly rely on label recognition, liquid code verification, or single-channel flow rate monitoring to determine the path. However, these methods generally suffer from problems such as dependence on manual intervention, weak adaptability to dynamic changes, and delayed response to changes in path status and behavior. They are difficult to effectively identify misconnected path behaviors that are both hidden and diverse. In addition, some methods lack the structured use of historical path behavior data and cannot establish a targeted and generalizable risk identification model, making it difficult to avoid misjudgments and omissions. Summary of the Invention

[0004] This invention provides a method and system for verifying the medication delivery route of a nutrition pump based on a misconnection identification mechanism. By constructing an initial path behavior map that integrates operational behavior data and path status data, it identifies misconnection path behavior units and extracts key abnormal features. It then generates a path misconnection risk prediction model by combining a rule set and historical records, and performs dynamic verification and early warning response during real-time infusion. This enables intelligent and high-precision identification and closed-loop alarm of misconnection paths, significantly improving the safety and management efficiency of nutrition pump medication delivery routes.

[0005] A method for validating nutrient pump medication pathways based on a misconnection identification mechanism includes the following steps: S1. Collect operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. The operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. The path status data includes channel response sequence and drug identification code information when the path is connected. Construct an initial path behavior map to characterize abnormal path features. S2, based on the initial path behavior map, identify the misconnection path behavior units, extract the path data that show abnormal channel response characteristics and undesirable drug liquid identification combinations, construct the misconnection identification rule set, and combine it with the historical path cross record library to generate the target path misconnection risk prediction model. S3. Apply the target path misconnection risk prediction model to the real-time infusion path of the current nutrient pump and perform dynamic verification on each infusion path. If the path behavior characteristics match the misconnection identification rule set and the target path misconnection risk prediction model judges it as a high-risk path, then trigger the medication path alarm signal.

[0006] Optionally, S1 includes: S11. Before and after infusion initiation, multi-point sampling is performed at key time periods (5 seconds before initiation, instantaneous initiation, and within 10 seconds after initiation) to collect operational behavior data of the nutrient pump, including initiation delay time. (Time difference between the issuance of the control command and the actual response of the pump), pressure feedback change curve (Pressure sensor feedback during pumping), residual liquid flow rate sequence (Changes in the flow rate of residual drug solution in the initial stage of pumping); S12, After the path is successfully connected, collect path status data, including the channel response sequence. Medicine identification code ; S13, Unify and align the collected operational behavior data and path status data to construct an initial path behavior map for characterizing path anomaly features. .

[0007] Optionally, S13 includes: S131, Time axis normalization is performed on the operational behavior data and path status data, with the unified benchmark being the pump start trigger time. Reconstruct the relative timeline; S132 uses linear interpolation to uniformly map runtime behavior data to a fixed time step. At equally spaced time points; S133, for each alignment time point Above, extract the current estimated values ​​of all operational behavior data and path status data channels to construct a path behavior status vector. ; S134, the obtained multiple path behavior state vectors The data is arranged in chronological order to form a path behavior data sequence. ; S135, construct the path behavior state sequence as a graph structure, which consists of nodes and edges, represented as follows: ; in, This forms the initial graph for path behavior. Each node in the graph represents the path behavior state at a given point in time. For graph edges, represent the continuous evolution of path states over time.

[0008] Optionally, S2 includes: S21. From the initial path behavior map, select nodes that meet the criteria of channel response change exceeding a set threshold or drug identification code not matching the current prescription as abnormal candidate nodes. Then, gather the abnormal candidate nodes and their temporally continuous adjacent nodes in the initial path behavior map to form a misconnected path behavior unit, and extract the combination of its channel response change pattern and drug identification code. S22. Based on the extracted misconnection path behavior units, abnormal features are summarized, a misconnection identification rule set is constructed, and historical path cross-records are combined to generate a target path misconnection risk prediction model.

[0009] Optionally, S21 includes: S211, in the initial path behavior graph In, for each node Perform anomaly detection; if the node channel response variation amplitude Exceeding the set anomaly detection threshold or its corresponding drug identification code Not part of the set of identifiers allowed by the current prescription That is, satisfying or Then the node Nodes identified as anomalous candidate nodes are grouped together with their temporally consecutive adjacent nodes in the initial path behavior graph to form misconnection path behavior units. For each misconnection path behavior unit, its corresponding channel response change pattern is extracted. and the combination of drug liquid identification codes involved. , as a sample of misconnection features; S212, for each abnormal candidate node Find temporally consecutive neighboring nodes in the initial path behavior graph. and If an adjacent node exists in the initial path behavior graph, then it is joined with... These are grouped together into a single misconnection path behavior unit, forming a candidate set of misconnection behaviors. ; S213, for each misconnection path behavior unit Extract the channel response change patterns respectively Combination with drug liquid identification code Finally, the feature set of misconnection behavior is obtained. .

[0010] Optionally, S22 includes: S221, misconnection path behavior unit Extracted channel response change patterns Combination with drug liquid identification code As input data, the K-Means clustering algorithm is applied to cluster misconnection behaviors, identifying different types of misconnection patterns and forming a set of misconnection behavior patterns. ; S222, based on the set of misconnection behavior patterns By summarizing the common characteristics within each cluster, a set of rules for misconnection identification is formed. Each rule Represented as: ; in, For the first The range of channel response modes corresponding to each rule. For the first The combination of illegal drug liquid identifiers corresponding to the rule; S223, Introducing a historical path cross-recording library The system combines the current path behavior state sequence to construct a rule matching vector, and generates a target path misconnection risk prediction model based on this vector, which is used to assess the misconnection risk of the current path.

[0011] Optionally, S3 includes: S31. During the nutrient pump infusion process, the behavior state sequence of the current path is acquired in real time and matched with the pre-built misconnection identification rule set. For each path, based on whether it meets the channel response change mode range and illegal drug liquid identification combination in the rule conditions, a corresponding rule matching vector is generated to mark whether the current path has structural features of potential misconnection behavior. If it meets at least one rule, it is marked as a suspected misconnection path. S32, for suspected misconnected paths, call the target path misconnection risk prediction model to calculate the risk score. If the risk score is not lower than the set path misconnection risk score judgment threshold, it is judged as a high-risk path and the medication path alarm signal is immediately triggered.

[0012] Optionally, S31 includes: S311, during the nutrient pump infusion process, based on a preset time step. Collect real-time operational behavior data and path status data of the current path to construct a time-series state vector of the path. And form a path behavior sequence at continuous time steps. ; S312, path behavior sequence Misconnection identification rule set Perform a comparison; S313, combine all rule matching results to form the rule matching vector for the current path. If it exists If the current path matches at least one misconnection rule, then the path is marked as a suspected misconnection path. For rules The matching results.

[0013] Optionally, S32 includes: S321, Calculate the risk score based on the target path misconnection risk prediction model. And set a threshold for determining the risk of misconnection of a path for risk assessment. If it is, then it is determined to be a high-risk path, among which, The threshold for determining the risk of misconnection of a route; S322: When the score determines that the path is high-risk, the medication path alarm signal is immediately triggered, and the prompting mechanism is activated through the nutrition pump control interface, including a buzzer, screen flashing, and status lights.

[0014] A nutrition pump medication pathway verification system based on a misconnection identification mechanism, used to implement the aforementioned nutrition pump medication pathway verification method based on a misconnection identification mechanism, includes the following modules: Data acquisition module: Collects operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. The operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. The path status data includes channel response sequence and drug identification code information, and constructs an initial path behavior map. Anomaly identification module: Based on the initial path behavior map, it identifies misconnection path behavior units that exhibit abnormal channel response characteristics and undesirable drug liquid identification combinations, constructs a misconnection identification rule set, and generates a target path misconnection risk prediction model by combining it with the historical path cross-record library. Path verification and alarm module: The target path misconnection risk prediction model is applied to the real-time infusion path of the current nutrient pump to dynamically verify whether the path behavior matches the misconnection identification rule set. If it is judged to be a high-risk path, a medication path alarm signal is triggered.

[0015] The beneficial effects of this invention are: This invention constructs an initial path behavior map, which integrates multidimensional features of nutrient pump operation behavior data and path status data to form a complete, time-aligned path behavior representation. This enables the capture of anomalies at key nodes before and after infusion initiation, effectively compensating for the perception blind spots of traditional single-point flow rate or label recognition methods in dynamic processes, thereby improving the coverage and accuracy of path anomaly identification.

[0016] This invention, by extracting the combined features of channel response change patterns and drug identification codes, and combining cluster analysis and rule induction methods, can systematically identify multiple types of misconnection behavior patterns, and establish a rule set and risk prediction model with adaptive feature expression capabilities. This mechanism not only has the ability to distinguish complex situations such as multi-path intersections and mixed drug paths, but also can achieve clustering and structural generalization of unknown misconnection patterns, improving the comprehensiveness of misconnection identification and the universality of the model.

[0017] This invention applies a target path misconnection risk prediction model to the real-time infusion process of a nutrition pump. Combined with rule-matching vector generation and a dynamic scoring mechanism, the system can achieve real-time judgment and hierarchical control of path misconnection risk. After identifying high-risk paths, it automatically triggers medication path alarm signals, constructing a closed-loop infusion safety assurance system, significantly reducing the risk of medication misconnection, and improving the level of clinical medication safety and management intelligence. Attached Figure Description

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

[0019] Figure 1 This is a schematic diagram of the verification method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the system functional modules according to an embodiment of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Those skilled in the art may employ other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0021] like Figure 1 As shown, a method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism includes the following steps: S1. Collect operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. Operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. Path status data includes channel response sequence and drug identification code information when the path is connected. Construct an initial path behavior map to characterize abnormal path features. S2, based on the initial path behavior map, identify the misconnection path behavior units, extract the path data that show abnormal channel response characteristics and undesirable drug liquid identification combinations, construct the misconnection identification rule set, and combine it with the historical path cross record library to generate the target path misconnection risk prediction model. S3. Apply the target path misconnection risk prediction model to the real-time infusion path of the current nutrient pump and perform dynamic verification on each infusion path. If the path behavior characteristics match the misconnection identification rule set and the target path misconnection risk prediction model judges it as a high-risk path, then trigger the medication path alarm signal.

[0022] S1 includes: S11. Before and after infusion initiation, multi-point sampling is performed at key time periods (5 seconds before initiation, instantaneous initiation, and within 10 seconds after initiation) to collect operational behavior data of the nutrient pump, including initiation delay time. (Time difference between the issuance of the control command and the actual response of the pump), pressure feedback change curve (Pressure sensor feedback during pumping), residual liquid flow rate sequence (Changes in the flow rate of residual drug solution in the initial stage of pumping); The estimated residual liquid flow velocity is expressed as: ; in, The path segment length, The point at which the pressure begins to rise. This refers to the point in time when the pressure reaches its peak. S12, After the path is successfully connected, collect path status data, including the channel response sequence. Medicine identification code ; The channel response sequence is used to identify the presence of transient anomalous changes by calculating the response difference, as shown below: ; in, For the path at a point in time The channel response value, The difference between continuous responses For the path at a point in time The channel response value; If there exist multiple consecutive time points that satisfy If it is, then it is marked as an abnormal connection path status, where, This is the abnormal threshold for changes in the channel response; ; in, The channel response difference in the historical normal connection path The mean, The corresponding standard deviation is... Amplification factor; S13, Unify and align the collected operational behavior data and path status data to construct an initial path behavior map for characterizing path anomaly features. .

[0023] S13 includes: S131, Time axis normalization is performed on the operational behavior data and path status data, with the unified benchmark being the pump start trigger time. The reconstructed relative timeline is represented as follows: ; in, For the first The relative time of each data record The timestamp of the original data collection; S132 uses linear interpolation to uniformly map runtime behavior data to a fixed time step. At equally spaced time points, it is represented as: ; in, For the first Each behavioral channel at standard time The estimated value, For channel At the original time point The sampled values, Alignment time with the target; S133, for each alignment time point Above, extract the current estimated values ​​of all operational behavior data and path status data channels to construct a path behavior status vector. , represented as: ; in, For time points Pressure feedback value, This represents the residual liquid flow rate. This indicates a startup delay state (0 indicates normal, 1 indicates timeout). This is the channel response value. For identification code of the medicine liquid; S134, the obtained multiple path behavior state vectors The data is arranged in chronological order to form a path behavior data sequence. , represented as: ; in, The number of time steps within the path behavior observation window; S135, construct the path behavior state sequence as a graph structure, which consists of nodes and edges, represented as follows: ; in, This forms the initial graph for path behavior. Each node in the graph represents the path behavior state at a given point in time. For graph edges, represent the continuous evolution of path states over time.

[0024] S2 includes: S21. From the initial path behavior map, select nodes that meet the criteria of channel response change exceeding a set threshold or drug identification code not matching the current prescription as abnormal candidate nodes. Then, gather the abnormal candidate nodes and their temporally continuous adjacent nodes in the initial path behavior map to form a misconnected path behavior unit, and extract the combination of its channel response change pattern and drug identification code. S22. Based on the extracted misconnection path behavior units, abnormal features are summarized, a misconnection identification rule set is constructed, and historical path cross-records are combined to generate a target path misconnection risk prediction model.

[0025] S21 includes: S211, in the initial path behavior graph In, for each node Perform anomaly detection; if the node channel response variation amplitude Exceeding the set anomaly detection threshold or its corresponding drug identification code Not part of the set of identifiers allowed by the current prescription That is, satisfying or Then the node Nodes identified as anomalous candidate nodes are grouped together with their temporally consecutive adjacent nodes in the initial path behavior graph to form misconnection path behavior units. For each misconnection path behavior unit, its corresponding channel response change pattern is extracted. and the combination of drug liquid identification codes involved. As a misconnection feature sample, it is represented as: ; ; ; in, , , These are the channel response values ​​for the previous node, the current node, and the [other node's] node, respectively. , , The first , , The liquid identification code of each node; ; in, This represents the mean of the differences in channel response changes during the steady-state period. The standard deviation of the channel response variation. To determine the threshold amplification factor; S212, for each abnormal candidate node Find temporally consecutive neighboring nodes in the initial path behavior graph. and If an adjacent node exists in the initial path behavior graph, then it is joined with... These are grouped together into a single misconnection path behavior unit, forming a candidate set of misconnection behaviors. ; S213, for each misconnection path behavior unit Extract the channel response change patterns respectively Combination with drug liquid identification code Finally, the feature set of misconnection behavior is obtained. , represented as: .

[0026] S22 includes: S221, misconnection path behavior unit Extracted channel response change patterns Combination with drug liquid identification code As input data, the K-Means clustering algorithm is applied to cluster misconnection behaviors, identifying different types of misconnection patterns and forming a set of misconnection behavior patterns. Specifically, it includes: (1) Definition of eigenvector: The eigenvector is defined as follows: The feature representation of each misconnected path behavior unit is a vector, represented as follows: ; in, For the first Feature vector of each misconnected path behavior unit The feature encoding function encodes the channel response sequence and the drug liquid identifier set into a unified numerical vector; (2) K-Means clustering process: Let the clustering be as follows: The goal of K-Means is to minimize the sum of squared Euclidean distances from all sample points to the center of their respective clusters. The objective function is: ; ; in, For the first A cluster, For the first The center point of each cluster, The set number of clusters; (3) Output results: The final set of misconnection behavior patterns is represented as follows: ; S222, based on the set of misconnection behavior patterns By summarizing the common characteristics within each cluster, a set of rules for misconnection identification is formed. Each rule Represented as: ; in, For the first The range of channel response modes corresponding to each rule. For the first The combination of illegal drug liquid identifiers corresponding to the rule; ; in, For the first The positional mean of all channel response sequences in the cluster. For positional standard deviation, Amplification factor to control response tolerance bandwidth; ; in, For candidate drug solution identifier combinations, For the set of legal drug pathway pairings, Based on the support threshold definition, let this combination appear in [the context of support thresholds]. The proportion is greater than 20%; S223, Introducing a historical path cross-recording library A rule-matching vector is constructed by combining the current path behavior state sequence, and a target path misconnection risk prediction model is generated based on this vector to assess the misconnection risk of the current path. Specifically, this includes: (1) Construction of misconnection rule matching vector: ; ; in, For the rule matching vector, For the first Rules Is it hit by the current path? The total number of rules; (2) Path misconnection risk scoring function: ; in, Assess the risk of misconnection along the current path. For the first The weight of each rule.

[0027] S3 includes: S31. During the nutrient pump infusion process, the behavior state sequence of the current path is acquired in real time and matched with the pre-built misconnection identification rule set. For each path, based on whether it meets the channel response change mode range and illegal drug liquid identification combination in the rule conditions, a corresponding rule matching vector is generated to mark whether the current path has structural features of potential misconnection behavior. If it meets at least one rule, it is marked as a suspected misconnection path. S32, for suspected misconnected paths, call the target path misconnection risk prediction model to calculate the risk score. If the risk score is not lower than the set path misconnection risk score judgment threshold, it is judged as a high-risk path and the medication path alarm signal is immediately triggered.

[0028] S31 includes: S311, during the nutrient pump infusion process, based on a preset time step. Collect real-time operational behavior data and path status data of the current path to construct a time-series state vector of the path. And form a path behavior sequence at continuous time steps. , represented as: ; ; in, This is the pressure feedback value at that point in time. The residual liquid flow rate, This is the channel response value. For drug liquid identification code, This represents the total number of sampling time steps within the current infusion cycle. S312, path behavior sequence Misconnection identification rule set Perform a comparison; S313, combine all rule matching results to form the rule matching vector for the current path. If it exists If the current path matches at least one misconnection rule, then the path is marked as a suspected misconnection path. For rules The matching results.

[0029] S32 includes: S321, Calculate the risk score based on the target path misconnection risk prediction model. And set a threshold for determining the risk of misconnection of a path for risk assessment. If it is, then it is determined to be a high-risk path, among which, The threshold for determining the risk of misconnection of a route; ; in, This indicates that the first step is to compute the first step for a given sample set. quantile (set as) 0.75), For the reason A scoring sequence consisting of risk scores calculated from known misconnected path samples; S322: When the score determines that the path is high-risk, the medication path alarm signal is immediately triggered, and the prompting mechanism is activated through the nutrition pump control interface, including a buzzer, screen flashing, and status lights.

[0030] like Figure 2 As shown, a nutrition pump medication route verification system based on a misconnection identification mechanism is used to implement the aforementioned nutrition pump medication route verification method based on a misconnection identification mechanism, and includes the following modules: Data acquisition module: Collects operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. Operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. Path status data includes channel response sequence and drug identification code information, and constructs an initial path behavior map. Anomaly identification module: Based on the initial path behavior map, it identifies misconnection path behavior units that exhibit abnormal channel response characteristics and undesirable drug liquid identification combinations, constructs a misconnection identification rule set, and generates a target path misconnection risk prediction model by combining it with the historical path cross-record library. Path verification and alarm module: The target path misconnection risk prediction model is applied to the real-time infusion path of the current nutrition pump to dynamically verify whether the path behavior matches the misconnection identification rule set. If it is judged to be a high-risk path, the medication path alarm signal is triggered.

[0031] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0032] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for verifying the medication delivery pathway of a nutrient pump based on a misconnection identification mechanism, characterized in that, Includes the following steps: S1. Collect operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. The operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. The path status data includes channel response sequence and drug identification code information when the path is connected. Construct an initial path behavior map to characterize abnormal path features. S2, based on the initial path behavior map, identify the misconnection path behavior units, extract the path data that show abnormal channel response characteristics and undesirable drug liquid identification combinations, construct the misconnection identification rule set, and combine it with the historical path cross record library to generate the target path misconnection risk prediction model. S3. Apply the target path misconnection risk prediction model to the real-time infusion path of the current nutrient pump and perform dynamic verification on each infusion path. If the path behavior characteristics match the misconnection identification rule set and the target path misconnection risk prediction model judges it as a high-risk path, then trigger the medication path alarm signal.

2. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 1, characterized in that, S1 includes: S11. Before and after infusion initiation, multi-point sampling is performed during key time periods to collect operational data of the nutrient pump, including start-up delay time. Pressure feedback change curve Residual liquid flow velocity sequence ; S12, After the path is successfully connected, collect path status data, including the channel response sequence. Medicine identification code ; S13, Unify and align the collected operational behavior data and path status data to construct an initial path behavior map for characterizing path anomaly features. .

3. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 2, characterized in that, S13 includes: S131, Time axis normalization is performed on the operational behavior data and path status data, with the unified benchmark being the pump start trigger time. Reconstruct the relative timeline; S132 uses linear interpolation to uniformly map runtime behavior data to a fixed time step. At equally spaced time points; S133, for each alignment time point Above, extract the current estimated values ​​of all operational behavior data and path status data channels to construct a path behavior status vector. ; S134, the obtained multiple path behavior state vectors The data is arranged in chronological order to form a path behavior data sequence. ; S135, construct the path behavior state sequence as a graph structure, which consists of nodes and edges, represented as follows: ； in, This forms the initial graph for path behavior. Each node in the graph represents the path behavior state at a given point in time. For graph edges, represent the continuous evolution of path states over time.

4. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 3, characterized in that, S2 includes: S21. From the initial path behavior map, select nodes that meet the criteria of channel response change exceeding a set threshold or drug identification code not matching the current prescription as abnormal candidate nodes. Then, gather the abnormal candidate nodes and their temporally continuous adjacent nodes in the initial path behavior map to form a misconnected path behavior unit, and extract its channel response change pattern and drug identification code combination. S22. Based on the extracted misconnection path behavior units, abnormal features are summarized, a misconnection identification rule set is constructed, and historical path cross-records are combined to generate a target path misconnection risk prediction model.

5. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 4, characterized in that, S21 includes: S211, in the initial path behavior graph In, for each node Perform anomaly detection; if the node channel response variation amplitude Exceeding the set anomaly detection threshold or its corresponding drug identification code Not part of the set of identifiers allowed by the current prescription That is, satisfying or Then the node Nodes identified as anomalous candidate nodes are grouped together with their temporally consecutive adjacent nodes in the initial path behavior graph to form misconnection path behavior units. For each misconnection path behavior unit, its corresponding channel response change pattern is extracted. and the combination of drug liquid identification codes involved. , as a feature sample of misconnection; S212, for each abnormal candidate node Find temporally consecutive neighboring nodes in the initial path behavior graph. and If an adjacent node exists in the initial path behavior graph, then it is joined with... These are grouped together into a single misconnection path behavior unit, forming a candidate set of misconnection behaviors. ; S213, for each misconnection path behavior unit Extract the channel response change patterns respectively Combination with drug liquid identification code Finally, the feature set of misconnection behavior is obtained. .

6. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 5, characterized in that, S22 includes: S221, misconnection path behavior unit Extracted channel response change patterns Combination with drug liquid identification code As input data, the K-Means clustering algorithm is applied to cluster misconnection behaviors, identifying different types of misconnection patterns and forming a set of misconnection behavior patterns. ; S222, based on the set of misconnection behavior patterns By summarizing the common characteristics within each cluster, a set of rules for misconnection identification is formed. Each rule Represented as: ； in, For the first The range of channel response modes corresponding to each rule. For the first The combination of illegal drug liquid identifiers corresponding to the rule; S223, Introducing a historical path cross-recording library The system combines the current path behavior state sequence to construct a rule matching vector, and generates a target path misconnection risk prediction model based on this vector, which is used to assess the misconnection risk of the current path.

7. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 6, characterized in that, S3 includes: S31. During the nutrient pump infusion process, the behavior state sequence of the current path is acquired in real time and matched with the pre-built misconnection identification rule set. For each path, based on whether it meets the channel response change mode range and illegal drug liquid identification combination in the rule conditions, a corresponding rule matching vector is generated to mark whether the current path has structural features of potential misconnection behavior. If it meets at least one rule, it is marked as a suspected misconnection path. S32, for suspected misconnected paths, call the target path misconnection risk prediction model to calculate the risk score. If the risk score is not lower than the set path misconnection risk score judgment threshold, it is judged as a high-risk path and the medication path alarm signal is immediately triggered.

8. The method for verifying the medication delivery route of a nutrient pump based on a misconnection identification mechanism according to claim 7, characterized in that, S31 includes: S311, during the nutrient pump infusion process, based on a preset time step. Collect real-time operational behavior data and path status data of the current path to construct a time-series state vector of the path. And form a path behavior sequence at continuous time steps. ; S312, path behavior sequence Misconnection identification rule set Perform a comparison; S313, combine all rule matching results to form the rule matching vector for the current path. If it exists If the current path matches at least one misconnection rule, then the path is marked as a suspected misconnection path. For rules The matching results.

9. A method for verifying the medication delivery pathway of a nutrient pump based on a misconnection identification mechanism according to claim 8, characterized in that, S32 includes: S321, Calculate the risk score based on the target path misconnection risk prediction model. And set a threshold for determining the risk of misconnection of a path for risk assessment. If it is, then it is determined to be a high-risk path, among which, The threshold for determining the risk of misconnection of a route; S322: When the score determines that the path is high-risk, the medication path alarm signal is immediately triggered, and the prompting mechanism is activated through the nutrition pump control interface, including a buzzer, screen flashing, and status lights.

10. A nutrition pump medication route verification system based on a misconnection identification mechanism, used to implement the nutrition pump medication route verification method based on a misconnection identification mechanism as described in any one of claims 1-9, characterized in that, Includes the following modules: Data acquisition module: Collects operational behavior data and path status data of the nutrient pump at multiple key time periods before and after infusion start-up. The operational behavior data includes pump start-up delay, pressure feedback changes and residual liquid flow behavior. The path status data includes channel response sequence and drug identification code information, and constructs an initial path behavior map. Anomaly identification module: Based on the initial path behavior map, it identifies misconnection path behavior units that exhibit abnormal channel response characteristics and undesirable drug liquid identification combinations, constructs a misconnection identification rule set, and generates a target path misconnection risk prediction model by combining it with the historical path cross-record library. Path verification and alarm module: The target path misconnection risk prediction model is applied to the real-time infusion path of the current nutrient pump to dynamically verify whether the path behavior matches the misconnection identification rule set. If it is judged to be a high-risk path, a medication path alarm signal is triggered.

Images