Medical instrument data informatization monitoring method and system

Abstract

The invention discloses a medical instrument data informatization monitoring method and system, and particularly relates to the technical field of medical instrument data informatization monitoring, and the method comprises the steps: obtaining an execution request of a medical instrument calling task, and extracting a preset operation node sequence, an execution sequence, a time constraint parameter and a behavior chain structure corresponding to the medical instrument calling task based on the execution request, constructing flow structure information of the medical instrument calling task; a hidden track is constructed based on the process structure information corresponding to the medical instrument calling task, non-interactive marks are set for nodes of the hidden track, no equipment instruction is triggered, and user interface display is not performed; abnormal adjustment, leapfrogging execution and human intervention operation in the task execution process are recognized by constructing a standard hidden track and a real task track, executing path structure comparison and behavior consistency scoring.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical device data informatization monitoring, more particularly, the present application relates to a medical device data informatization monitoring method and system. BACKGROUND

[0002] The current medical device data information monitoring method generally relies on the execution state reporting, log recording and alarm clearing identification of the equipment to determine whether the task is completed. As long as the system receives the task completion signal returned by the equipment and forms the corresponding operation record in the database, the system considers that the task has been normally executed. This determination mechanism is only based on the existence of surface data and time closed loop, ignoring the problems of abnormal adjustment, step execution or human intervention in the operation process.

[0003] In actual clinical scenarios, medical staff can make the task flow appear as a complete closed loop on the data by adjusting the operation sequence, delaying execution, skipping confirmation or supplementing operation, etc. However, in essence, the task has not been truly completed. For example, when the equipment performs abnormally or the alarm is not handled in time, the operator can make the system mistakenly believe that the task has been successfully completed by temporarily powering off, restarting the equipment or bypassing part of the confirmation link. Since the system's judgment standard for task completion does not include verification of operation rhythm, behavior logic and response rationality, as long as the final record chain is closed, it is assumed that the task execution is valid. The system cannot identify whether there is human intervention or operation deviation.

[0004] This processing logic hides the real execution situation of the task status, making some incomplete or abnormally executed tasks incorrectly marked as normally completed in the system, which seriously affects the authenticity and traceability of the monitoring data. In scenarios where equipment is frequently used and personnel operate intensively, such problems are more likely to occur. Since the system lacks a behavior discrimination mechanism, it cannot actively identify the source of the risk, and thus the key abnormalities in the execution process are ignored for a long time.

[0005] Therefore, the current medical device data information monitoring method lacks the ability to verify the rationality of the operation process and the credibility of the data record in the task completion status determination mechanism. The system can only determine whether the data exists, but cannot determine whether the record truly reflects the device behavior and actual execution situation. This problem makes the monitoring data appear complete in form, but in essence, there is a structural distortion that deviates from the actual situation, which has become the most core risk gap in the current medical device data informatization monitoring. SUMMARY

[0006] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a medical device data informatization monitoring method and system, which identifies abnormal adjustment, step execution and human intervention operation in the task execution process by constructing a standard hidden track and a real task track, performing path structure comparison and behavior consistency scoring.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for information-based monitoring of pharmaceutical and medical device data, comprising:

[0008] S1: Obtain the execution request of the drug and medical device requisition task, extract the preset operation node sequence, execution order, time constraint parameters and behavior chain structure corresponding to the drug and medical device requisition task based on the execution request, and construct the process structure information of the drug and medical device requisition task;

[0009] S2: Construct a hidden trajectory based on the process structure information corresponding to the drug and medical device call task, set non-interactive markers for the nodes of the hidden trajectory, do not trigger any device commands, and do not display the user interface;

[0010] S3: During the execution of the drug and medical device requisition task, the actual instruction issuance time, equipment feedback time, task status update information, and operation node timing are collected as behavior discrimination data. Based on the collected behavior discrimination data, the process structure information of the actual execution process of the drug and medical device requisition task is restored, and the real task trajectory is constructed.

[0011] S4: Perform a structural comparison between the real task trajectory and the corresponding hidden trajectory. The structural comparison includes node sequence consistency analysis, response time delay analysis, and path integrity analysis. During the structural comparison, extract three types of deviation features: sequence offset, time offset, and path missing. Calculate the trajectory consistency score based on the three types of deviation features.

[0012] S5: Compare the trajectory consistency score with the corresponding preset threshold, and perform a joint judgment based on the three types of auxiliary indicators. If the trajectory consistency score is lower than the credibility threshold and at least one type of auxiliary indicator shows structural abnormality, the drug and medical device dispatch task is judged as an abnormal task; otherwise, it is judged as a normal task.

[0013] In a preferred embodiment, in S1, a task identifier field is identified from the execution request of the drug and medical device requisition task, and the structure template data corresponding to the task identifier field is called. The structure template data includes a preset operation node sequence, a preset execution order, time constraint parameters and a behavior chain structure associated with the corresponding drug and medical device requisition task.

[0014] The preset operation nodes in the structure template data are sorted according to the preset execution order to form a node chain arranged in chronological order;

[0015] Based on the time constraint parameter, time window information is configured for each operation node in the node chain, the time window information includes the start time and completion time interval allowed by the node, the logical dependency relationship between nodes is identified in combination with the behavior chain structure, the corresponding time constraint parameter and dependent node identification information are written in the data structure of each operation node, and the process structure information with timing attribute and dependency structure is generated.

[0016] In a preferred embodiment, in S2, based on the node chain constructed in the process structure information of the drug and instrument calling task, each node is connected in the preset execution order to form a hidden track consistent with the process structure of the corresponding drug and instrument calling task;

[0017] The interactive attribute of each operation node in the hidden track is written, specifically including: adding a state control field area in the data structure corresponding to the operation node, writing an interactive state field in the added state control field area, and setting the interactive state field to a disabled state during writing, to generate a hidden track with interactive disabled attribute.

[0018] In a preferred embodiment, in S3, during the execution of the drug and instrument calling task, task control events triggered by operation instructions are continuously recorded, and behavior judgment data including instruction issuing time, device feedback time, task state update information and operation node execution timing are collected from the task control events;

[0019] Based on the task state update information and operation node execution timing in the behavior judgment data, the logical dependency relationship between each operation node is identified, the corresponding timestamp information and dependent node identification information are written in the data structure of each operation node, and the process structure information consistent with the actual execution process is generated;

[0020] The operation nodes in the actual process structure information are connected in turn according to the execution time sequence to form a real task track.

[0021] In a preferred embodiment, in S4, the real task track is compared with the corresponding hidden track in structure, and the structure comparison includes node sequence consistency analysis, response time delay analysis and path integrity analysis;

[0022] By positioning the operation nodes with consistent identification in the two tracks, the execution position index of each pair of operation nodes in their respective tracks is obtained, and the difference value of the execution position index is calculated to obtain the execution displacement of each pair of operation nodes in the real task track relative to the hidden track, and the node identification, displacement direction and displacement value information of all nodes with non-zero displacement are extracted as the sequence offset feature representing the sequence misplacement of the tracks;

[0023] The response time of each operation node in the two trajectories is calculated respectively, and the response time of the same identification node is calculated by difference, to obtain the response time difference of each node. The nodes with response time difference exceeding the preset tolerance threshold are summarized as an abnormal node set, and the response time difference value and the execution position in the task trajectory of each abnormal node are extracted as the time offset feature constituting the response time anomaly;

[0024] The preset operation nodes in the hidden trajectory are compared with the executed operation nodes in the real task trajectory, and the nodes in the hidden trajectory that do not appear in the real task trajectory are identified. The node identification, preset trigger position and preset execution time information in the non-appearing nodes are extracted as the path missing feature constituting the path interruption.

[0025] In a preferred embodiment, S4 further includes integrating the three types of deviation features, i.e., the sequence offset feature, the time offset feature and the path missing feature, into a structural deviation feature set. The number of operation nodes involved in each type of deviation feature is counted, and the proportion of the number of operation nodes involved in each type of deviation feature to the total number of nodes in the hidden trajectory is calculated to generate the sequence offset rate, the time offset rate and the path missing rate.

[0026] The sequence offset rate, the time offset rate and the path missing rate are respectively multiplied by the corresponding offset weight to obtain three types of weighted offset values. The sum of the three types of weighted offset values is calculated to generate a total offset value. Based on the full score, the total offset value is deducted from the full score by subtraction operation, and the trajectory consistency score is output.

[0027] In a preferred embodiment, in S5, the three types of auxiliary indicators are whether the number of sequence offset nodes is greater than or equal to two, whether the number of response delay abnormal nodes is greater than or equal to three, and whether the number of path interruption nodes is greater than or equal to one. If the respective determination conditions are met, the corresponding auxiliary indicators are determined to have the structural abnormality feature.

[0028] A drug and medical device data information monitoring system, comprising a structure construction module, a trajectory generation module, an extraction module, a structure comparison module and a determination module;

[0029] The structure construction module is used to obtain an execution request of a drug and medical device calling task, extract a preset operation node sequence, an execution sequence, a time constraint parameter and a behavior chain structure corresponding to the drug and medical device calling task based on the execution request, and construct process structure information of the drug and medical device calling task.

[0030] The trajectory generation module constructs a hidden trajectory based on the process structure information corresponding to the drug and medical device calling task. The nodes of the hidden trajectory are marked as non-interactive, do not trigger any device instructions, and do not display the user interface.

[0031] The extraction module is used to collect the instruction issuing time, device feedback time, task state update information and operation node timing actually occurring in the process of the drug equipment calling task execution as behavior discrimination data, and to restore the process structure information of the actual execution process of the drug equipment calling task based on the collected behavior discrimination data, and to construct a real task trajectory;

[0032] The structure comparison module is used to compare the structure of the real task trajectory with the corresponding hidden trajectory, which includes node sequence consistency analysis, response time delay analysis and path integrity analysis, extracts three types of deviation features of sequence offset, time offset and path loss, and calculates the trajectory consistency score based on the three types of deviation features;

[0033] The determination module is used to compare the trajectory consistency score with the corresponding preset threshold, and to perform joint determination combined with the three types of auxiliary indexes, if the trajectory consistency score is lower than the credibility threshold and at least one type of auxiliary index appears structural abnormal feature, the drug equipment calling task is determined as an abnormal task, otherwise it is determined as a normal task.

[0034] The technical effects and advantages of the present application are:

[0035] 1. The present application constructs hidden trajectory and real task trajectory and performs multi-dimensional structure comparison to identify abnormal adjustment and human intervention in the drug equipment calling task, solving the problem that the existing system cannot verify the authenticity of task execution;

[0036] 2. The standard trajectory and real trajectory are constructed based on the process structure information, and the structure is restored combined with the operation sequence, time and node dependence to realize path-level tracing of the task execution process;

[0037] 3. The three types of deviation features of sequence, time and path are extracted and quantified as offset rate, and the trajectory consistency score is calculated to accurately evaluate the overall execution deviation of the real trajectory relative to the standard trajectory;

[0038] 4. Combined with the joint determination of the three types of auxiliary indexes, the sensitivity and robustness of abnormal task recognition are improved, single misjudgment of score is avoided, and the monitoring ability for complex deviation is improved. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 The method step framework flowchart of the present application;

[0040] Figure 2 The system module structure diagram of the present application;

[0041] Figure 3 The task structure extraction flowchart of the present application;

[0042] Figure 4 The real trajectory construction and structure comparison flowchart of the present application;

[0043] Figure 5 Abnormality determination flowchart of the present application. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0045] With reference to the drawings in the description Figures 1-5 A drug and medical instrument data informationization monitoring method according to an embodiment of the present application comprises:

[0046] S1: An execution request of a drug and medical instrument calling task is acquired, a preset operation node sequence, an execution order, a time constraint parameter and a behavior chain structure corresponding to the drug and medical instrument calling task are extracted based on the execution request, a flow structure information of the drug and medical instrument calling task is constructed, and the flow structure information is taken as an input basis for drug and medical instrument task monitoring;

[0047] S2: A hidden track is constructed based on the flow structure information corresponding to the drug and medical instrument calling task, a non-interactive label is set for a node of the hidden track, no device instruction is triggered, and no user interface is displayed, the hidden track is taken as a standardized reference path, and is used to ensure that the drug and medical instrument calling task cannot be accessed, modified or intervened during an execution process of the drug and medical instrument calling task;

[0048] S3: During the execution process of the drug and medical instrument calling task, an instruction issuing time actually occurring, a device feedback time, task state update information and an operation node time sequence are collected as behavior discrimination data, flow structure information of an actual execution process of the drug and medical instrument calling task is restored based on the collected behavior discrimination data, a real task track is constructed, and the real task track is used to represent an execution path of the drug and medical instrument calling task in actual running;

[0049] S4: The real task track is compared with the corresponding hidden track in structure, the structure comparison includes node order consistency analysis, response time delay analysis and path integrity analysis, three types of deviation features of sequence offset, time offset and path loss are extracted in the structure comparison process, and a track consistency score is calculated based on the three types of deviation features, and the track consistency score is used to quantify an execution offset degree between the real task track and the hidden track;

[0050] S5: comparing the trajectory consistency score with a corresponding preset threshold value, performing joint determination in combination with three types of auxiliary indicators, if the trajectory consistency score is lower than the credibility threshold value and at least any one of the three types of auxiliary indicators has a structural abnormality feature, determining that the drug and medical instrument calling task is an abnormal task, otherwise determining that it is a normal task, the three types of auxiliary indicators including whether the number of sequence offset nodes is greater than or equal to two, whether the number of response delay abnormal nodes is greater than or equal to three, and whether the number of path interruption nodes is greater than or equal to one.

[0051] In S1, a task identification field is identified from an execution request of a drug and medical instrument calling task, a structure template data corresponding to the task identification field is called, and the structure template data includes a preset operation node sequence, a preset execution sequence, a time constraint parameter, and a behavior chain structure associated with a corresponding drug and medical instrument calling task;

[0052] The preset operation nodes in the structure template data are sorted according to the preset execution sequence to form a node chain arranged in time sequence, and the node chain is used to construct a time sequence basis of the process structure information;

[0053] Based on the time constraint parameter, a time window information is configured for each operation node in the node chain, the time window information includes a start time and a completion time interval allowed by the node, a logical dependency relationship between nodes is identified in combination with the behavior chain structure, the corresponding time constraint parameter and the dependent node identification information are written in the data structure of each operation node, and process structure information with time sequence attributes and dependency structure is generated, which is used to support the structural mapping and comparison processing of the hidden trajectory and the real task trajectory of the drug and medical instrument calling task.

[0054] In S2, based on the node chain constructed in the process structure information of the drug and medical instrument calling task, the nodes are connected according to the preset execution sequence to form a hidden trajectory consistent with the process structure of the corresponding drug and medical instrument calling task, and the hidden trajectory is used as a standardized reference path of the task execution process;

[0055] An interaction attribute is written for each operation node in the hidden trajectory, specifically including: adding a state control field area in the data structure corresponding to the operation node, writing an interaction state field in the added state control field area, and setting the interaction state field to a disabled state during the writing process, completing the interaction attribute writing, and generating a hidden trajectory with an interaction disabled attribute, the hidden trajectory with the interaction disabled attribute is used to prevent the operation node from being accessed, triggered or displayed during the execution process of the drug and medical instrument calling task.

[0056] In S3, during the execution process of the drug and medical instrument calling task, task control events generated by operation instructions are continuously recorded, and behavior judgment data including instruction issuing time, device feedback time, task state update information, and operation node execution time sequence are collected from the task control events;

[0057] Based on the task state update information in the behavior discrimination data and the operation node execution time sequence, the logical dependency relationship between each operation node is identified, the corresponding timestamp information and dependent node identification information are written in the data structure of each operation node, and the actual process structure information is generated, which is used to represent the path structure and control logic of the drug and instrument calling task in the actual execution process.

[0058] The operation nodes in the actual process structure information are connected in turn according to the execution time sequence to form a real task trajectory, which is used to represent the complete operation path formed by the drug and instrument calling task in the actual execution process, and serves as the basis for structural comparison with the hidden trajectory.

[0059] In S4, the real task trajectory is compared with the corresponding hidden trajectory, and the structural comparison includes node sequence consistency analysis, response time delay analysis and path integrity analysis.

[0060] By locating the operation nodes with the same identification in the two trajectories, the execution position index of each pair of operation nodes in their respective trajectories is obtained, and the difference value of the execution position index is calculated to obtain the execution displacement of each pair of operation nodes in the real task trajectory relative to the hidden trajectory. The node identification, displacement direction and displacement value information of all nodes with non-zero displacement are extracted as the sequence offset feature representing the sequence misplacement of the trajectory.

[0061] The response time of each operation node in the two trajectories is calculated, and the response time of the nodes with the same identification is calculated. The response time difference of each node is obtained, and the nodes with response time difference exceeding the preset tolerance threshold are summarized as an abnormal node set, and the response time difference value and the execution position in the task trajectory of each abnormal node are extracted as the time offset feature constituting the response time abnormality.

[0062] The set comparison is performed between all preset operation nodes in the hidden trajectory and the executed operation nodes in the real task trajectory, the nodes in the hidden trajectory that do not appear in the real task trajectory are identified, and the node identification, preset trigger position and preset execution time information of the non-appeared nodes are extracted as the path missing feature constituting the path interruption.

[0063] It should be noted that for the formula structure involved in the present scheme, the dimensionless term can be used as a proportional or structural adjustment factor. When combined with quantities with units, it only plays a numerical scaling role and does not introduce new physical dimensions, so it will not change or confuse the unit system of the whole expression. Such combination of "dimensionless term and quantity unit term" can be understood as the complex structure expression form commonly used in mathematical and physical modeling, which meets the principle of dimensional consistency and has a clear physical interpretation basis.

[0064] Secondly, in the formula structure of the present scheme, if multiple variable terms with different physical units are involved, including but not limited to time, mass or energy variables, their joint occurrence is to express the cooperative modeling relationship of multiple physical mechanisms. Each variable can be mapped by a function, combined by a ratio or adjusted by a normalization to form a unified structure with clear units and explicit meaning, and the overall expression conforms to the principle of dimensional consistency and the general norm of engineering modeling.

[0065] In the present scheme, constants, weights, adjustment factors, threshold parameters, proportionality coefficients, etc. are all adjustable control parameters for different application environments. Their values depend on the target device configuration, data input characteristics and performance optimization goals, and are set within a reasonable range through model verification, performance constraints or engineering calibration in the implementation stage. Although these parameters do not have a unique value, they have clear adjustment logic and calculation path, and belong to the deterministic setting process in engineering implementation. The purpose of such setting is to ensure that the scheme has both general adaptability and reproducibility and operability, without affecting its technical clarity and implementability.

[0066] In S4, the sequence offset feature, the time offset feature and the path missing feature are integrated into a set of structural deviation features. The number of operation nodes involved in each type of deviation feature is counted, and the proportion of the number of operation nodes involved in each type of deviation feature to the total number of hidden trajectory nodes is calculated to generate the sequence offset rate, the time offset rate and the path missing rate.

[0067] The sequence offset rate, the time offset rate and the path missing rate are respectively multiplied by the corresponding offset weight to obtain three types of weighted offset values. The total offset value is obtained by summing the three types of weighted offset values. Based on the full score, the total offset value is deducted from the full score by subtraction operation to output the trajectory consistency score. The trajectory consistency score is used to quantify the overall consistency of the true task trajectory of the drug and equipment calling task with respect to the hidden trajectory in the three dimensions of structural sequence, response time and path integrity.

[0068] The trajectory consistency score Ψ is defined as:

[0069]

[0070]

[0071] where Φ (s) represents the sequence offset response feature function; Φ (t) represents the response time offset feature function; Φ m) represents the path missing feature function; and the symbol Ξ(·) represents the offset response fusion operator, which is used for nested aggregation of three types of offset values; Ξ(·) represents the trajectory offset integration function, which is used to merge the three types of offset response into a unified offset function; 1-Ξ(·) represents the inverse mapping score in the form of subtraction operation, which is used to deduct the offset degree from the consistency of the full score.

[0072] Furthermore, in equation Ψ, the sequence offset function Φ is integrated through the trajectory offset function. (s) Response time offset function Φ (t) With path missing function Φ (m) The total offset value is generated by fusion, and then the trajectory consistency score Ψ is obtained by subtraction.

[0073] I represents the set of sequentially offset nodes, which indicates the set of node indices that are misaligned at the execution location; π r (i) represents the execution order index of the sequential offset node i in the actual task trajectory, numbered in ascending order according to the task occurrence time; π h (i) represents the execution order index of the sequential offset node i in the hidden trajectory, sorted according to the preset path; For the set of operation nodes in the hidden trajectory, Indicates the standard execution path; The total number of hidden trajectory nodes is used to normalize the order and time offset; χ(·) represents the order difference normalization function, which is defined as a monotonically increasing function; δ i δ represents the path influence depth of the sequential offset node i. i This indicates the topological distribution strength of the edges connecting the node before and after it, and is used to measure the disturbance propagation capability. The path graph structure representing the hidden trajectory; The path importance response function of the sequential offset node i in the path graph structure of the hidden trajectory;

[0074] Furthermore, in Φ (s) In the formula, the execution order difference of nodes with the same identifier in the real trajectory and the hidden trajectory is calculated by the order difference normalization function χ, and then the path graph structure of the node in the hidden trajectory is introduced by the path importance response function Λ. The structural sensitivity of the structure is analyzed, and the influence of nodal displacement intensity and structural influence is integrated to generate the sequential offset response characteristic function Φ(s).

[0075] τ represents the set of nodes with abnormal response times; r (j) represents the response time of the abnormal node j in the actual trajectory, i.e., the time difference between executing the instruction and receiving feedback from the device; τh (j) represents the standard response time of the response time abnormal node j in the hidden trajectory, i.e. the preset reference response time; ε(j) represents the preset response tolerance threshold of the response time abnormal node j; ω(·) represents a time offset normalization function, which is used for normalizing the response time error; δ j represents the path influence depth of the response time abnormal node j; represents the path importance response function of the response time abnormal node j in the path atlas structure of the hidden trajectory;

[0076] Further, in Φ t) , the normalized value of the response time difference of the corresponding node relative to the tolerance range is calculated by the time offset normalization function ω, and the time sequence dependence degree of the node in the trajectory structure is extracted in combination with the path importance response function Λ, to generate the response time offset feature function Φ (t) of the response time abnormality strength;

[0077] is a path missing node set, which represents the nodes in the hidden trajectory that do not appear in the real trajectory; ρ(k) represents the triggering position index of the path missing node k in the path atlas structure of the hidden trajectory, which is provided by ; is a node triggering position mapping function of the path atlas structure of the hidden trajectory, which reflects the logical starting point position of each node in the path; ξ(x, y) represents a path breakage response function, which is used to evaluate the structural risk brought by the path missing node x in the path atlas structure position y of the hidden trajectory; η(k) is a breakage importance function of the path missing node k, which represents the path completeness destruction degree caused by it in the path atlas structure of the hidden trajectory;

[0078] Further, in Φ (m) , the preset triggering position of the missing node is mapped to the path atlas structure of the hidden trajectory by the path breakage response function ξ , and the criticality of the missing node to the task path is extracted in combination with the breakage importance function η, to generate the path missing feature function Φ m) .

[0079] In S5, the three types of auxiliary indicators are whether the number of sequence offset nodes is greater than or equal to two, whether the number of response delay abnormal nodes is greater than or equal to three, and whether the number of path interruption nodes is greater than or equal to one, and if the respective determination conditions are met, the corresponding auxiliary indicators are determined to have the structural abnormality feature.

[0080] A drug and device data information monitoring system, comprising a structure construction module, a trajectory generation module, an extraction module, a structure comparison module, and a determination module.

[0081] The structure construction module is configured to obtain an execution request of a drug and equipment calling task, extract a preset operation node sequence, an execution sequence, a time constraint parameter and a behavior chain structure corresponding to the drug and equipment calling task based on the execution request, construct flow structure information of the drug and equipment calling task, and use the flow structure information as an input basis for monitoring of the drug and equipment task;

[0082] The trajectory generation module is configured to construct a hidden trajectory based on the flow structure information corresponding to the drug and equipment calling task, set a non-interactive mark for nodes of the hidden trajectory, not trigger any device instruction, and not display a user interface, so as to use the hidden trajectory as a standardized reference path for ensuring that the hidden trajectory cannot be accessed, modified or intervened in the execution process of the drug and equipment calling task.

[0083] The extraction module is configured to collect an actually occurring instruction issuing time, a device feedback time, task state update information and an operation node time sequence as behavior discrimination data in the execution process of the drug and equipment calling task, restore flow structure information of an actual execution process of the drug and equipment calling task based on the collected behavior discrimination data, and construct a real task trajectory, which is used to represent an execution path of the drug and equipment calling task in actual operation.

[0084] The structure comparison module is configured to compare the real task trajectory with the corresponding hidden trajectory, the comparison including node sequence consistency analysis, response time delay analysis and path integrity analysis, extract three types of deviation features including sequence offset, time offset and path loss, and calculate a trajectory consistency score based on the three types of deviation features, which is used to quantify an execution deviation degree between the real task trajectory and the hidden trajectory.

[0085] The determination module is configured to compare the trajectory consistency score with a corresponding preset threshold, perform joint determination in combination with three types of auxiliary indexes, determine that the drug and equipment calling task is an abnormal task if the trajectory consistency score is lower than a credibility threshold and at least one of the three types of auxiliary indexes has a structural abnormality feature, and otherwise, determine that the drug and equipment calling task is a normal task, the three types of auxiliary indexes including whether a sequence offset node number is greater than or equal to two, whether a response delay abnormal node number is greater than or equal to three, and whether a path interruption node number is greater than or equal to one.

[0086] It is necessary to make an overall description, including but not limited to: the present scheme is around the problem that state information cannot truly reflect operation behavior in the execution process of a drug and equipment calling task, proposes a realness verification mechanism based on structure flow and execution trajectory comparison, and aims to realize in-depth identification and deviation identification of task behavior level; the whole method is based on a behavior chain, constructs a structure flow, generates a standard trajectory, collects real behavior data, performs trajectory comparison and scoring, and finally realizes determination of an abnormal task;

[0087] The specific implementation process is as follows:

[0088] Firstly, the task identification field is identified, the associated process template is called, the operation node sequence, time sequence, time constraint parameter and behavior chain structure are extracted, and the process structure information with time sequence attribute and dependency structure is constructed, which is used to standardize the logical framework of all path generation; subsequently, the hidden track is constructed based on the process structure information, and the interaction disabling attribute is written in each operation node, so that the hidden track is inaccessible, triggerable and displayable during the task execution process, serving as a structural reference for comparison with the real task track;

[0089] During the actual task running process, the task control events triggered by the operation instructions are recorded, and the behavior judgment data is collected, including the instruction issuing time, device feedback time, task state update information and operation node execution time sequence; based on the behavior judgment data, the real task track is constructed, the dependency recognition and time sequence connection are performed, and the real path structure is generated, which is used to restore the actual task execution process;

[0090] The hidden track and the real task track are compared in structure, including three dimensions of sequence consistency analysis, response time delay analysis and path integrity analysis, the consistent operation nodes are located, the execution position indexes of each node in the two tracks are obtained, the execution displacement is calculated, the node identification, displacement direction and value of the existing position misplacement are extracted, and the sequence offset feature is generated; the response time difference of the operation node is calculated, the nodes exceeding the response time tolerance threshold are identified, the response time difference and execution position are extracted, and the time offset feature is generated; the preset operation nodes and the executed nodes are compared in set, the nodes not appearing in the hidden track are identified, the node identification, trigger position and execution time information are extracted, and the path missing feature is generated;

[0091] The three types of deviation features are integrated into a structural deviation set, and the proportion of the number of nodes involved in the sequence offset rate, the time offset rate and the path missing rate is calculated, which is formed by the total number of nodes in the hidden track; the three types of offset rates are multiplied by the corresponding operation weights respectively, summed to generate the total offset value, and then subtracted from the full score to generate the track consistency score, which is used to quantify the overall consistency degree of the real task track in the three dimensions of structural sequence, response time and path integrity;

[0092] Finally, the track consistency score is compared with the credibility threshold, and the joint determination is performed in combination with the three types of auxiliary features of the number of sequence offset nodes, the number of response abnormal nodes and the number of path interruption nodes; when the score is insufficient and any auxiliary feature meets the abnormal standard, the abnormal task judgment result is output, otherwise it is judged as a normal task; the whole method realizes the complete path from path generation to abnormal identification through the steps of structure construction, behavior collection, track reconstruction and score determination, has executability and high adaptability, and improves the accuracy and authenticity of the pharmaceutical and medical device task state monitoring.

[0093] The above merely provides the preferred embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall fall within the protection scope of the present application.

Claims

1. A method for monitoring the information of a drug and a device, characterized in that, Comprise: S1: Obtain the execution request of the drug and instrument calling task, extract the preset operation node sequence, execution order, time constraint parameter and behavior chain structure corresponding to the drug and instrument calling task based on the execution request, and construct the process structure information of the drug and instrument calling task; S2: Based on the process structure information of the drug and instrument calling task, a hidden track is constructed, the nodes of the hidden track are marked as non-interactive, no device instruction is triggered, and no user interface is displayed; S3: In the process of executing the drug and instrument calling task, the instruction issuing time, device feedback time, task state update information and operation node time sequence actually occurring are collected as behavior judgment data, the process structure information of the actual execution process of the drug and instrument calling task is restored based on the collected behavior judgment data, and a real task track is constructed; S4: The real task track is compared with the corresponding hidden track in structure, the structure comparison includes node sequence consistency analysis, response time delay analysis and path integrity analysis, three types of deviation features of sequence deviation, time deviation and path loss are extracted in the structure comparison process, and the track consistency score is calculated based on the three types of deviation features; S5: The track consistency score is compared with the corresponding preset threshold, and the joint determination is performed combined with three types of auxiliary indexes, if the track consistency score is lower than the credibility threshold and at least one type of auxiliary index appears structural abnormal feature, it is determined that the drug and instrument calling task is an abnormal task, otherwise it is determined as a normal task.

2. The drug and instrument data information monitoring method according to claim 1, wherein: In S1, the task identification field is identified from the execution request of the drug and instrument calling task, the structure template data corresponding to the task identification field is called, and the structure template data includes the preset operation node sequence, the preset execution order, the time constraint parameter and the behavior chain structure associated with the corresponding drug and instrument calling task; The preset operation nodes in the structure template data are sorted according to the preset execution order to form a node chain arranged in time sequence; Based on the time constraint parameter, time window information is configured for each operation node in the node chain, the time window information includes the starting time and completion time interval allowed by the node, the logical dependency relationship between nodes is identified combined with the behavior chain structure, the corresponding time constraint parameter and dependent node identification information are written in the data structure of each operation node, and the process structure information with time sequence attribute and dependency structure is generated.

3. The drug and instrument data information monitoring method according to claim 2, wherein: In S2, based on the node chain constructed in the process structure information of the drug and instrument calling task, the nodes are connected according to the preset execution order to form a hidden track consistent with the process structure of the corresponding drug and instrument calling task; The interactive attribute of each operation node in the hidden track is written, specifically including: adding a state control field area in the data structure corresponding to the operation node, writing an interactive state field in the added state control field area, and setting the interactive state field to disabled state during writing, to generate a hidden track with interactive disabled attribute.

4. The method according to claim 3, wherein: in S3, during the execution of the drug and medical device calling task, task control events triggered by operation instructions are continuously recorded, and behavior judgment data including instruction issuing time, device feedback time, task state update information and operation node execution time sequence are collected from the task control events; based on the task state update information and operation node execution time sequence in the behavior judgment data, the logical dependency relationship between each operation node is identified, and corresponding timestamp information and dependent node identification information are written in the data structure of each operation node to generate process structure information consistent with the actual execution process; the operation nodes in the actual process structure information are sequentially connected in the order of execution time to form a real task trajectory.

5. The method according to claim 4, wherein: in S4, the real task trajectory and the corresponding hidden trajectory are structurally compared, including node sequence consistency analysis, response time delay analysis and path integrity analysis; by locating the operation nodes with the same identification in the two trajectories, the execution position index of each pair of operation nodes in their respective trajectories is obtained, and the difference of the execution position index is calculated to obtain the execution displacement of each pair of operation nodes in the real task trajectory relative to the hidden trajectory, and the node identification, displacement direction and displacement value information of all nodes with non-zero displacement are extracted as sequence offset features representing the sequence dislocation of the trajectories; the response time of each operation node in the two trajectories is calculated, and the response time of the nodes with the same identification is calculated, and the response time difference of each node is obtained, and the nodes with the response time difference exceeding the preset tolerance threshold are summarized as an abnormal node set, and the response time difference and the execution position of each abnormal node in the task trajectory are extracted as time offset features constituting the response time abnormality; all preset operation nodes in the hidden trajectory and executed operation nodes in the real task trajectory are compared in set, and nodes in the hidden trajectory that do not appear in the real task trajectory are identified, and the node identification, preset trigger position and preset execution time information of the non-appearing nodes are extracted as path missing features constituting the path interruption.

6. The method according to claim 5, wherein: in S4, the sequence offset features, time offset features and path missing features are integrated into a structural deviation feature set, the number of operation nodes involved in each type of deviation feature is counted, and the proportion of the number of operation nodes involved in each type of deviation feature to the total number of nodes in the hidden trajectory is calculated to generate sequence offset rate, time offset rate and path missing rate; the sequence offset rate, time offset rate and path missing rate are multiplied by the corresponding offset weight to obtain three types of weighted offset values, and the sum of the three types of weighted offset values is calculated to generate a total offset value, and based on the full score, the total offset value is deducted from the full score by subtraction to output a trajectory consistency score.

7. The method according to claim 6, wherein: In S5, the three types of auxiliary indexes are whether the number of sequentially offset nodes is greater than or equal to two, whether the number of response delay abnormal nodes is greater than or equal to three, and whether the number of path interruption nodes is greater than or equal to one. If the respective determination conditions are met, the corresponding auxiliary indexes are determined to have structural abnormal characteristics.

8. A drug and equipment data informationization monitoring system, comprising a drug and equipment data informationization monitoring method according to claim 7, comprising a structure construction module, a trajectory generation module, an extraction module, a structure comparison module, and a determination module, characterized in that: The structure construction module is used to obtain an execution request of a drug and equipment calling task, extract a preset operation node sequence, an execution sequence, a time constraint parameter, and a behavior chain structure corresponding to the drug and equipment calling task based on the execution request, and construct process structure information of the drug and equipment calling task. The trajectory generation module constructs a hidden trajectory based on the process structure information corresponding to the drug and equipment calling task, sets a non-interactive mark for the nodes of the hidden trajectory, does not trigger any device instructions, and does not display a user interface. The extraction module is used to collect instruction issuance time, device feedback time, task state update information, and operation node timing actually occurring in the execution process of the drug and equipment calling task as behavior discrimination data, restore process structure information of the actual execution process of the drug and equipment calling task based on the collected behavior discrimination data, and construct a real task trajectory. The structure comparison module is used to compare the real task trajectory with the corresponding hidden trajectory, which includes node sequence consistency analysis, response time delay analysis, and path integrity analysis, extracts three types of deviation characteristics of sequence offset, time offset, and path loss during the structure comparison, and calculates a trajectory consistency score based on the three types of deviation characteristics. The determination module is used to compare the trajectory consistency score with a corresponding preset threshold, perform joint determination in combination with the three types of auxiliary indexes, and determine that the drug and equipment calling task is an abnormal task if the trajectory consistency score is lower than a credibility threshold and at least any one type of auxiliary index has a structural abnormal characteristic, otherwise, it is determined to be a normal task.

Images