A time axis key node automatic acquisition system for chest pain diagnosis and treatment

Abstract

The application discloses a time axis key node automatic acquisition system for chest pain diagnosis and treatment, relates to the field of information and communication technology of medical data, and comprises a pre-stored interval template containing starting or ending anchor point rules; an event stream carrying chest pain patient identification and source system time stamp derived from one or more medical information systems is acquired in real time; an active interval instance in memory is created and maintained for the chest pain patient; a candidate ending event is inferred; source system time reference drift is identified, and a reliability label is attached to time-consuming data. The application realizes automatic and real-time monitoring of the time consumption of the diagnosis and treatment process. The system not only supports starting and ending anchor point matching, automatically identifies time reference drift and attaches a reliability label, but also guarantees the reliability of time-consuming data in a real complex environment without relying on strict time synchronization of the whole hospital system.

Description

Technical Field

[0001] This invention relates to the field of information and communication technology for medical data, and in particular to an automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment. Background Technology

[0002] In the current medical information technology system, ensuring that every diagnosis and treatment action, such as registering for a patient with chest pain, issuing a medical order, and publishing an examination report, can be recorded completely and traceably as an atomic data event with a precise timestamp is a mature and necessary approach to ensuring the integrity of the original data.

[0003] However, when this technology is applied to clinical scenarios with extremely high requirements for treatment efficiency, such as acute chest pain, its inherent design principles and the core need for real-time monitoring of the treatment process in such scenarios become constraints. This is because the focus of quality control in such scenarios is no longer the isolated event itself, but the treatment interval time composed of the start and end points of multiple key events. Under the existing technology, the information system itself lacks the ability to identify the logic of the treatment interval; it can only process discrete events. This makes the time analysis of any treatment interval a mining task that is detached from the treatment process and requires manual intervention. Medical staff or quality control personnel must first use their professional knowledge to manually filter out specific start and end point events from the massive event streams in different systems, and then perform matching and calculation.

[0004] Currently, to improve this situation, the industry has attempted to simplify analysis by developing data query tools or data visualization panels. However, these improvements do not change the fact that the processing method is post-analysis, and cannot provide any real-time information about the progress of critical intervals during the treatment process. This lag in processing means that the role of information systems in time-sensitive treatment is always limited to passive data recording. Specifically, existing technologies have several shortcomings: the identification and calculation of treatment intervals rely on manual backtracking and matching by end users, which cannot achieve automation and real-time processing; the system lacks an inherent understanding of the process concept of treatment intervals, making it unable to track the real-time status of intervals during treatment; and due to the lag in its data processing method, the system cannot provide immediate warning information when critical treatment intervals are about to expire or have already expired. Summary of the Invention

[0005] In view of the problems existing in the existing automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment, this invention is proposed.

[0006] Therefore, the problem that this invention aims to solve is that existing medical information systems, due to their use of event atomic storage, cannot manage treatment intervals that have quality control significance, thus making it difficult to achieve automated and real-time monitoring of the efficiency of key treatment processes.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, embodiments of the present invention provide an automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment, comprising: a medical information system, including a system time (HIS), a source system time (LIS), and an electrocardiogram system; an interval definition library for pre-storing interval templates containing start or end anchor point rules, wherein the interval templates are configured to pre-store multiple chest pain-specific interval templates; an event monitoring module for real-time acquisition of event streams from one or more medical information systems carrying chest pain patient identifiers and source system timestamps; an interval instance management module for creating and maintaining an active interval instance in memory for a chest pain patient when an event matches the start anchor point rule of any chest pain-specific interval template based on the event stream; an adaptive correction module for handling anchor point mismatches and inferring candidate end events; and a time base compensation module for identifying source system time base drift and attaching a credibility marker to the time-consuming data.

[0009] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the chest pain-specific interval template in the interval definition library further includes an intermediate checkpoint anchor rule configured to define at least one intermediate checkpoint anchor rule between the start anchor rule and the end anchor rule.

[0010] The interval instance management module also includes being configured to continuously monitor and match the corresponding intermediate checkpoint anchor rules when an active interval instance is in the created state, and to preprocess the time consumption of the continuous sub-stages constituting the treatment interval based on the internal progress status.

[0011] Match the event stream with the end anchor rule of the active interval instance. When a match is successful, close the active interval instance and calculate the time consumed.

[0012] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the adaptive correction module includes: when the running time of any active interval instance exceeds a preset time and is still not closed, it automatically infers candidate termination events from the events related to chest pain patients that are not matched with any anchor point rules, which are filtered by the interval instance management module, based on the correlation rules, and closes the active interval instance through a single confirmation operation of a human-computer interaction interface module.

[0013] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the time reference compensation module includes continuous statistical analysis of the source system timestamps of the event stream, taking the source system of the event as the dimension, identifying drift anomalies in the source system time reference, and compensating for the time consumption or adding credibility markers based on the drift anomalies.

[0014] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the correlation rules include configuring the adaptive correction module, extracting the core semantic keywords of the end anchor point rules of active interval instances that have not been closed due to timeout, and filtering out all events that occurred after the creation of the active interval instance from events that have never matched any anchor point rules.

[0015] For each selected event, a pre-defined scoring model is used to calculate the event's relevance score. The event with the highest score is selected as the candidate termination event, and the event relevance score is... The calculation formula is:

[0016] ;

[0017] in, Match weights to preset keywords. A normalized text similarity value between 0 and 1 is calculated based on the descriptive information and core semantic keywords of the selected events. The preset time proximity weight, This is a normalized time proximity value between 0 and 1, calculated based on the time difference between the occurrence time and the active interval instances of the selected events.

[0018] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the interval definition library further includes a configuration for defining a chest pain-specific interval template and clinical logical relationship rules between it and one or more other chest pain-specific interval templates.

[0019] The clinical logical relationship rules include a clinical context arbitrator module, which includes configuration options for the clinical context arbitrator module.

[0020] When multiple active interval instances with logical relationship rules are created for the same chest pain patient, the visualization priority status of one or more other related active interval instances on the human-computer interaction interface module is dynamically adjusted according to the closure status of one of the active interval instances or a preset quality control status mark, following the logical relationship rules.

[0021] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the time reference compensation module further includes configuring the time reference compensation module to calculate and maintain the moving average and moving standard deviation of the source system timestamp intervals of adjacent events in the event stream in real time for each source system.

[0022] When the timestamp interval of a newly arrived event in the source system deviates from the moving average by more than a preset multiple of the moving standard deviation, it is determined that the time base of the source system has drifted abnormally.

[0023] The additional credibility markers include marking the time-consuming results of active interval instances closed by events originating from the anomalous source system on the human-computer interaction interface module as data source time bases to be verified.

[0024] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the human-computer interaction interface module includes a diagnosis and treatment interval dashboard, which is configured to provide an independent display unit for each type of quality control interval in the form of a graphical progress bar array.

[0025] Using a pre-defined visual coding system, the system can distinguish and display in real time the status of each active region instance, including those that have been created, are in time, are closed and meet the closure criteria, as well as those that have timed out and are set to low priority due to clinical logic arbitration.

[0026] As a preferred embodiment of the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment described in this invention, the event listening module includes a stateless event listener. The stateless event listener is configured to passively listen to the standardized medical information exchange protocol message bus to acquire event streams. After acquiring each event, it does not store any historical state information and distributes the metadata of the acquired events to the interval instance management module and the time base compensation module.

[0027] The interval definition library also includes a graphical configuration interface configured to be used by non-IT professionals.

[0028] The graphical configuration interface includes features configured to allow non-IT professionals to manipulate chest pain specific interval templates using a structured form.

[0029] Secondly, embodiments of the present invention provide an automatic acquisition method for key timeline nodes in chest pain diagnosis and treatment, comprising: pre-storing interval templates containing start or end anchor point rules, wherein the interval templates refer to multiple chest pain-specific interval templates configured to be pre-stored; acquiring in real time event streams from one or more medical information systems carrying chest pain patient identifiers and source system timestamps; based on the event stream, when an event matches the start anchor point rule of any chest pain-specific interval template, creating and maintaining an active interval instance in memory for the chest pain patient; handling anchor point mismatches and inferring candidate end events; identifying source system time base drift and attaching credibility markers to time-consuming data.

[0030] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any step of the above-mentioned automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment.

[0031] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any of the steps of the above-described automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment.

[0032] The beneficial effects of this invention are as follows: By constructing an interval template with key nodes in chest pain diagnosis and treatment as the core, this invention transforms multi-source medical event flows into active interval instances with lifecycles and states in real time, realizing automated and real-time monitoring of the time consumed in the diagnosis and treatment process. The system not only supports start and end anchor point matching, but also introduces intermediate checkpoints to realize sub-stage bottleneck early warning. It also dynamically adjusts the priority of parallel diagnosis and treatment paths through the clinical logical relationship between intervals. At the same time, based on the statistical analysis of event source timestamps, it automatically identifies time base drift and adds credibility markers, ensuring the reliability of time-consuming data in real and complex environments without relying on strict time synchronization of the entire hospital system. Attached Figure Description

[0033] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0034] Figure 1 A flowchart of an automatic timeline key node acquisition system for chest pain diagnosis and treatment provided in an embodiment of the present invention.

[0035] Figure 2This is a comparison diagram of time reference drift in an automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment, provided as an embodiment of the present invention. Detailed Implementation

[0036] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0037] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0038] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0039] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0040] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0041] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0042] Example 1

[0043] Reference Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides an automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment, comprising:

[0044] This invention first constructs an interval definition library, which is configured to provide a graphical configuration interface for non-IT professionals. This interface allows quality control personnel to define and manage the interval template name, quality control duration, and the event source system event type and event content keywords corresponding to the start and end anchor rules using a structured form. A specific interval template's data structure may include: interval ID, interval name, quality control threshold (e.g., 600 seconds), a start event anchor rule, and an end event anchor rule. The start event anchor rule is defined as a logical condition combination, such as (Event Source: HIS System AND Event Type: Patient Registration AND Registration Department: Emergency Department), and the end event anchor rule is also a similar logical condition combination, such as (Event Source: ECG System AND Event Type: Report Release). In this way, the clinical quality control focus intervals are established as a structured and manageable object within the information system.

[0045] To associate predefined interval templates with real-time diagnostic and treatment activities, the system is configured with an event listening module. In a preferred embodiment, this module is a stateless event listener that passively listens to standardized medical information exchange protocol message buses (such as HL7 or FHIR message streams) to acquire event streams from one or more medical information systems, carrying patient identifiers and source system timestamps. Upon acquiring each event, the stateless event listener does not store any historical state information but directly distributes the event's metadata (including patient identifier, event source system, source system timestamp, event type, and content description) to the interval instance management module and the time base compensation module. Upon receiving an event, the interval instance management module executes a two-way anchoring procedure: first, it matches the event with the starting anchor point rules of all interval templates in the interval definition library. For example, if a patient A's registration event occurs at 13:00, its metadata... When the start anchor point rule of the gate-to-ECG time interval template is met, the interval instance management module immediately creates an active interval instance corresponding to the template in memory for patient A. This instance is marked as created and in time, and its start time is recorded as 13:00. Secondly, for the same event, the interval instance management module also matches it with the end anchor point rule of all active interval instances created for that patient in memory. When an ECG report event related to patient A published at 13:05 is listened to, if its metadata meets the aforementioned end anchor point rule of the gate-to-ECG time active interval instance, the interval instance management module immediately closes the active interval instance, updates its status to closed, records the end time as 13:05, and calculates the interval duration as 300 seconds. This method of reconstructing discrete event streams into a series of parallel and manageable process objects transforms the monitoring of treatment efficiency from post-hoc analysis to real-time presentation.

[0046] In real-world medical information environments, differences in how different systems describe the same diagnostic and treatment behavior can lead to inaccurate matching of preset end-anchor rules, known as anchor mismatch. This prevents completed treatment intervals from closing within the system. To address this, the system employs an adaptive correction module. This module is triggered when the runtime of any active interval instance exceeds its defined quality control threshold in the interval template. Upon triggering, the module first extracts core semantic keywords from the end-anchor rules of the expired, unclosed active interval instance. For example, for the ECG report release anchor, the keyword set can be preset as ECG, ECG, and ECG monitoring. Simultaneously, the interval instance management module filters out all unanchored events related to the patient that occurred after the creation of the active interval instance. Finally, the adaptive correction module calculates the relevance score for each selected unanchored event using a preset scoring model.

[0047] :

[0048] in, To assign weight to keywords, The normalized text similarity value between 0 and 1 is calculated between the description information of the selected event and the aforementioned core semantic keywords. As the time proximity weight, The normalized time proximity value, between 0 and 1, is calculated based on the time difference between the occurrence time of the selected event and the creation time of the instance in the active region. It should be noted that the weight... and The settings can be adjusted based on the judgment of the main reasons for anchor point mismatch. If the mismatch is mainly caused by non-standard text descriptions, then the settings can be adjusted accordingly. Set to a higher value; an example setting is... ,and ; The similarity can be calculated using algorithms such as the Jaccard similarity coefficient or cosine similarity based on word vectors. The calculation can be achieved through an inverse proportional function, allowing events that are closer in time to each other to receive higher scores. For example, if a timeout instance starts at 13:00, with a quality control duration of 10 minutes, and the system triggers a correction mechanism at 13:11, with the keyword "ECG," there are currently events to be anchored: A (13:09, ECG test result) and B (13:10, troponin test). The description information for event A includes the keyword synonym "ECG." The value is higher, while event B's is higher. The value is 0, therefore event A will receive the highest relevance score. This is inferred as a candidate termination event; however, this candidate termination event does not directly close the interval, but is confirmed through the operation of a human-computer interaction module. For example, next to the corresponding timeout entry in a treatment interval dashboard, a suggestion is given: the test result (13:09) must be completed and confirmed by medical staff before it can be used to close the active interval instance; the correlation score in the adaptive correction module The computational model's internal parameters need to be solidified through a standardized offline calibration procedure; specifically, the normalized time proximity value... The computation function is determined to be:

[0049] ;

[0050] in, The time difference between the candidate termination event and the creation time of the active interval instance, and the baseline quality control threshold. The weighting parameters are determined by extracting all time-consuming samples of the same interval from the historical event stream data of the target hospital and taking the 95th percentile of their statistical distribution; subsequently, the weighting parameters are determined by... and The optimization process involves executing on a subset of cases selected from historical data that have known correct termination events but contain non-standard text descriptions. This is done by setting... The search space is [0.1, 0.9] with a step size of 0.1, and maintains... The constraints are used to iteratively calculate the factor that causes the percentage of correctly inferred cases to reach its peak. and The values ​​are then used as the final configuration for this deployment environment.

[0051] To improve the reliability of output time-consuming data, especially in environments where time bases in various business systems may drift, a time base compensation module is configured. This module performs continuous statistical analysis on the timestamps of the source systems in the event stream, using the source system as the dimension, to identify drift in the source system's time base. Specifically, for each source system, the module calculates and maintains in real time the moving average and moving standard deviation of the source system timestamp intervals of adjacent events in the event stream. When the value of the source system timestamp interval of a newly arriving event deviates from the moving average by more than a preset multiple (e.g., 3 times) of the moving standard deviation, it is determined that the time base of that source system has drifted abnormally. The source system is internally marked as having a time reference pending verification. Correspondingly, any instance of an active interval closed by events originating from this anomalous source system will have its calculated time consumption result appended with a credibility flag on the human-computer interface module, such as being marked as a data source time reference pending verification. Similarly, the threshold for judging drift anomalies used by the time reference compensation module, and the logical relationship rules upon which the clinical context arbitrator module is based, are also configured through preliminary data analysis. Specifically, for any source system, the threshold for judging time reference drift is calculated by extracting timestamp data from the system within a known stable operating cycle and calculating the standard deviation of the reference shift between adjacent events. And a standard deviation multiple that can cover 99.9% of the interval samples. and product The threshold is set; the logical relationship rules followed by the clinical context arbitrator are stored in a structured form in the interval definition library. Each rule consists of a trigger condition field, a target object field, and an execution action field. For example, the trigger condition can be defined as a specific interval instance closing with a specific result, the target object can be defined as all other active interval instances belonging to a certain diagnostic pathway, and the execution action is defined as adjusting its visual priority status to a preset value. Finally, the status of all active interval instances, including those created and timing, closed and meeting the target, closed and timed out, and those set to low priority due to clinical logic arbitration, are presented through a human-computer interaction interface module. This module can be specifically implemented as a diagnostic interval dashboard, providing an independent display unit for each type of quality control interval in the form of a graphical progress bar array, and using a preset visual encoding system (such as different colors or animation effects) to distinguish and display the status of each active interval instance in real time, thereby providing medical staff and quality control personnel with an immediate view of the progress status of key intervals.

[0052] In a preferred embodiment, an automatic acquisition method for key timeline nodes in chest pain diagnosis and treatment includes: pre-storing interval templates containing start or end anchor point rules, wherein the interval templates are configured to pre-store multiple chest pain-specific interval templates; acquiring in real time event streams from one or more medical information systems carrying chest pain patient identifiers and source system timestamps; based on the event stream, when an event matches the start anchor point rule of any chest pain-specific interval template, creating and maintaining an active interval instance in memory for the chest pain patient; handling anchor point mismatches and inferring candidate end events; identifying source system time base drift and attaching credibility markers to the time-consuming data.

[0053] In summary, this invention constructs an interval template centered on key nodes in chest pain diagnosis and treatment, transforming multi-source medical event flows into active interval instances with lifecycles and states in real time. This enables automated and real-time monitoring of the time consumed in the diagnosis and treatment process. The system not only supports start and end anchor point matching but also introduces intermediate checkpoints to achieve sub-stage bottleneck warnings. Furthermore, it dynamically adjusts the priority of parallel diagnosis and treatment paths through clinical logical relationships between intervals. Simultaneously, based on statistical analysis of event source timestamps, it automatically identifies time base drift and adds credibility markers, ensuring the reliability of time-consuming data in real and complex environments without relying on strict time synchronization across the entire hospital system.

[0054] To address the challenge of automatically constructing multiple structured timelines, dynamically adjusting path priorities based on clinical logic, and implementing bottleneck warnings at intermediate checkpoints when differentiating between acute ACS and PE in patients with atypical acute chest pain, this manual proposes the following implementation:

[0055] Example 2

[0056] Reference Figures 1-2 This is the second embodiment of the present invention. This embodiment provides a dual-path (ACS and PE) differential diagnosis and treatment scenario for patients with atypical acute chest pain. Specifically, it demonstrates how the system of the present invention automatically creates parallel active interval instances, dynamically adjusts priorities based on clinical logic rules, and implements sub-stage bottleneck warnings through intermediate checkpoints, thereby transforming discrete events into a structured, arbitrable, and warning-enabled real-time diagnosis and treatment process.

[0057] In the diagnosis and treatment process of a patient with acute chest pain, due to its atypical clinical presentation, the etiology needs to be differentiated simultaneously into acute coronary syndrome (ACS) and pulmonary embolism (PE). Therefore, the medical team needs to initiate and track multiple diagnosis and examination processes in parallel. However, with existing information technology support, data events from different business systems are presented in discrete, unrelated list formats, requiring medical staff to rely on a high level of cognitive load to manually integrate and compare the progress status and time nodes of these two parallel treatment paths. The system of this invention is deployed in this application scenario. After the patient completes registration at 12:00, event monitoring begins. The module captures the registration event and distributes it to the interval instance management module. Based on multiple preset interval templates related to the differential diagnosis of chest pain in the interval definition library, the interval instance management module creates and activates multiple active interval instances in memory for the patient simultaneously through matching the starting event anchor point rules. These include the door-to-ECG time interval instance and the troponin-outcome interval instance belonging to the ACS diagnostic pathway, and the D-dimer-outcome interval instance and the pulmonary artery CTA-report interval instance belonging to the PE diagnostic pathway. At this moment, on the treatment interval dashboard presented by the human-computer interaction module, these two sets of intervals, corresponding to ACS and PE respectively, are displayed. All interval entries were simultaneously activated and presented as created and timed, reflecting the parallel processing state in the early stages of diagnosis and treatment. At 12:45, a troponin-outcome event related to the patient, originating from the LIS system, was detected. This event carried a positive test result. Upon receiving this event, the interval instance management module, by ending the matching of the event anchor rule, set the state of the active troponin-outcome interval instance to closed. This state change was captured by the clinical context arbitrator module, which then queried the interval definition library and retrieved a preset logical relationship rule that defined the troponin-outcome interval instance. A positive closure state has a mutually exclusive relationship with all interval templates belonging to the PE diagnostic pathway. In this process, the structured state information output by the interval instance management module, i.e., a specific interval is closed with a specific result, becomes the direct input for the clinical context arbitrator module to perform logical judgment, enabling the functions of the two modules to work together. According to this rule, the clinical context arbitrator module sends a state modification instruction to all active interval instances currently activated for this patient and belonging to the PE diagnostic pathway, i.e., D-dimer-result interval instances and pulmonary artery CTA-report interval instances, to adjust their status from created and dynamically timed to low priority.

[0058] In the above scenario, after the troponin-outcome interval instance closes and the ACS diagnostic pathway is confirmed as high priority, a door-to-balloon dilation active interval instance with a total quality control time of 90 minutes continues to time. The interval template corresponding to this instance, in the interval definition library, has a preset intermediate checkpoint anchor rule for catheterization lab activation, in addition to the start and end anchor points, and a suggested time of 30 minutes is set for the sub-stage from patient registration to catheterization lab activation. In the subsequent progress of this scenario, when the system detects the catheterization lab activation event, the internal progress status of the door-to-balloon dilation active interval instance is updated, and its first sub-stage time is independently calculated to be 45 minutes, exceeding the suggested time of 30 minutes. The system then highlights the first segment of the instance's progress bar in yellow on the treatment interval dashboard as a warning. This provides the medical team with a proactive warning about internal process bottlenecks, even when the total interval is far from timed out. Upon receiving the status modification command, the real-time display interface of the treatment interval dashboard changes accordingly. The D-dimer-Results and Pulmonary Artery CTA-Report interval entries, which were previously in a created and timed state, are transformed into a grayed-out, static, low-priority state, while all interval entries related to the ACS diagnostic pathway continue to maintain a high-priority display. Through automatic arbitration and dynamic adjustment of the logical relationships between parallel treatment intervals, the system presents a structured information view that replaces the previous method of requiring medical staff to manually correlate and prioritize discrete events. This allows the medical team to immediately focus their attention on the currently confirmed treatment direction.

[0059] To verify the significant advantages of this invention over traditional technologies in anchor point mismatch correction, real-time monitoring, time base drift fault tolerance, and intelligent arbitration of parallel paths, and to demonstrate that it can more accurately, reliably, and efficiently achieve automated, intelligent, and real-time acquisition and presentation of key time nodes in chest pain diagnosis and treatment, this specification proposes the following embodiments:

[0060] Example 3

[0061] Reference Figures 1-2 This is the third embodiment of the present invention. This embodiment provides multiple comparative examples of comparative experiments, which systematically verify the significant technical advantages of the present invention in terms of anchor mismatch correction, real-time performance, time base drift fault tolerance, and intelligent arbitration of parallel paths. It proves that compared with traditional precise matching, post-event BI analysis, and conventional systems without context awareness, it can more accurately, reliably, and efficiently achieve automated real-time monitoring of key intervals in chest pain diagnosis and treatment.

[0062] A comparative validation experiment was conducted. The experimental platform was based on a historically anonymized HL7 event stream dataset containing 1000 independent chest pain diagnosis and treatment cases to simulate the output of a multi-source heterogeneous information system in a real emergency environment. The experimental environment consisted of a server and an event stream injection tool. The server hardware configuration met the real-time operation requirements of the system of this invention. The event stream injection tool was responsible for injecting the event stream from the dataset into the event listening module of the system at time intervals consistent with the real scenario. To simulate the anchor point mismatch, the injection tool was configured to randomly select 200 cases in the dataset and modify the text description of the standard end event anchor point of the time interval from the door to the electrocardiogram, i.e., the electrocardiogram report release event, for example, changing it to "ECG examination completed", thereby creating a scenario in which the end anchor point cannot be identified by the exact matching rule.

[0063] The experiment was conducted in two groups: a control group and an experimental group. In the control group's system, the adaptive correction module was disabled; in the experimental group's system, the adaptive correction module was enabled, and its internal correlation score was... computational model The weight parameters in the settings are as follows: and This setting is based on the fact that the injected errors in this experiment were mainly semantic textual differences; the evaluation metric for the experiment was the correct correction rate of the timeout interval, which was defined as the percentage of all injected error cases in which the candidate termination event inferred by the adaptive correction module matched the original standard termination event before the injected tool was modified; during the experiment, the quality control threshold for the door-to-ECG time interval was set to 600 seconds, and the adaptive correction module of the experimental group was triggered when the runtime of the active interval instance exceeded this threshold; after the experiment was executed, in the control group, all 2 In the 00 cases where errors were injected, the corresponding gate-to-ECG time active interval instances remained in a timeout state due to the inability to match the end anchor point, and were not closed until the data stream ended, with a timeout interval correctness rate of 0%. In the experimental group, the adaptive correction module was triggered after the active interval instances of 200 cases timed out, and inferred candidate end events based on correlation rules. After comparison with the original data, the inferred candidate end events of 188 cases were correct, with a timeout interval correctness rate of 94%. The key data records of the experiment are shown in Table 1 below.

[0064] Table 1 is a comparison table of experimental data for verifying the effectiveness of the adaptive correction module.

[0065]

[0066] Table 1 shows that when the adaptive correction module is disabled, the system is unable to handle the problem of non-standard text descriptions of end anchor point events; when the adaptive correction module is enabled, the system can identify and suggest the correct end event with a 94% accuracy rate without manual intervention, thereby correcting the interval timing error caused by anchor point mismatch.

[0067] To further verify the essential differences and technical advantages of the adaptive correction module of the present invention compared with conventional automation solutions in the field in solving the anchor mismatch problem, the following comparative examples are provided.

[0068] Comparative Example 1: This comparative example aims to simulate the performance of an automatic diagnostic interval acquisition system built on conventional technical approaches without an integrated adaptive correction module when processing real-world data. Except for the following key differences, the system architecture, interval template (door-to-ECG time, quality control threshold 600 seconds), event monitoring module, and interval instance management module used in this comparative example are consistent with the system of the present invention in Example 2. The key difference is that the system in this comparative example does not include an adaptive correction module, and its interval instance management module can only rely on keywords in the preset anchor point rules for precise and hard text matching when attempting to close an active interval instance.

[0069] An anonymized real-world historical dataset was used, covering the medical event flow of all 300 independent acute chest pain patients within the most recent calendar month. After manual verification and annotation, this dataset contained 45 cases (15%) of end-of-interview events in the time interval from the door to the electrocardiogram (ECG). The text descriptions in these events differed from the pre-defined ECG report publication rules in the interval template due to inconsistencies such as synonyms, abbreviations, or word order (e.g., "ECG examination completed," "12-lead ECG results issued"). These are potential causes of anchor point mismatch. When this dataset was injected into the comparative system, the system was able to correctly identify and close all 255 cases. (85%) of the interval instances had standardized end-event descriptions; however, for all 45 instances with non-standard text descriptions, the system failed to close the corresponding active interval instances due to the inability to achieve precise matching. These 45 instances ultimately exceeded the 600-second quality control threshold, were marked as closed and timed out by the system, and triggered timeout alarms. This directly resulted in quality control personnel having to manually trace and correct each of these 45 false positive alarms, greatly increasing the additional workload and seriously affecting the accuracy of the system's output data. The experimental results of this comparative example are compared with the processing results using the present invention, and the specific data are shown in Table 2 below:

[0070] Table 2. Performance comparison between the present invention and Comparative Example 1 in processing datasets containing anchor mismatch risk.

[0071]

[0072] Table 2 shows that in the test of the system of this invention, for 45 non-standard cases, the adaptive correction module successfully inferred and closed 43 of them (accuracy rate of approximately 95.6%), and only 2 cases failed to be corrected due to low text similarity and poor temporal proximity, which is more in line with real engineering scenarios. The test results clearly show that the automated system using conventional exact matching technology has inherent defects in its design principle when facing the problem of non-standard event descriptions that are common in real medical environments. It cannot avoid a large number of anchor mismatches, which leads to serious false positive alarms and relies on a lot of manual intervention, thus limiting the practical application value of the system. In contrast, the present invention, by introducing an adaptive correction module, can effectively deal with such uncertainties and automatically solve the anchor mismatch problem with extremely high accuracy, significantly improving the objectivity of time-consuming data collection and the credibility of the system.

[0073] Comparative Example 2: This comparative example aims to compare the differences between the present invention and post-analysis solutions based on traditional data warehouses and business intelligence (BI) tools in terms of the timeliness of diagnosis and treatment efficiency monitoring and human resource costs.

[0074] Scenario: The quality control department of a hospital needs to conduct statistical analysis on the time from the door to the electrocardiogram for all emergency chest pain patients in the previous quarter in order to evaluate the efficiency of the process.

[0075] Using conventional techniques: Quality control personnel first need to submit a data extraction request to the information center. Information technology personnel use ETL (Extract, Transform, Load) tools to extract patient registration events (starting point) from the HIS system database and report release events (ending point) from the ECG system database. Since the data structures and patient ID formats of the two systems may differ, the technicians need to write scripts to clean, correlate, and match the data. This process usually takes 2-3 working days. After the data is imported into the data warehouse, quality control personnel use BI tools to generate statistical charts through drag-and-drop and configuration. The entire process is completely manual, which is not only time-consuming and labor-intensive, but also has a significant lag in its analysis results. Problems can only be discovered after the quarter ends, and no real-time intervention can be made for delays in the diagnosis and treatment process.

[0076] Using this invention: After deployment, all time intervals from the gate to the electrocardiogram are defined as interval templates. The system automatically and in real-time captures event streams and creates and closes active interval instances. Quality control personnel can view real-time data accurate to the minute, the pass rate, and timeout warnings at any time through the treatment interval dashboard without any technical personnel intervention. When potential delays occur, the system can issue warnings in real time, rather than summarizing afterward.

[0077] Comparative conclusions: As shown in Table 3 below, this invention fundamentally solves the problem of lag in traditional solutions by changing the monitoring mode from manual retrospection after the fact to automatic insight during the process, and elevates the role of the information system from a passive data recording tool to a proactive process management and decision support tool.

[0078] Table 3. Performance Comparison of the Invention and Comparative Example 2 in the Diagnosis and Treatment Efficiency Monitoring Paradigm.

[0079]

[0080] Table 3 describes a performance comparison between the present invention and Comparative Example 2 in the diagnostic and treatment efficiency monitoring paradigm.

[0081] Comparative Example 3 compares the reliability of the output results of the present invention with that of a conventional interval acquisition system without time base compensation function when faced with time base drift of the data source.

[0082] Scenario setting: The electrocardiogram system of a certain hospital is a standalone physical server. Its NTP (Network Time Protocol) service occasionally fails, causing its system clock to drift ahead of the standard time (such as the HIS system time) by about 90 seconds.

[0083] A conventional system, capable of automatically collecting time intervals, is employed. However, its design relies on an idealized assumption: that all source systems are perfectly synchronized. When a patient registers at 13:00:00 (HIS system time) and completes an ECG at 13:08:00, the ECG system clock is 90 seconds ahead, resulting in a report release timestamp of 13:09:30. The conventional system directly uses these two timestamps to calculate the time elapsed, yielding a result of 570 seconds (9 minutes and 30 seconds). This result is numerically inaccurate; it incorrectly classifies a compliant interval of 480 seconds (8 minutes) as a suspected timeout interval close to the 10-minute quality control threshold, thus outputting misleading and unreliable data. The conventional solution is a hospital-wide IT infrastructure upgrade to ensure mandatory, strict time synchronization across all systems, which is costly and difficult to implement in practice.

[0084] The present invention employs a time base compensation module that continuously performs statistical analysis on the timestamp intervals of the ECG system event stream during system operation. When an NTP service failure causes clock drift, the module detects statistically significant anomalies in the moving average and standard deviation of the timestamp intervals. The system then determines that the time base of the source system has drifted and marks it as pending verification. When the aforementioned ECG report event (carrying a 13:09:30 timestamp) is used for a closed active interval instance, although the system will still calculate a duration of 570 seconds, a confidence marker indicating that the data source time base is pending verification will be automatically added to the result on the treatment interval dashboard.

[0085] Comparative conclusion: Conventional technical solutions can produce silent errors in non-ideal IT environments, compromising the objectivity of their output data. This invention, through its built-in time base compensation module, does not attempt to correct an erroneous timestamp whose true drift amount is unknown. Instead, it adds a layer of credibility information about the data source quality to the final time-consuming data, significantly improving the reliability and honesty of the system's output in real, complex IT environments. This provides crucial evidence for subsequent manual review.

[0086] Comparative Example 4: This comparative example aims to compare the information presentation efficiency and clinical value of the present invention with that of a conventional monitoring panel using an indiscriminate independent timer when processing parallel treatment pathways.

[0087] Scenario setting: As in Example 1, a patient with acute chest pain needs to be diagnosed simultaneously for two possible causes: acute coronary syndrome (ACS) and pulmonary embolism (PE). Therefore, the system starts the timing of multiple treatment intervals in parallel.

[0088] The conventional approach involves a rudimentary electronic whiteboard or monitoring panel capable of starting independent countdown or positive timers for different quality control items (such as troponin and D-dimer result times). During diagnosis and treatment, all timers operate independently. When the troponin result is positive at 12:45, indicating a highly definitive ACS diagnosis, the timers on the monitoring panel related to the PE diagnostic pathway—specifically the D-dimer result time and pulmonary CTA report time—will not change. They will continue running, and may even trigger red alerts due to exceeding their respective quality control thresholds. This presentation method is completely out of sync with the evolution of clinical thinking, generating useless information noise and potentially distracting the medical team, leading to alarm fatigue.

[0089] The present invention employs a clinical context arbitrator module that incorporates rules regarding the clinical logical relationships between different diagnostic pathways. When a troponin-outcome interval instance closes with a positive result, the module is triggered and, based on preset mutual exclusion rules, automatically and dynamically adjusts the visualization status of all active interval instances (such as D-dimer-outcome interval instances) related to the PE diagnostic pathway that are still being timed to low priority (e.g., displayed in gray).

[0090] Comparative Conclusion: Conventional technical solutions provide isolated and unrelated data presentations, requiring medical staff to rely on their own experience to perform secondary processing and filtering of information at the cognitive level. This invention, by introducing a clinical context arbitrator, embeds clinical logic into the system's processing flow, achieving automatic and real-time synchronization between the focus of information presentation and the clinical diagnostic approach. This transforms the system from a simple information bulletin board into an intelligent assistant capable of understanding clinical context and preprocessing and prioritizing information, significantly reducing the user's cognitive load and improving the efficiency and value of information interaction.

[0091] To address the challenge of real-time detection and correction of source system time base drift through the collaborative work of various system modules, thereby ensuring the accuracy and reliability of time consumption data for critical intervals in chest pain diagnosis and treatment, and achieving credible, real-time, and visualized monitoring, this specification proposes the following embodiments:

[0092] Example 4

[0093] Reference Figures 1-2 This is the fourth embodiment of the present invention. This embodiment provides the overall architecture and data flow coordination mechanism of the system of the present invention, and explains how the time base compensation module identifies the source system clock drift and works in conjunction with the interval instance management, adaptive correction and human-computer interaction modules to achieve real-time, reliable and visual monitoring of the time consumption of key intervals in chest pain diagnosis and treatment.

[0094] This embodiment combines Figures 1 to 2 Description of the automatic data acquisition system for key timeline nodes in chest pain diagnosis and treatment, such as... Figure 2 As shown in the figure, the horizontal axis is time series in minutes and the vertical axis is event time interval in seconds. It shows the changes in the timestamp intervals of event streams from three different systems: HIS system, LIS system, and electrocardiogram system. The event time intervals of the HIS system and LIS system fluctuate within a stable range, while the event time intervals of the electrocardiogram system show abnormal values ​​with a continuous increase and violent jitter after about 30 minutes, which shows that the time base of the system has drifted. The time base compensation module of the present invention is designed to identify and mark such anomalies.

[0095] like Figure 1As shown, multiple medical information systems act as event sources, continuously generating cross-source event streams carrying patient identifiers and timestamps. An event listening module is responsible for acquiring and distributing this event stream in real time. The distributed event metadata is split into two paths: one is sent to the interval instance management module, and the other is sent to the time base compensation module. The interval instance management module creates, maintains, and closes active interval instances in memory based on anchor point rules retrieved from the interval definition library, and calculates the time consumption. Then, it outputs the structured interval status and time consumption information to a human-computer interaction interface module. When the interval instance management module detects that the timeout has not been closed... When an instance is closed, the adaptive correction module is triggered. After inferring the candidate end event, the module submits a confirmation request to the human-computer interaction interface module and returns the candidate end event to close the instance based on the received user confirmation instruction. At the same time, the time base compensation module performs independent statistical analysis on the event flow to identify the time base drift of the source system and attaches the resulting confidence mark to the human-computer interaction interface module. Finally, medical and quality control personnel can obtain graphical instance information for each interval, which includes real-time status, time consumption results and confidence mark, through the human-computer interaction interface module.

[0096] To address the challenge of optimizing the weight parameters of an adaptive correction module using offline calibration methods in heterogeneous medical information system environments, thereby improving the accuracy of event termination inference in the event of anchor mismatch, this specification proposes the following implementation:

[0097] Example 5

[0098] Reference Figures 1-2 This is the fifth embodiment of the present invention, which provides an offline calibration method for adaptive correction module weight parameters for heterogeneous hospital system environments, in order to optimize the correction accuracy under anchor mismatch.

[0099] The automatic data acquisition system of this invention needs to be connected to a network environment consisting of multiple heterogeneous medical information systems built at different times. The event descriptions of the same diagnostic and treatment behavior differ in textual standardization among these systems. This difference in textual standardization is addressed in the adaptive correction module for correlation scoring. Computational model Weight parameters and The following procedure can be used for offline calibration: The execution of this procedure begins in the data preparation stage. From the database of the hospital to be deployed, extract all completed anonymized historical event stream data related to chest pain diagnosis and treatment within a certain period, such as the past 6 months. Quality control personnel manually label the start and end events of the accurate time interval from the door to the electrocardiogram for no less than 500 cases based on the medical records, forming a benchmark dataset containing factual criteria. Subsequently, from this benchmark dataset, select 100 cases whose text descriptions of the end events do not conform to the standard anchor point rules in the interval definition library, but are manually confirmed as correct, as the offline calibration dataset.

[0100] During the calibration execution phase, the system runs in offline mode and performs iterative calculations on the aforementioned 100 calibration datasets; taking into account the weight parameters... and There are The constraint relationship simplifies the calibration process to a single parameter. Finding the optimal value; setting The search space is [0.1, 0.9], and the search step size is 0.1, thus forming 9 sets of candidate weight parameter pairs. The system will use these 9 sets of parameter pairs sequentially to process all 100 cases in the calibration dataset. That is, after the time interval from the door to the electrocardiogram in each case expires, the adaptive correction module is triggered, and the inferred candidate termination event is recorded. By comparing the system's inference results with the factual standards in the dataset, the correct correction rate of the timeout interval corresponding to each set of candidate weight parameter pairs is calculated. After calculating all 9 sets of parameter pairs, a set of correction rate data is obtained. One set of calibration results shows that when Set to 0.6. When set accordingly to 0.4, the correct correction rate for the timeout interval reaches a peak of 95%; when When the value is below or above 0.6, the correction rate decreases to varying degrees; accordingly, the output of this calibration procedure is that, in the information technology environment of this specific hospital, a set of preferred weight parameters for the correlation score calculation model are configured as follows: and By executing this calibration procedure, parameter settings are transformed into an engineering process with defined inputs, closed execution steps, and verifiable outputs.

[0101] To address how the system can automatically optimize the keyword library and time base drift judgment threshold in anchor point rules based on user feedback and historical data during long-term operation, thereby achieving continuous adaptive improvement in the accuracy of diagnosis and treatment event identification and the ability to detect time synchronization anomalies, this specification proposes the following embodiments:

[0102] Example 6

[0103] Reference Figures 1-2 This is the sixth embodiment of the present invention, which provides an adaptive maintenance method that automatically optimizes anchor point rules and time drift thresholds based on user feedback and historical data.

[0104] After deployment, the system of this invention can execute a periodic keyword library maintenance procedure. The system can run a built-in maintenance program, which first retrieves and counts the pairing relationships between the candidate end events for closing timeout active interval instances confirmed by the user through the human-computer interaction interface module and their corresponding core semantic keywords within a certain operating cycle. Then, the program performs frequency analysis on these pairing relationships, identifies new keywords in the descriptions of frequently successfully paired events, and generates a suggestion list. If the core keyword "ECG" for the door-to-ECG time interval is frequently paired with the user-confirmed 12-lead ECG examination event description, then 12-lead ECG will be recommended as a new synonym with high confidence. After reviewing the list, the administrator can batch confirm the addition of these new keywords to the synonym set of the end anchor rule of the corresponding interval template in the interval definition library, thereby iteratively optimizing the anchor rule in a data-driven manner.

[0105] To calibrate the threshold used to determine drift anomalies in the time base compensation module, the system can execute a preliminary baseline analysis procedure before connecting to a new event source system. This procedure first extracts the source system timestamp data from the historical database of the source system, such as the LIS system, covering all events generated within a known stable clock cycle, such as the past month. Subsequently, the calibration program processes this timestamp data, calculates the timestamp intervals between adjacent events, and performs statistical distribution analysis on all interval values ​​to obtain the baseline moving average and baseline moving standard deviation of the time interval distribution under stable operating conditions. Based on this statistical distribution, the procedure further calculates and determines the standard deviation multiple that can cover 99.9% of the historical time interval sample points, such as 3.5 times, and sets this value as the threshold for determining time base drift anomalies in the source system. Through this procedure, a statistically based, quantitative drift judgment threshold is determined for a specific data source.

[0106] To address the challenge of building a proprietary baseline operating model based on an institution's historical data before system deployment, scientifically solidifying quality control thresholds, assessing data integrity risks, and calibrating adaptive correction algorithm parameters to achieve precise system adaptation to the local medical environment, this specification proposes the following embodiments:

[0107] Example 7

[0108] Reference Figures 1-2This is the seventh embodiment of the present invention. This embodiment provides a standardized procedure for building an institution-specific baseline operating model based on historical data before system deployment, which is used to adapt to the local data environment and solidify quality control thresholds and correct algorithm parameters.

[0109] To ensure that its internal statistical analysis and adaptive correction functions are accurately adapted to the institution's unique data environment, a standardized offline analysis procedure is required to build a baseline operating model specific to the institution. This procedure takes as input a manually reviewed and anonymized historical event stream dataset covering at least 1,000 independent chest pain treatment cases within the institution over the past 12 months, labeled with fact-standard treatment intervals. The procedure first performs a baseline assessment of the dataset's integrity by reversing the search for all labeled end events and verifying whether a corresponding start event exists within the maximum reasonable time window of 30 minutes. This identifies and counts the number of isolated end events, and the statistical results are recorded in the model as one of the benchmark indicators for assessing the risk of data loss from upstream data sources.

[0110] Based on this dataset, the procedure establishes a performance baseline for key diagnostic and treatment processes. For the time interval from door to ECG, the program analyzes the actual time taken for all successfully matched and closed instances, calculates its statistical distribution, and embeds the 95th percentile time (e.g., 550 seconds) as the baseline quality control threshold for this interval within the medical institution into the model. Simultaneously, the procedure finalizes the parameters of the internal algorithm model for the adaptive correction module, specifying the function used to calculate the normalized time proximity value as follows:

[0111] ;

[0112] in, The time difference between the occurrence time of the candidate end event and the creation time of the active interval instance. The baseline quality control thresholds determined in the aforementioned steps are used to generate a baseline operating model for the specific medical institution. This model includes data integrity baseline indicators, process performance baseline parameters, and a deterministic algorithm model. The model is then loaded into the online system as the basis for real-time judgment and adaptive correction.

[0113] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An automatic timeline key node acquisition system for chest pain diagnosis and treatment, characterized in that: include, Medical information systems include system time (HIS), source system (LIS), and electrocardiogram (ECG) systems; An interval definition library is used to pre-store interval templates containing start or end anchor point rules. The interval templates refer to multiple chest pain-specific interval templates that are configured to be pre-stored. The event listening module is used to acquire event streams from one or more medical information systems in real time, carrying the chest pain patient identifier and the source system timestamp. The interval instance management module is used to create and maintain an active interval instance in memory for a chest pain patient when an event matches the starting anchor point rule of any chest pain-specific interval template, based on the event flow. The adaptive correction module is used to handle anchor point mismatch and infer candidate termination events; The time base compensation module is used to identify time base drift in the source system and add a credibility marker to the time-consuming data; Furthermore, the adaptive correction module includes: when the running time of any active interval instance exceeds a preset time and is still not closed, it automatically infers candidate termination events from events related to chest pain patients that do not match any anchor rules, which are filtered by the interval instance management module, based on correlation rules, and closes the active interval instance through a single confirmation operation of a human-computer interaction interface module; wherein, the correlation rules include configuring the adaptive correction module to extract the core semantic keywords of the termination anchor rules of active interval instances that have not been closed within the time limit, and filtering all events that occurred after the creation of the active interval instance from events that do not match any anchor rules; For each selected event, a pre-defined scoring model is used to calculate the event's relevance score. The event with the highest score is selected as the candidate termination event, and the event relevance score is... The calculation formula is: ； in, Match weights to preset keywords. A normalized text similarity value between 0 and 1 is calculated based on the descriptive information and core semantic keywords of the selected events. The preset time proximity weight, This is a normalized time proximity value between 0 and 1, calculated based on the time difference between the occurrence time and the active interval instances of the selected events. The time base compensation module includes continuous statistical analysis of the source system timestamps of the event stream, taking the source system of the event as the dimension, identifying drift anomalies in the source system time base, and compensating for the time consumption or adding credibility markers based on the drift anomalies. The interval definition library also includes configuration of the interval definition library, used to define a chest pain specific interval template and clinical logical relationship rules between it and one or more other chest pain specific interval templates; The clinical logical relationship rules include a clinical context arbitrator module, which includes configuration options for the clinical context arbitrator module. When multiple active interval instances with logical relationship rules are created for the same chest pain patient, the visualization priority status of one or more other related active interval instances on the human-computer interaction interface module is dynamically adjusted according to the closure status of one of the active interval instances or a preset quality control status mark, following the logical relationship rules.

2. The automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment as described in claim 1, characterized in that: The chest pain-specific interval template in the interval definition library also includes an intermediate checkpoint anchor rule configured to define at least one intermediate checkpoint anchor rule between the start anchor rule and the end anchor rule. The interval instance management module also includes being configured to continuously monitor and match the corresponding intermediate checkpoint anchor rules when an active interval instance is in the created state, and to preprocess the time consumption of the continuous sub-stages constituting the treatment interval based on the internal progress status. Match the event stream with the end anchor rule of the active interval instance. When a match is successful, close the active interval instance and calculate the time consumed.

3. The automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment as described in claim 1, characterized in that: The time base compensation module also includes configuring the time base compensation module to calculate and maintain the moving average and moving standard deviation of the source system timestamp intervals of adjacent events in the event stream in real time for each source system. When the timestamp interval of a newly arrived event in the source system deviates from the moving average by more than a preset multiple of the moving standard deviation, it is determined that the time base of the source system has drifted abnormally. The additional credibility markers include marking the time-consuming results of active interval instances closed by events originating from the anomalous source system on the human-computer interaction interface module as data source time bases to be verified.

4. The automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment as described in claim 1, characterized in that: The human-computer interaction interface module includes a treatment zone dashboard, which is configured to provide an independent display unit for each type of quality control zone in the form of a graphical progress bar array. Using a pre-defined visual coding system, the system can distinguish and display in real time the status of each active region instance, including those that have been created, are in time, are closed and meet the closure criteria, as well as those that have timed out and are set to low priority due to clinical logic arbitration.

5. The automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment as described in claim 1, characterized in that: The event listening module includes a stateless event listener, which is configured to passively listen to the standardized medical information exchange protocol message bus to obtain event streams. After obtaining each event, it does not store any historical state information and distributes the metadata of the obtained event to the interval instance management module and the time base compensation module. The interval definition library also includes a graphical configuration interface configured to be used by non-IT professionals. The graphical configuration interface includes features configured to allow non-IT professionals to manipulate chest pain specific interval templates using a structured form.

6. An automatic acquisition method for key timeline nodes in chest pain diagnosis and treatment, based on the automatic acquisition system for key timeline nodes in chest pain diagnosis and treatment as described in any one of claims 1 to 5, characterized in that: include, Pre-store interval templates containing start or end anchor point rules, wherein the interval templates refer to multiple chest pain-specific interval templates configured to be pre-stored; Real-time acquisition of event streams from one or more medical information systems, carrying chest pain patient identifiers and source system timestamps; Based on the event flow, when an event matches the starting anchor rule of any chest pain-specific interval template, an active interval instance in memory is created and maintained for the chest pain patient. Handle anchor point mismatch and infer candidate termination events; Identify time base drift in the source system and attach credibility markers to time-consuming data.

Images