An automatic processing method and system for emergency data quality control analysis

Abstract

The application provides an automatic processing method and system for emergency data quality control analysis, and relates to the technical field of medical information processing. First, the time series data of the patient and the real-time record of the medical resource are acquired to generate emergency process flow data, and the signal abnormal interval in the time series data is identified, and the target resource event is associated. Then, the emergency process flow data is divided into stages to obtain a clinical stage unit. In the clinical stage unit, the identified signal abnormal interval and the associated target resource event are combined to generate a quality control analysis focus. The logic judgment rule corresponding to the clinical scene feature is screened, the quality control logic is generated, and finally the automatic calculation of the quality control logic is performed on the quality control analysis focus to output the quality control abnormal event record. The technical scheme provided by the application not only realizes the automatic conversion from mixed data to closed-loop quality control decision, but also effectively improves the accuracy, scene and efficiency of quality control.

Description

Technical Field

[0001] This application relates to the field of medical information processing technology, and in particular to an automated processing method and system for emergency data quality control analysis. Background Technology

[0002] With the continuous improvement of information technology in emergency medicine, a massive amount of time-series data on patient physiological parameters and records of medical resource scheduling events have been generated. In order to improve medical quality and safety, there is an urgent technical need for clinical management to conduct systematic and automated quality control analysis of the entire emergency process.

[0003] Currently, the technical solution adopted is an anomaly detection algorithm for time series of single physiological parameters. This solution first collects and digitizes the continuous waveform signals output by monitoring devices such as ECG and blood pressure in real time. Then, through preset thresholds or waveform pattern matching algorithms, it directly identifies segments in the signal that exceed the normal physiological range or have abnormal morphology, and marks these segments as suspected data quality problems or clinical events.

[0004] However, such solutions primarily rely on isolated judgments based on single-dimensional data waveforms, lacking correlation analysis between the specific clinical scenario and medical resource context when anomalies occur. For example, they cannot distinguish whether signal anomalies are due to equipment failure, patient movement interference, or occur within a reasonable timeframe for critical treatment procedures performed by medical staff. This results in alarms or quality control prompts being disconnected from the actual clinical process, leading to a high false alarm rate. Furthermore, they cannot assess whether the temporal logic of medical operations and changes in patient status is compliant, failing to meet the need for integrated quality control of the entire emergency treatment process. Summary of the Invention

[0005] This application provides an automated processing method and system for emergency data quality control analysis, which solves the problem in the prior art that the quality control judgment is disconnected from clinical scenarios and resource events, resulting in one-sided analysis results and insufficient practicality.

[0006] In a first aspect, this application provides an automated processing method for emergency data quality control analysis, including: Acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data; Based on the emergency process flow data, identify the abnormal signal intervals in the time series data and associate them with target resource events that occur in the same spatiotemporal dimension in the real-time medical resource records; Based on a predefined emergency clinical pathway, the emergency process flow data is divided into stages to obtain clinical stage units, which are used to reflect different treatment urgency levels. Within the clinical phase unit, a quality control analysis focus is generated by combining the identified abnormal signal intervals and associated target resource events; Based on the clinical scenario characteristics included in the quality control analysis focus, logical judgment rules corresponding to the clinical scenario characteristics are selected from the pre-set emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus. The quality control logic is automatically calculated for the quality control analysis focus to output a quality control anomaly event record marked with a quality control anomaly event.

[0007] Optionally, time-series data of emergency patients during their visit and real-time records of medical resources can be acquired to generate emergency process flow data, including: Physiological parameter measurements were collected from monitoring equipment of emergency patients, and the corresponding collection time was marked for each physiological parameter measurement to obtain time series data; Real-time acquisition of bed status change events and medical staff task allocation events from the emergency department resource management system, and marking the corresponding event occurrence time for each event to obtain real-time records of medical resources; All physiological parameter measurements in the time series data are arranged in chronological order of the acquisition times to form an ordered sequence of physiological parameters. All events recorded in the real-time medical resource data are arranged in chronological order of their occurrence to form an ordered sequence of resource events. The ordered sequence of physiological parameters and the ordered sequence of resource events are interleaved and combined in chronological order to form emergency process flow data.

[0008] Optionally, based on the emergency process flow data, identify abnormal signal intervals in the time series data and associate them with target resource events occurring in the same spatiotemporal dimension in the real-time medical resource records, including: Extract ordered sequences of physiological parameters from the emergency process flow data; In the ordered sequence of physiological parameters, points where the change in physiological parameter measurement values ​​between adjacent acquisition times exceeds a preset change threshold are detected and identified as first-type candidate anomalies. In the ordered sequence of physiological parameters, points where the same physiological parameter measurement value appears consecutively are detected as second-type candidate abnormal points; Based on the first type of candidate anomalies and the second type of candidate anomalies, the signal abnormality time period is determined in the ordered sequence of physiological parameters; Extract the ordered sequence of resource events from the emergency process flow data; In the ordered sequence of resource events, find the first target resource event that overlaps in time with the time period of the abnormal signal and whose location identifier matches the location of the monitoring device that generated the ordered sequence of physiological parameters; The first target resource event is associated with the time period of the signal anomaly, and thus considered as a target resource event occurring in the same spatiotemporal dimension.

[0009] Optionally, the emergency process flow data is divided into stages according to a predefined emergency clinical pathway to obtain clinical stage units, which are used to reflect different levels of treatment urgency, including: Extract the ordered sequence of resource events from the emergency process flow data; Based on a predefined emergency clinical pathway, several standard medical node event types are identified, including triage completion, first medication administration, outpatient examination, and hospitalization application. In the ordered sequence of resource events, events that match the standard medical node event type are searched and used as stage-separating events; Record the occurrence time of each phase separation event in the ordered sequence of resource events, and use it as the phase separation time; According to the chronological order of the stage separation times, continuous time segments are divided in the emergency process flow data; Each of the aforementioned time segments is defined as a clinical phase unit.

[0010] Optionally, within the clinical phase unit, a quality control analysis focus is generated by combining the identified abnormal signal intervals and associated target resource events, including: Within the time segment corresponding to the clinical stage unit, filter out signal abnormality time periods whose time range is completely contained within the time segment; Acquire target resource events that are associated with each time period of signal anomaly that has been filtered; Determine whether there is any overlap in time between the abnormal signal time period and the time of occurrence of the event corresponding to the acquired target resource event; When it is determined that there is a time overlap, the abnormal time periods of the signal that overlap in time and the target resource event are combined to form a quality control analysis focus.

[0011] Optionally, based on the clinical scenario characteristics included in the quality control analysis focus, logical judgment rules corresponding to the clinical scenario characteristics are selected from a pre-set emergency quality control standard knowledge base to generate quality control logic, including: Extract signal anomaly type features from the signal anomaly time periods included in the quality control analysis focus; Extract stage type features from the clinical stage unit where the quality control analysis focus is located; Extract resource event type features from the target resource events included in the quality control analysis focus; The extracted signal anomaly type features, stage type features, and resource event type features are combined to generate combined features; In the emergency quality control standard knowledge base, search the logical judgment rule base. Each logical judgment rule in the logical judgment rule base has a corresponding rule trigger feature pre-set. The combined features are matched with the pre-defined rule triggering features of each logical judgment rule; Filter out the logical judgment rules that match the combined features, and use them as target logical judgment rules; The selected target logic judgment rules are arranged according to the predefined rule execution order to generate quality control logic.

[0012] Optionally, the quality control logic is automatically calculated for the quality control analysis focus to output a quality control anomaly event record marked with a quality control anomaly event, including: From the quality control analysis focus, obtain ordered sequence fragments of physiological parameters containing abnormal signal time periods, and ordered sequence fragments of resource events containing target resource events; Extract the first target logic judgment rule in the quality control logic according to the execution order of the predefined rules in the quality control logic; The ordered sequence fragments of the physiological parameters and the ordered sequence fragments of the resource events are substituted into the judgment conditions of the first target logic judgment rule for comparison and calculation to obtain the first intermediate judgment result; Determine whether the first intermediate judgment result satisfies the conclusion condition of the first target logic judgment rule; If the conclusion of the first target logic judgment rule is satisfied, then the conclusion corresponding to the first target logic judgment rule is recorded as the first intermediate conclusion. Repeat the steps of extracting the target logic judgment rules from the quality control logic, substituting and comparing the calculations, judging whether the conditions for the conclusion to be valid, and recording the conclusion, until all the target logic judgment rules in the quality control logic have been executed, so as to obtain the intermediate conclusion combination. Based on the aforementioned intermediate conclusions, a final conclusion is generated targeting the focus of the quality control analysis. The quality control analysis focus, the quality control logic, and the final conclusion are linked and bound together to generate a quality control anomaly event record.

[0013] Secondly, this application provides an automated processing system for emergency data quality control analysis, comprising: The acquisition module is used to acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data; The identification module is used to identify abnormal signal intervals in the time series data based on the emergency process flow data, and associate them with target resource events that occur in the same spatiotemporal dimension in the real-time medical resource record. The segmentation module is used to divide the emergency process flow data into stages according to the predefined emergency clinical pathway to obtain clinical stage units, which are used to reflect different treatment urgency levels. The generation module is used to generate quality control analysis focus within the clinical stage unit by combining the identified abnormal signal intervals and associated target resource events; The filtering module is used to filter logical judgment rules corresponding to the clinical scenario characteristics contained in the quality control analysis focus from a pre-set emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus. The calculation module is used to perform automated calculations of the quality control logic on the quality control analysis focus to output quality control abnormal event records marked with quality control abnormal events.

[0014] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to realize an automated processing method for emergency data quality control analysis as described in the first aspect above.

[0015] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements an automated processing method for emergency data quality control analysis as described in the first aspect.

[0016] This application integrates time-series data of emergency patients with real-time records of medical resources into unified process flow data. When anomalies are identified, it actively correlates them with medical resource events occurring in the same spatiotemporal dimension. This effectively breaks down data barriers between different information systems, enabling the judgment of data quality or process compliance to shift from isolated data point analysis to correlation analysis combined with specific medical operation scenarios. This reduces misjudgments caused by reasonable circumstances such as medical operation interference, and improves the accuracy and clinical relevance of quality control analysis.

[0017] Furthermore, by dividing the process into stages based on the emergency clinical pathway, the identified abnormalities and event combinations are placed within specific treatment stages to form the focus of quality control analysis. This drives the corresponding rules in the knowledge base to perform automated logical calculations. This method achieves a leap from discrete alarms to systematic quality control based on multi-dimensional rules for the entire clinical process. It generates quality control records containing complete contextual data, events, logic, and conclusions, which not only significantly improves the automation and efficiency of quality control work, but also makes the quality control results have good interpretability and traceability, providing a direct and reliable basis for subsequent medical quality improvement.

[0018] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of an automated processing method for emergency data quality control analysis provided in this application is shown; Figure 2 This invention provides a schematic diagram of the structure of an automated processing system for emergency data quality control analysis. Figure 3 A schematic diagram of the structure of a computing device provided in this application is shown. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solution of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0022] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Figure 1 This application provides a flowchart of an automated processing method for emergency data quality control analysis, such as... Figure 1 As shown, the method includes: Step 101: Obtain time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data.

[0025] Optionally, step 101 may specifically include: Step 1011: Collect physiological parameter measurements from the monitoring equipment of emergency patients and mark the corresponding collection time for each physiological parameter measurement to obtain time series data.

[0026] Step 1012: Obtain bed status change events and medical staff task allocation events in real time from the emergency department resource management system, and mark the corresponding event occurrence time for each event to obtain real-time records of medical resources.

[0027] Step 1013: Arrange all the measured values ​​of physiological parameters in the time series data according to the order of the acquisition time to form an ordered sequence of physiological parameters.

[0028] Step 1014: Arrange all events in the real-time records of the medical resources according to the chronological order of their occurrence to form an ordered sequence of resource events.

[0029] Step 1015: The ordered sequence of physiological parameters and the ordered sequence of resource events are interleaved and combined in chronological order to form emergency process flow data.

[0030] In this step, time-series data refers to the continuous collection of physiological parameter measurements, such as heart rate and blood pressure readings, from the monitoring equipment of emergency patients, with precise acquisition time stamps. These measurements are used to record the changes in the patient's physiological state in chronological order.

[0031] Real-time medical resource records refer to event information such as changes in bed status and assignments of medical staff that are obtained in real time from the hospital's emergency department resource management system and have precise time stamps. This information is used to track the use and allocation of medical resources during the treatment process.

[0032] Emergency process flow data refers to a unified data stream formed by alternately splicing ordered sequences of physiological parameters and ordered sequences of resource events in chronological order. This data stream provides a timeline basis for subsequent analysis that integrates the patient's physiological state and the context of medical operations.

[0033] Physiological parameter measurements refer to the raw values ​​read directly from monitoring devices such as electrocardiogram monitors and blood pressure monitors connected to the patient. These values ​​reflect the patient's real-time vital signs and are acquired in real time through the device's data output port.

[0034] An emergency department resource management system refers to an information-based software system within a hospital used to manage and track the status and tasks of resources such as beds, medical staff, and equipment within the emergency department. It provides digital records of resource allocation events, which can be obtained through the application programming interface or data export function provided by the management system.

[0035] A bed status change event refers to a specific piece of information recorded in the emergency department resource management system regarding a change in the status of a certain emergency bed, such as from vacant to occupied, or from occupied to disinfected. It is used to mark the moment when the resource occupancy status changes, and is obtained by listening to or querying the status change log of the resource management system.

[0036] A medical staff task assignment event refers to a specific piece of information recorded in the emergency department resource management system regarding the assignment of a specific medical task, such as a bedside electrocardiogram or intravenous puncture, to a specific medical staff member. This information is used to track the person responsible for performing the medical action and the start time, and is obtained by retrieving the task scheduling records from the management system.

[0037] The moment of occurrence refers to the point in time when a bed status change event or a medical staff task assignment event is officially recorded in the management system. It is used to provide a precise timestamp for each resource event and is usually automatically generated and recorded by the emergency department resource management system when the event is created.

[0038] An ordered sequence of physiological parameters refers to a data list formed by sorting all collected physiological parameter measurements from morning to night according to their corresponding collection times. It is used to organize discrete measurement points into an ordered data line that progresses over time. It is obtained by sorting the original time series data by timestamp.

[0039] The ordered sequence of resource events refers to a list of events formed by sorting all acquired bed status change events and medical staff task assignment events according to their corresponding event occurrence times from early to late. It is used to organize discrete resource events into an ordered event line that develops over time. It is obtained by sorting the original medical resource records by timestamp in real time.

[0040] In this step, the data acquisition interface program deployed on the medical intranet first connects to the network or serial port of the emergency monitoring equipment in a polling manner. The program parses the measured values ​​of physiological parameters in digital format from the data stream continuously sent by the equipment in real time. At the moment of reading each measurement value, a precise timestamp is extracted from the data frame or generated by the system clock to tag the measurement value with the acquisition time, thereby obtaining structured time series data. Secondly, by writing scripts to call the API interface provided by the hospital's emergency department resource management system, data query requests that have been authenticated are sent at specific time intervals, such as every second, and the structured data returned by the system is received. From this data, two core parts are parsed out: bed status change records and medical staff task allocation events. The event content of each record and its precise timestamp generated in the system database are extracted, thereby obtaining a structured real-time record of medical resources. Next, for the time series data, all data entries are loaded, and the quick sort algorithm is used to reorganize them using the collection time field in each data entry as the sort key. By selecting a benchmark value and recursively partitioning, a physiological parameter ordered sequence in which all data points are strictly arranged in chronological order is finally output. Simultaneously, the same sorting logic is performed on the real-time records of medical resources, but the sorting key is the event occurrence time, generating a resource event ordered sequence in which all events are arranged in chronological order. Finally, a two-pointer merging algorithm is used to process the two sorted sequences. The algorithm first initializes two pointers to point to the head of the two sequences respectively. In the loop, the timestamp values ​​of the data items currently pointed to by the two pointers are continuously compared. Data items with smaller timestamp values ​​are taken out of their sequences and appended to a new cache list in order. At the same time, the pointer of the sequence is moved forward one position. This process of comparison, retrieval and appending is continued until all data items of the two sequences have been retrieved and merged into the cache list. The final cache list, which interweaves physiological data points and resource event points according to a globally unified timeline, is the temporally aligned emergency process flow data required for subsequent analysis.

[0041] For example, patient A enters the emergency department of a hospital due to chest pain and is connected to an electrocardiogram (ECG) monitor. The monitor first begins to continuously measure data such as heart rate and blood pressure, and during the measurement process, the acquisition time is automatically marked with millisecond precision and transmitted to the server via the hospital network to form time-series data. At the same time, the emergency department resource management system automatically records resource events related to the patient: when the patient is assigned to bed 3 in the resuscitation area, a bed occupancy record with a timestamp is generated; then, after the doctor issues an ECG examination order, a task allocation record with a timestamp is generated. These constitute a real-time record of medical resources. Subsequently, a sorting algorithm is used to arrange all physiological measurements from earliest to latest according to their timestamps, generating an ordered sequence of physiological parameters; at the same time, all resource events are sorted according to their occurrence time, generating an ordered sequence of resource events; finally, these two ordered sequences are merged alternately according to their time sequence, such as arranging the heart rate value at 14:00:00, the bed event at 14:00:05, and the blood pressure value at 14:00:10 in sequence, forming a complete emergency process flow data that records the patient's physiological changes and the use of medical resources.

[0042] Step 102: Based on the emergency process flow data, identify the abnormal signal intervals in the time series data and associate them with target resource events that occur in the same spatiotemporal dimension in the real-time medical resource records.

[0043] Optionally, step 102 may specifically include: Step 1021: Extract an ordered sequence of physiological parameters from the emergency process flow data.

[0044] Step 1022: In the ordered sequence of physiological parameters, detect points in which the change in the measured value of the physiological parameter between adjacent acquisition times exceeds a preset change threshold, and use them as first-class candidate anomalies.

[0045] Step 1023: In the ordered sequence of physiological parameters, detect points where the same physiological parameter measurement value appears consecutively, as second-type candidate abnormal points.

[0046] Step 1024: Based on the first type of candidate anomalies and the second type of candidate anomalies, determine the signal abnormality time period in the ordered sequence of physiological parameters.

[0047] Step 1025: Extract the ordered sequence of resource events from the emergency process flow data.

[0048] Step 1026: In the ordered sequence of resource events, find a first target resource event that overlaps in time with the abnormal signal time period and whose location identifier matches the location of the monitoring device that generated the ordered sequence of physiological parameters.

[0049] Step 1027: Associate the first target resource event with the abnormal signal time period as target resource events occurring in the same spatiotemporal dimension.

[0050] In this step, the signal abnormality interval refers to a continuous time period in time series data where the measured values ​​of physiological parameters do not clearly conform to the normal physiological change pattern due to equipment failure, signal interference, or non-physiological operation. It is used to identify the period when possible data quality problems or special clinical events occur. It is obtained by performing abnormal detection on the ordered sequence of physiological parameters and defining the time range.

[0051] The same spatiotemporal dimension refers to the situation where time overlaps and spatial location matches. Here, it refers to specific emergency beds or treatment areas. It is used to ensure that the associated abnormal signal interval and the target resource event occur in the same location and time period of the same patient. This is determined by comparing the overlap of timestamps and the consistency of location identifiers.

[0052] Target resource events refer to specific bed status change events or medical staff task assignment events that are selected from real-time records of medical resources and match a certain period of signal abnormality in both time and space. They are used to provide possible external operational explanations or clinical context for signal abnormalities and are obtained through spatiotemporal matching queries.

[0053] Preset change threshold: This refers to a pre-defined numerical limit used to determine whether the change between two consecutive measurements of a physiological parameter is abnormally drastic. For example, if a patient's current heart rate is 80 beats / min, and the heart rate is measured at 85 beats / min in the next second, the change is 5 beats / min, which is less than the threshold of 30 beats / min. This is considered a normal physiological fluctuation and will not be marked as abnormal. It is used to identify abrupt abnormalities and is pre-configured by domain experts based on physiological knowledge and equipment characteristics.

[0054] The first category of candidate anomalies refers to individual data acquisition time points in an ordered sequence of physiological parameters that are selected by comparing the differences between adjacent measurements, and whose changes exceed a preset threshold. These points are used to mark abnormal moments that may be caused by signal jumps, equipment interruptions, etc.

[0055] The second category of candidate anomalies refers to data points in an ordered sequence of physiological parameters that remain completely unchanged across multiple consecutive acquisition times by detecting the duration of consecutive identical values. These data points are used to mark abnormal platform segments that may be caused by equipment signal freezing, sensor detachment, or other reasons.

[0056] The abnormal signal time period refers to one or more continuous time intervals defined in the ordered sequence of physiological parameters, with the first or second type of candidate anomalies as the core, through time clustering and boundary expansion rules. These intervals are considered as an overall abnormality and are used to integrate discrete anomalies to extract anomaly periods with practical analytical significance.

[0057] The first target resource event refers to the first bed status change event or medical staff task allocation event that matches a specific abnormal signal time period in the ordered sequence of resource events, after filtering by both time and space conditions, and is used as the main candidate event for establishing a correlation.

[0058] In this step, the emergency process flow data is first read, and all data items marked as physiological parameters are extracted through the data type filter. Since the flow data itself is already sorted by time, these data items are simply stored in a new list in the original order to obtain an ordered sequence of physiological parameters. Next, the sliding window traversal algorithm is used to process the ordered sequence of physiological parameters. The absolute value of the difference between the measured values ​​of two adjacent data points in the ordered sequence of physiological parameters is calculated in turn. After each calculation, a threshold comparison function is called to compare the difference with a pre-loaded preset change threshold for the ordered sequence type of physiological parameters. If the difference exceeds the threshold, the timestamp of the next data point is recorded in the first type of candidate anomaly list, thus completing the detection of numerical jump anomalies. Next, a second round of scanning is performed on the ordered sequence of the same physiological parameters to detect signal freezing. The sliding window traversal algorithm maintains a counter. When two consecutive data points are found to be completely equal, the counter is incremented. When the number of consecutive equal points exceeds the preset minimum number of continuous points, such as 3 points, the time of all points will be recorded in the second type of candidate abnormal point list starting from the beginning of the continuous segment and the abnormality of the value remaining unchanged for a long time will be marked. Then, all timestamps of the first and second types of candidate anomalies are read and processed using a time density-based clustering algorithm. After sorting all timestamps, the time difference between adjacent points is calculated. If the time difference is less than the preset merging tolerance window, such as 5 seconds, these points are grouped into the same cluster. Then, based on the time boundary of each cluster, a fixed buffer time, such as 2 seconds, is extended forward and backward to generate a signal anomaly time period with a clear start and end time for each cluster. The emergency process flow data was then processed again, but an event type filter was used to extract all data items marked as resource events, forming an ordered sequence of resource events. Subsequently, for each abnormal signal time period, a two-condition search was performed in the ordered sequence of resource events. A composite index query technique based on time and location was used to quickly locate events that met two conditions: first, the occurrence time of the event fell within the start and end time range of the abnormal time period; second, the location identifier recorded in the event, such as the bed number, was completely consistent with the patient location identifier associated with the current analysis. The first event found that met both conditions was identified as the first target resource event. Finally, an association mapping is established to bind each abnormal signal time period with its corresponding first target resource event, generating an association record. This record together constitutes the target resource events that occur in the same spatiotemporal dimension, providing a clear clinical operational context for the abnormal signal.

[0059] For example, following the specific implementation of the previous step, firstly, an ordered sequence of physiological parameters is extracted from the emergency process flow data of patient B, and this ordered sequence of physiological parameters is traversed; secondly, when it is detected that the patient's heart rate suddenly increases from 110 beats / min to 155 beats / min within two consecutive seconds, and the change exceeds the preset change threshold of 30 beats / min / second, that second is immediately marked as a first-type candidate anomaly point; at the same time, it is found that the patient's blood pressure value remains unchanged at 85 / 50 mmHg for multiple consecutive sampling points, triggering the signal freeze detection rule, and this period is marked as multiple second-type candidate anomaly points; Next, these discrete outliers were merged. A temporal clustering algorithm was used to group nearby outliers together and extend the buffer time, ultimately forming two distinct signal anomaly time periods: a brief period of sudden heart rate changes and a sustained period of blood pressure signal stagnation. Subsequently, an ordered sequence of resource events for patient B was extracted from the same emergency process flow data. For the blood pressure signal stagnation period, a query was performed within the resource event sequence to find resource events that overlapped temporally with this period and whose location perfectly matched the patient's location. The query revealed that a nurse performing a central venous puncture task event temporally covered this outlier period and matched the location. Finally, this event was identified as the first target resource event related to this signal anomaly, and the two were associated and bound together to form a target resource event record indicating that the blood pressure signal stagnation occurred during the central venous puncture procedure.

[0060] Step 103: According to the predefined emergency clinical pathway, the emergency process flow data is divided into stages to obtain clinical stage units, which are used to reflect different treatment urgency levels.

[0061] Optionally, step 103 may specifically include: Step 1031: Extract the ordered sequence of resource events from the emergency process flow data.

[0062] Step 1032: Based on the predefined emergency clinical pathway, determine multiple standard medical node event types, including triage completion, first medication administration, outpatient examination, and hospitalization application.

[0063] Step 1033: In the ordered sequence of resource events, search for events that match the standard medical node event type and use them as stage separation events.

[0064] Step 1034: Record the time of occurrence of each stage separation event in the ordered sequence of resource events as the stage separation time.

[0065] Step 1035: Divide the emergency process flow data into continuous time segments according to the chronological order of the stage separation times.

[0066] Step 1036: Define each of the time segments as a clinical stage unit.

[0067] In this step, the predefined emergency clinical pathway refers to a standardized process model that is pre-developed and entered into the system by medical experts. It describes the key medical steps and sequence that an emergency patient should undergo from admission to discharge under ideal conditions. It is used to provide the computer with the logical basis for dividing the treatment stages. This is done by accessing the hospital's clinical knowledge base, such as: triage completed - first electrocardiogram completed - first administration of emergency medication - completion of echocardiography - hospitalization application.

[0068] Clinical phase units refer to consecutive time periods with specific clinical significance that are cut along the timeline of a patient's complete and continuous emergency treatment process according to a predefined emergency clinical pathway. These units are used to reflect the different treatment stages and urgency levels of the patient during the emergency period.

[0069] Standard medical node event types refer to the clearly defined, landmark event categories in predefined emergency clinical pathways that represent the completion of key clinical decisions or operations, and are used as search templates to identify phase transition points in time series.

[0070] Triage completion is a standard medical node event type, specifically referring to the event at the emergency department entrance where the initial assessment and classification of the severity of a patient's condition is completed, and the patient is assigned to a corresponding treatment area. This event marks the official start of the emergency treatment process.

[0071] First-time administration is a standard medical milestone event type, specifically referring to the event in which a patient is given the first intravenous, intramuscular, or oral medication during emergency room visits. It is used to mark a key turning point from the assessment and decision-making phase to the active treatment phase.

[0072] Out-of-department examination is a type of standard medical node event, specifically referring to the event where a patient temporarily leaves the emergency department to undergo examinations in other departments such as the CT room or ultrasound department due to diagnostic needs. It is used to mark the transition of a patient from leaving the primary monitoring environment to entering the examination process.

[0073] Hospitalization request is a type of standard medical event, specifically referring to the event in which a doctor formally initiates a hospitalization request for a patient. It is used to mark the beginning of the waiting period for hospitalization or the handover phase after the basic completion of emergency treatment.

[0074] A stage-separated event refers to a specific event instance that is identified in the ordered sequence of the patient's actual resource events and matches a certain standard medical node event type. It is used to accurately locate the specific point in time when the stage transition occurs on the timeline and is obtained by searching and matching in the event sequence.

[0075] The stage split time refers to the precise time of occurrence of each stage split event within its ordered sequence of resource events, used as the absolute time point coordinate for cutting emergency process flow data on the global time axis.

[0076] In this step, the emergency process flow data is first filtered by a data type filter, and all data items identified as resource events are selected according to the preset type identifier field in the data items. Since the emergency process flow data itself is strictly arranged in chronological order, these event data items are simply stored into a new list in the order of reading to obtain an ordered sequence of resource events. Next, the configuration file or knowledge base storing the emergency clinical pathway definition is accessed. By parsing the configuration file or knowledge base query, the key node code defined in the path, such as TRIAGE_COMPLETE, is extracted to form a standard medical node event type list. Then, an event type matching search is performed on the ordered sequence of resource events. A string exact matching algorithm is used to traverse each resource event in the sequence and compare the event type field of each resource event with the obtained standard type list one by one. When the string exact matching algorithm finds that the type field of an event is completely consistent with a type in the list, the match is determined to be successful, and the event instance is marked as a stage separator event. Then, for each marked stage separation event, the value of the event occurrence time field is read from its data structure through field extraction operation, and this time value is saved as an independent stage separation moment. All found separation events will generate corresponding separation moments, and then these moment points are sorted in ascending order using the quicksort algorithm. Then, the sorted list of stage separation times is processed using a time interval partitioning algorithm. The first time segment is drawn by taking the start time of the entire emergency process, i.e. the start time of the streaming data, as the initial starting point and the first separation time as the end point. Then, the previous separation time is taken as the starting point of the next segment and the next separation time as the end point, and the subsequent segments are drawn. This process is iterated until the last segment is drawn by taking the last separation time as the starting point and the end time of the consultation as the end point. Finally, through the stage semantic mapping rules, each divided time segment is assigned a preset clinical stage name. For example, the mapping rule defines the segment from the start to the completion of triage as the triage assessment stage. This time segment, which is given specific clinical semantics, is defined as a clinical stage unit.

[0077] For example, following the specific implementation of the previous step, firstly, an ordered sequence of resource events is extracted from the emergency process flow data of patient B, which includes events such as completing primary triage, performing central venous puncture, transferring the patient to the CT room, and submitting a hospitalization request; secondly, a predefined emergency clinical pathway model for trauma patients is loaded, and key standard medical node event types are extracted from it, including triage completion, first invasive procedure, outpatient examination, and hospitalization request; then, a matching search is performed in the resource event sequence to identify that the event of completing primary triage matches the triage completion type, performing central venous puncture matches the first invasive procedure, transferring the patient to the CT room matches the outpatient examination, and submitting a hospitalization request matches the hospitalization request. These events are then marked as key stage-separating events, and their occurrence times are recorded as stage-separating times. Subsequently, based on these separation times, the complete timeline of the patient from admission to discharge is divided into continuous time segments, such as the first segment from admission to triage completion, the second segment from triage completion to the first invasive procedure, and so on. Finally, each time segment is given a clinical definition to generate clinical stage units. The first segment is defined as the triage assessment stage, the second segment as the resuscitation and rescue stage, the third segment as the outpatient examination stage, and the fourth segment as the treatment completion and handover stage. Each unit clearly corresponds to a different stage of the treatment process.

[0078] Step 104: Within the clinical phase unit, generate quality control analysis focus by combining the identified abnormal signal intervals and associated target resource events.

[0079] Optionally, step 104 may specifically include: Step 1041: Within the time segment corresponding to the clinical stage unit, filter out the signal abnormality time period whose time range is completely contained within the time segment.

[0080] Step 1042: Obtain the target resource events that have been associated with each of the filtered signal anomalous time periods.

[0081] Step 1043: Determine whether there is any overlap in time between the abnormal signal time period and the time of occurrence of the event corresponding to the acquired target resource event.

[0082] Step 1044: When it is determined that there is a time overlap, the time overlapped signal abnormality time period and the target resource event are combined to form a quality control analysis focus.

[0083] In this step, the quality control analysis focus refers to a minimum analytical unit within a specific clinical stage unit, which is composed of an abnormal signal time period and its temporally overlapping, associated target resource events, and is used to encapsulate a problem scenario to be checked with clear spatiotemporal boundaries and clinical context. It is generated by combining abnormal time periods within the stage with associated events.

[0084] In this step, the current clinical stage unit is first operated on. By reading the start and end time fields in its data structure, its corresponding time segment is obtained. Then, all generated signal abnormal time segment lists are traversed, and a time inclusion judgment is performed on each time segment. This inclusion judgment is implemented through a comparison algorithm to check whether the start time of the abnormal time segment is greater than or equal to the start time of the clinical stage time segment, and at the same time, check whether the end time of the abnormal time segment is less than or equal to the end time of the clinical stage time segment. Only time segments that fully meet these two comparison conditions will be filtered out and stored in a special list of abnormalities to be processed for this stage. Secondly, for each selected abnormal signal time period, perform a correlation query, access the mapping table that stores the correlation between abnormal time periods and resource events, use the unique identifier of the current abnormal time period as the query key, and retrieve the complete data record of the target resource event bound to it from the mapping table; Next, time overlap verification is performed on each pair of abnormal time periods and target resource events. The start and end times of the abnormal signal time period and the event occurrence time in the target resource event record are read respectively. Through a simple logical judgment operation, it is verified whether the event occurrence time is greater than or equal to the start time of the abnormal time period and less than or equal to the end time of the abnormal time period. If the judgment condition is true, it is confirmed that there is time overlap between the two. Finally, after the time overlap verification is passed, the data encapsulation operation is performed. The start and end times, anomaly type, and other attributes of the current abnormal signal period are packaged together with the event type, content, and occurrence time of the target resource event and filled into a new, predefined structure object. This newly created structure object is then defined as a quality control analysis focus.

[0085] For example, following the specific implementation of the previous step, firstly, the clinical phase unit of Patient B's resuscitation and rescue stage is analyzed. Based on the time range of this clinical phase unit, a signal abnormality time period that occurs entirely within this clinical phase is selected from all abnormal time periods of Patient B, such as a period of ECG signal dropout. Secondly, the established association relationship is queried to obtain the target resource event pre-associated with this abnormal time period, such as an event record of a nurse performing a patient repositioning. Next, time overlap verification is performed to determine whether the specific time of the nurse performing the patient repositioning event falls within the time interval of the ECG signal dropout period. After verification, the two overlap in time. Then, the time-overlapping ECG signal dropout period and the nurse performing the patient repositioning event are combined and encapsulated to generate an independent quality control analysis focus. Finally, this analysis focus is clearly pointed to the specific scenario of ECG signal dropout occurring during the patient repositioning operation in the resuscitation and rescue stage, for subsequent special compliance verification.

[0086] Step 105: Based on the clinical scenario characteristics included in the quality control analysis focus, select logical judgment rules corresponding to the clinical scenario characteristics from the preset emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus.

[0087] Optionally, step 105 may specifically include: Step 1051: Extract signal anomaly type features from the signal anomaly time periods included in the quality control analysis focus.

[0088] Step 1052: Extract stage type features from the clinical stage unit where the quality control analysis focus is located.

[0089] Step 1053: Extract resource event type features from the target resource events included in the quality control analysis focus.

[0090] Step 1054: Combine the extracted signal anomaly type features, stage type features, and resource event type features to generate combined features.

[0091] Step 1055: Search the logical judgment rule base in the emergency quality control standard knowledge base. Each logical judgment rule in the logical judgment rule base has a corresponding rule trigger feature.

[0092] Step 1056: Match the combined features with the preset rule triggering features of each logical judgment rule.

[0093] Step 1057: Select logical judgment rules that match the combined features as target logical judgment rules.

[0094] Step 1058: Arrange the selected target logic judgment rules according to the predefined rule execution order to generate quality control logic.

[0095] In this step, clinical scenario features refer to the set of key attributes that can summarize the specific clinical situation from a quality control analysis focus, used to accurately describe the complete scenario at what stage, what abnormality occurred, and what operations were being performed.

[0096] The pre-built emergency quality control standard knowledge base refers to a digital knowledge set that is pre-built and stored in the system and contains a large number of emergency medical quality inspection rules. It is used to provide a basis for judgment for automated quality control. It is established by encoding clinical guidelines, operating procedures and quality standards into computer-executable rules. For example: IF The duration of the abnormality is less than 30 seconds AND the turning operation record is double-checked by two nurses THEN It is determined to be an acceptable physiological signal interference; ELSE It is determined to be an abnormal equipment connection or non-standard operation, which needs to be checked.

[0097] Logical judgment rules refer to specific, executable if-then judgment logic stored in the emergency quality control standard knowledge base. They are used to perform compliance checks on data and events in a specific clinical scenario and are pre-written and entered by knowledge engineers according to medical standards.

[0098] Quality control logic refers to a sequence of rules, consisting of one or more logical judgment rules dynamically selected from the knowledge base and arranged in a specific order, to address the specific quality control analysis focus at the moment. This sequence guides the system to perform step-by-step, specialized, automated checks on the focus.

[0099] Signal anomaly type features refer to classification labels extracted from the signal anomaly time periods included in the quality control analysis focus, used to identify abnormal data patterns, such as signal jumps or signal freezes, and used to characterize the nature of data quality problems.

[0100] Stage type characteristics refer to the classification labels extracted from the clinical stage unit where the quality control analysis focus is located, used to identify the treatment process, such as the triage assessment stage or the resuscitation stage, to characterize the stage of diagnosis and treatment when the abnormality occurs.

[0101] Resource event type features refer to the classification labels extracted from the target resource events included in the quality control analysis focus, used to identify the type of medical operation, such as drug administration or repositioning, and used to characterize the clinical behavior in progress when the abnormality occurs.

[0102] A combined feature is a comprehensive feature identifier formed by connecting the extracted signal anomaly type features, stage type features, and resource event type features. It is used to uniquely map to specific rules in the knowledge base for processing such complex scenarios.

[0103] Rule triggering features refer to a pre-defined feature condition for each logical judgment rule in the emergency quality control standard knowledge base. Its format is the same as that of combination features. It is used to define the specific scenario to which the rule applies and is set by the knowledge engineer when the rule is created.

[0104] The target logical judgment rule refers to one or more rules selected from the logical judgment rule base of the knowledge base, whose rule triggering characteristics match the combined characteristics of the current quality control analysis focus. These rules are the rules that will be used for actual verification and are selected through feature matching algorithms.

[0105] The predefined rule execution order refers to the order in which multiple target logic judgment rules are sorted when generating quality control logic. This order is used to ensure the logic and efficiency of the verification judgment. It is determined by the priority attribute of the rule itself or the preset dependency relationship. For example, in the scenario of blood pressure signal freezing during central venous puncture in the resuscitation stage, it is determined whether the operation is allowed to cause the blood pressure signal to disappear temporarily. If it is allowed, it is directly judged as compliant and the process ends. If rule A determines that it is not allowed, rule B is triggered to further determine whether the duration of signal freezing exceeds the maximum interference time allowed for the operation.

[0106] In this step, the data structure of the quality control analysis focus is first parsed. Using field extraction operations, the abnormality type attribute values ​​are read from the signal abnormality time period objects contained in the quality control analysis focus data to obtain the signal abnormality type characteristics; the stage type attribute values ​​are read from the clinical stage unit objects associated with the quality control analysis focus to obtain the stage type characteristics; and the event type attribute values ​​are read from the target resource event objects contained in the focus to obtain the resource event type characteristics. Secondly, a string concatenation algorithm is used to concatenate the three feature strings with a specific delimiter according to a predefined format template, generating a combined feature string that uniquely identifies the current complex scenario; then, the pre-set emergency quality control standard knowledge base is accessed, and the logical judgment rule library storing all judgment rules is located and loaded through a database query request. Each logical judgment rule in this logical judgment rule library is associated with a preset rule trigger feature string. Then, a string exact matching algorithm is used to compare the generated combined feature string with the rule trigger feature string of each rule in the rule base one by one. The comparison is performed character by character. When the two strings are completely identical, it is determined that the match is successful, and the corresponding rule is marked as the target logic judgment rule. Then, all the selected target logic judgment rules are sorted. The priority value attribute defined in each rule is read, and the quick sorting algorithm is used to sort the entire list in ascending order according to the priority value, thereby determining the execution order of the rules. Finally, the sorted sequence of target logic judgment rules is encapsulated into an ordered executable instruction set, that is, the quality control logic for the current focus is generated.

[0107] For example, following a specific implementation of the previous solution, the quality control analysis focuses on the simultaneous occurrence of ECG lead dropout and patient repositioning during the resuscitation phase. Key features are extracted from this focus: the signal abnormality type feature is ECG_LEAD_OFF (ECG lead dropout), the phase type feature is RESUSCITATION (resuscitation), and the resource event type feature is REPOSITIONING (patient repositioning). Next, these three feature values ​​are concatenated into strings according to a predefined phase abnormality operation format to generate the combined feature RESUSCITATION_ECG_LEAD_OFF_REPOSITIONING. Then, a search and matching operation is performed in the rule base of a pre-defined emergency quality control standard knowledge base. The generated combined feature string is compared with the rule trigger features of each rule in the base. If a rule with identical trigger features is matched, that rule is selected as the target logic judgment rule. Finally, this matched rule is organized to generate a quality control logic specifically used to determine whether ECG signal dropout during patient repositioning during the resuscitation phase is compliant.

[0108] Step 106: Execute the automated calculation of the quality control logic on the quality control analysis focus to output a quality control anomaly event record marked with a quality control anomaly event.

[0109] Optionally, step 106 specifically includes: Step 1061: From the quality control analysis focus, obtain ordered sequence fragments of physiological parameters containing abnormal signal time periods, and ordered sequence fragments of resource events containing target resource events.

[0110] Step 1062: Extract the first target logic judgment rule in the quality control logic according to the execution order of the predefined rules in the quality control logic.

[0111] Step 1063: Substitute the ordered sequence fragments of physiological parameters and the ordered sequence fragments of resource events into the judgment conditions of the first target logic judgment rule for comparison and calculation to obtain the first intermediate judgment result.

[0112] Step 1064: Determine whether the first intermediate judgment result satisfies the conclusion condition of the first target logic judgment rule.

[0113] Step 1065: If the conclusion of the first target logic judgment rule is satisfied, then record the conclusion corresponding to the first target logic judgment rule as the first intermediate conclusion.

[0114] Step 1066: Repeat the steps of extracting the target logic judgment rules from the quality control logic, substituting and comparing the calculations, judging whether the conditions for the conclusion to be established are met, and recording the conclusion, until all the target logic judgment rules in the quality control logic have been executed to obtain the intermediate conclusion combination.

[0115] Step 1067: Based on the intermediate conclusions, generate final conclusions targeting the quality control analysis focus.

[0116] Step 1068: Associate and bind the quality control analysis focus, the quality control logic, and the final conclusion to generate a quality control anomaly event record.

[0117] In this step, a quality control anomaly event refers to a specific event scenario that has been formally determined to have a quality problem or require clinical attention after automated calculation by the quality control logic. It is used to identify a complete, analyzed instance of a quality control problem, and is obtained by performing automated calculations on the focus of the quality control analysis and generating conclusions.

[0118] A quality control anomaly log is a structured data object that links the original quality control analysis focus data, the complete quality control logic used in the calculation, and the final conclusions drawn from the calculation. It is used to record the process and results of an automated quality control analysis in a complete and traceable manner.

[0119] The first intermediate judgment result refers to the preliminary numerical or logical value output obtained after substituting the data fragment of the focus of quality control analysis into the first rule in the quality control logic for condition calculation. It is used to reflect the preliminary judgment situation under a single rule and is obtained by executing the judgment function of the rule.

[0120] The first intermediate conclusion refers to the judgment statement with clear semantics output from the corresponding rule when the first intermediate judgment result meets the conclusion conditions of the rule. It is used to record the verification opinion of a single rule and is obtained by triggering the rule conclusion output after the conditions are met.

[0121] Intermediate conclusions refer to a list of all intermediate conclusions collected after all rules in the quality control logic have been executed in sequence. This list is used to comprehensively reflect the set of independent judgment results of each rule on the same focus.

[0122] The final conclusion for a quality control analysis focus refers to the final and overall quality control judgment derived from the combination of intermediate conclusions and a comprehensive analysis according to a preset conclusion aggregation rule. It is used to give the final state label of the event and is obtained by processing the set of intermediate conclusions through a conclusion aggregation function.

[0123] In this step, firstly, a data slicing operation is performed on the quality control analysis focus. Based on the start and end timestamps of the signal abnormality time period in the quality control analysis focus, all data points within that time period are extracted from the original ordered sequence of physiological parameters to generate an ordered sequence fragment of physiological parameters. At the same time, the complete record of the target resource event contained in the quality control analysis focus is extracted as an ordered sequence fragment of resource event. Secondly, the sequential execution engine for the quality control logic is started, and the quality control logic data structure is read. This is an ordered list of target logic judgment rules. An index pointer is initialized to point to the beginning position of the list, and the first target logic judgment rule is extracted through pointer read operation. Next, the rule calculation function is called, taking the ordered sequence fragments of physiological parameters and the ordered sequence fragments of resource events as input parameters and passing them into the judgment condition function predefined by the current rule. This condition function executes the internally defined calculation logic, such as performing numerical comparisons or logical operations, and outputs a first intermediate judgment result. Then, the condition adjudication logic is executed, reading the conclusion condition of the current rule and comparing the first intermediate judgment result with the condition. If the comparison result meets the condition, the rule is triggered to output its predefined conclusion text, and the text is saved as the first intermediate conclusion through data recording operations. The process then enters a loop control flow, incrementing the index pointer to retrieve the next target logic judgment rule and repeatedly executing the calculation function, conditional decision, and data recording operations. This loop iteration continues until all rules in the list have been processed. At this point, the output conclusions of all rules are collected, forming an intermediate conclusion set. Subsequently, the conclusion aggregation function is called to process the intermediate conclusion set, performing a comprehensive analysis of all intermediate conclusions according to a preset aggregation strategy, such as using veto or consensus voting logic, to derive a single final conclusion. Finally, data encapsulation and persistence operations are performed, creating a new structured record that associates and binds the original quality control analysis focus, the complete quality control logic list on which the execution process was based, and the generated final conclusion, generating a complete quality control anomaly event record, which is then output to the storage system.

[0124] For example, following the specific implementation of the previous step, firstly, the quality control analysis focus of patient B during the resuscitation phase is obtained, including ECG lead detachment, repositioning, and the quality control logic generated by matching, which includes a judgment rule; secondly, the required data is extracted from this quality control analysis focus: based on the start and end times of the ECG lead detachment period, data for the corresponding time period is extracted from the complete ECG waveform data to form an ordered sequence fragment of physiological parameters; simultaneously, a complete description of the repositioning event performed under double-checking is extracted to form an ordered sequence fragment of resource events; then, the quality control logic is executed, extracting the unique target logic judgment rule, and substituting the physiological and resource data fragments into the judgment conditions of the target logic rule for calculation; then, the information is calculated. The total duration of the signal detachment is calculated, and it is verified whether the operation was performed by two people to obtain the first intermediate judgment result. For example, if the duration is equal to 20 seconds, it is a two-person operation. Then, it is judged whether this result meets the conclusion conditions of the rule, such as: the duration is less than 30 seconds and it is a two-person operation. When the conditions are met, the conclusion corresponding to the rule is recorded as the first intermediate conclusion: the operation-related signal interference is acceptable. Since there is only this one rule in the quality control logic, the first intermediate conclusion is directly used as the final conclusion. Finally, the original quality control analysis focus, the quality control logic used for execution, and the final conclusion of the operation-related signal interference being acceptable are associated and bound to generate a complete quality control anomaly event record, which is the final output of the automated calculation.

[0125] Figure 2 This application provides a schematic diagram of the structure of an automated processing system for emergency data quality control analysis, as shown below. Figure 2 As shown, the system includes: The acquisition module 21 is used to acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data. The identification module 22 is used to identify the abnormal signal interval in the time series data based on the emergency process flow data, and associate it with the target resource events that occur in the same spatiotemporal dimension in the real-time medical resource record; The segmentation module 23 is used to segment the emergency process flow data into stages according to a predefined emergency clinical pathway to obtain clinical stage units, which are used to reflect different treatment urgency levels. The generation module 24 is used to generate a quality control analysis focus within the clinical stage unit by combining the identified abnormal signal intervals and associated target resource events. The filtering module 25 is used to filter logical judgment rules corresponding to the clinical scenario features contained in the quality control analysis focus from a preset emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus. The calculation module 26 is used to perform automated calculations of the quality control logic on the quality control analysis focus to output a quality control abnormality event record marked with a quality control abnormality event.

[0126] Figure 2 The aforementioned automated processing system for emergency data quality control and analysis can perform... Figure 1 The implementation principle and technical effects of the automated processing method for emergency data quality control analysis described in the illustrated embodiment will not be repeated here. The specific methods by which each module and unit performs operations in the automated processing system for emergency data quality control analysis described in the above embodiments have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0127] In one possible design, Figure 2 An automated processing system for emergency data quality control analysis, as shown in the embodiment, can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0128] The processing component 32 is used for the above Figure 1 The embodiment describes an automated processing method for emergency data quality control analysis.

[0129] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An automated processing method for quality control analysis of emergency department data, characterized in that, include: Acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data; Based on the emergency process flow data, identify the abnormal signal intervals in the time series data and associate them with target resource events that occur in the same spatiotemporal dimension in the real-time medical resource records; Based on a predefined emergency clinical pathway, the emergency process flow data is divided into stages to obtain clinical stage units, which are used to reflect different treatment urgency levels. Within the clinical phase unit, a quality control analysis focus is generated by combining the identified abnormal signal intervals and associated target resource events; Based on the clinical scenario characteristics included in the quality control analysis focus, logical judgment rules corresponding to the clinical scenario characteristics are selected from the pre-set emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus. The quality control logic is automatically calculated for the quality control analysis focus to output a quality control anomaly event record marked with a quality control anomaly event.

2. The method according to claim 1, characterized in that, Acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data, including: Physiological parameter measurements were collected from monitoring equipment of emergency patients, and the corresponding collection time was marked for each physiological parameter measurement to obtain time series data; Real-time acquisition of bed status change events and medical staff task allocation events from the emergency department resource management system, and marking the corresponding event occurrence time for each event to obtain real-time records of medical resources; All physiological parameter measurements in the time series data are arranged in chronological order of the acquisition times to form an ordered sequence of physiological parameters. All events recorded in the real-time medical resource data are arranged in chronological order of their occurrence to form an ordered sequence of resource events. The ordered sequence of physiological parameters and the ordered sequence of resource events are interleaved and combined in chronological order to form emergency process flow data.

3. The method according to claim 1, characterized in that, Based on the emergency process flow data, identify abnormal signal intervals in the time series data and correlate them with target resource events occurring in the same spatiotemporal dimension in the real-time medical resource records, including: Extract ordered sequences of physiological parameters from the emergency process flow data; In the ordered sequence of physiological parameters, points where the change in physiological parameter measurement values ​​between adjacent acquisition times exceeds a preset change threshold are detected and identified as first-type candidate anomalies. In the ordered sequence of physiological parameters, points where the same physiological parameter measurement value appears consecutively are detected as second-type candidate abnormal points; Based on the first type of candidate anomalies and the second type of candidate anomalies, the signal abnormality time period is determined in the ordered sequence of physiological parameters; Extract the ordered sequence of resource events from the emergency process flow data; In the ordered sequence of resource events, find the first target resource event that overlaps in time with the time period of the abnormal signal and whose location identifier matches the location of the monitoring device that generated the ordered sequence of physiological parameters; The first target resource event is associated with the time period of the signal anomaly, and thus considered as a target resource event occurring in the same spatiotemporal dimension.

4. The method according to claim 1, characterized in that, Based on a predefined emergency clinical pathway, the emergency process flow data is divided into stages to obtain clinical stage units. These clinical stage units reflect different levels of treatment urgency and include: Extract the ordered sequence of resource events from the emergency process flow data; Based on a predefined emergency clinical pathway, several standard medical node event types are identified, including triage completion, first medication administration, outpatient examination, and hospitalization application. In the ordered sequence of resource events, events that match the standard medical node event type are searched and used as stage-separating events; Record the occurrence time of each phase separation event in the ordered sequence of resource events, and use it as the phase separation time; According to the chronological order of the stage separation times, continuous time segments are divided in the emergency process flow data; Each of the aforementioned time segments is defined as a clinical phase unit.

5. The method according to claim 1, characterized in that, Within the clinical phase unit, by combining the identified abnormal signal intervals and associated target resource events, a quality control analysis focus is generated, including: Within the time segment corresponding to the clinical stage unit, filter out signal abnormality time periods whose time range is completely contained within the time segment; Acquire target resource events that are associated with each time period of signal anomaly that has been filtered; Determine whether there is any overlap in time between the abnormal signal time period and the time of occurrence of the event corresponding to the acquired target resource event; When it is determined that there is a time overlap, the abnormal time periods of the signal that overlap in time and the target resource event are combined to form a quality control analysis focus.

6. The method according to claim 1, characterized in that, Based on the clinical scenario characteristics included in the quality control analysis focus, logical judgment rules corresponding to the clinical scenario characteristics are selected from a pre-set emergency quality control standard knowledge base to generate quality control logic, including: Extract signal anomaly type features from the signal anomaly time periods included in the quality control analysis focus; Extract stage type features from the clinical stage unit where the quality control analysis focus is located; Extract resource event type features from the target resource events included in the quality control analysis focus; The extracted signal anomaly type features, stage type features, and resource event type features are combined to generate combined features; In the emergency quality control standard knowledge base, search the logical judgment rule base. Each logical judgment rule in the logical judgment rule base has a corresponding rule trigger feature pre-set. The combined features are matched with the pre-defined rule triggering features of each logical judgment rule; Filter out the logical judgment rules that match the combined features, and use them as target logical judgment rules; The selected target logic judgment rules are arranged according to the predefined rule execution order to generate quality control logic.

7. The method according to claim 1, characterized in that, The quality control logic is automatically calculated for the quality control analysis focus to output a quality control anomaly event record marked with the quality control anomaly event, including: From the quality control analysis focus, obtain ordered sequence fragments of physiological parameters containing abnormal signal time periods, and ordered sequence fragments of resource events containing target resource events; Extract the first target logic judgment rule in the quality control logic according to the execution order of the predefined rules in the quality control logic; The ordered sequence fragments of the physiological parameters and the ordered sequence fragments of the resource events are substituted into the judgment conditions of the first target logic judgment rule for comparison and calculation to obtain the first intermediate judgment result; Determine whether the first intermediate judgment result satisfies the conclusion condition of the first target logic judgment rule; If the conclusion of the first target logic judgment rule is satisfied, then the conclusion corresponding to the first target logic judgment rule is recorded as the first intermediate conclusion. Repeat the steps of extracting the target logic judgment rules from the quality control logic, substituting and comparing the calculations, judging whether the conditions for the conclusion to be valid, and recording the conclusion, until all the target logic judgment rules in the quality control logic have been executed, so as to obtain the intermediate conclusion combination. Based on the aforementioned intermediate conclusions, a final conclusion is generated targeting the focus of the quality control analysis. The quality control analysis focus, the quality control logic, and the final conclusion are linked and bound together to generate a quality control anomaly event record.

8. An automated processing system for quality control and analysis of emergency department data, characterized in that, include: The acquisition module is used to acquire time-series data of emergency patients during their visit and real-time records of medical resources to generate emergency process flow data; The identification module is used to identify abnormal signal intervals in the time series data based on the emergency process flow data, and associate them with target resource events that occur in the same spatiotemporal dimension in the real-time medical resource record. The segmentation module is used to divide the emergency process flow data into stages according to the predefined emergency clinical pathway to obtain clinical stage units, which are used to reflect different treatment urgency levels. The generation module is used to generate quality control analysis focus within the clinical stage unit by combining the identified abnormal signal intervals and associated target resource events; The filtering module is used to filter logical judgment rules corresponding to the clinical scenario characteristics contained in the quality control analysis focus from a pre-set emergency quality control standard knowledge base to generate quality control logic. The quality control logic is used to perform special verification on the quality control analysis focus. The calculation module is used to perform automated calculations of the quality control logic on the quality control analysis focus to output quality control abnormal event records marked with quality control abnormal events.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement an automated processing method for emergency data quality control analysis as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The system contains a computer program that, when executed by a computer, implements an automated processing method for emergency data quality control analysis as described in any one of claims 1 to 7.

Images