Urine metering data synchronous nursing management system suitable for closed pipeline

Abstract

The application discloses urine metering data synchronous nursing management system suitable for closed pipeline, relates to medical data processing technical field, and includes the trend segmentation module, carries out adaptive trend segmentation to urine metering time sequence based on patient postoperative days and diuretic medication state; the boundary detection module traverses the segmentation result in reverse with the nursing shift time as the anchor point, determines the trend state boundary in combination with the urine volume data illusion filtering rule caused by nursing operation and generates the handover information item; the time limit topology module derives the clinical failure time with the urine trend slope and the alert threshold distance, constructs the multi-patient time limit topology network and executes the cascade conflict prediction; the dual-state synchronous module triggers the snapshot freezing based on the urine data fluctuation amplitude and executes the adaptive interruption decision. The application realizes the automatic extraction of urine trend semantics, the multi-patient time limit priority sequencing and the shift view synchronization in the nursing shift scene, and improves the accuracy and timeliness of nursing management.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, and in particular to a synchronous nursing management system for urine metering data applicable to closed-loop systems. Background Technology

[0002] In clinical nursing, patients in intensive care and postoperative observation settings often have indwelling urinary catheters connected to closed urine metering lines. Urine volume is a key indicator for assessing fluid balance, renal function, and circulatory status. Nursing staff work in shifts, and the complete transmission of urine trend information during shift handover is directly related to the continuity of clinical decision-making.

[0003] Currently, nursing management of urine measurement data mainly relies on automated urine meters to upload measurement data to the hospital information system. However, this data upload only transmits raw numerical values ​​and lacks the ability to extract trend semantics and adapt to handover scenarios. Furthermore, while the hospital information system can display urine volume trend curves, it cannot distinguish between fluctuations in urine volume caused by nursing procedures such as turning over or flushing, and the patient's actual clinical trend changes. This results in a mixture of operational illusions and pathological trends, interfering with nursing judgment. Moreover, existing nursing management systems treat handover as a process node, but urine measurement data analysis and handover information transmission are disconnected. They cannot embed urine trend semantics into the handover process, nor can they generate inspection priorities and cascading risk predictions based on the urgency of multiple patients' urine trends. Additionally, when both parties view the same patient's urine data on different terminals, the time difference in data refresh leads to inconsistencies in the time sequence of their views, affecting the alignment of handover information.

[0004] Existing urine volume management technologies are unable to automatically extract urine trend semantics, filter out nursing operation artifacts, generate time-priority data for multiple patients, and ensure data visibility for both parties during nursing handover scenarios. Figure 1 Therefore, this invention proposes a synchronous nursing management system for urine metering data applicable to closed-loop pipelines. Summary of the Invention

[0005] The purpose of this invention is to provide a synchronous nursing management system for urine measurement data suitable for closed-loop systems, which solves the problems of existing urine measurement management technologies in automatically extracting urine trend semantics, filtering out nursing operation artifacts, generating time priority for multiple patients, and ensuring data visibility for both parties during nursing handover scenarios. Figure 1 The problem.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: This invention is a synchronous nursing management system for urine measurement data applicable to closed-loop tubing, comprising: The trend segmentation module is used to calculate the segmentation sensitivity coefficient of the patient's urine data based on the number of days after surgery and the diuretic medication status, and to perform trend segmentation on the urine measurement time series collected by the closed tubing using the segmentation sensitivity coefficient. The boundary detection module is used to reversely traverse the segmented results with the nursing handover time as the anchor point, combine the urine volume data illusion filtering rules caused by nursing operations to determine the starting boundary of the current urine trend state, define the interval from the anchor point to the boundary time as the handover window and generate urine trend handover information items. The time-effect topology module is used to construct the time-effect topology network of all patients under the care of the nurse taking over the shift. It derives the clinical failure time of each handover information item by the distance between the urine trend slope and the urine volume warning threshold, and performs global failure time recalculation and cascading conflict prediction driven by the deterioration of urine volume trend when urine volume trend changes or nursing behavior events are triggered. The dual-state synchronization module is used to monitor the current fluctuation range of urine measurement time series during shift handover. When the fluctuation range falls back to the preset stable threshold range, it triggers the freezing of urine measurement data snapshot and pushes the frozen data synchronously to the terminals of both parties during shift handover. During on-call state, it performs adaptive interruption decision based on the clinical delay cost of urine volume trend warning and the nurse's attention status. The closed-loop calibration module is used to reconcile the urine trend judgment records of each module with the actual urine volume data one by one after the end of the shift, trace the root cause parameters of deviation along the step chain and perform targeted calibration.

[0007] Furthermore, in the trend segmentation module, the segmentation sensitivity coefficient... The calculation method is as follows:

[0008] in, Based on sensitivity, This is a decay function based on the patient's disease type and the number of days post-surgery; the shorter the number of days post-surgery, the more... The larger the kidney, the more likely it is to be post-kidney surgery. The rate of decline is lower than that of non-kidney surgery. As a drug sensitivity factor based on diuretic use status, when using diuretics When not in use , Based on nursing risk level mapping factors, This is a normalization factor based on the normal fluctuation range of the patient's historical urine output; the greater the historical fluctuation range, the better. The smaller; Trend segmentation is based on the change in the first derivative between adjacent time windows. and The comparison results are used to perform the splitting, when At that time, segmentation is performed at the corresponding moment.

[0009] Furthermore, during the reverse traversal process, the boundary detection module further includes the following rules for filtering spurious urine data: First, query the set of nursing operation records within the current shift. If the time difference between the candidate boundary time and any nursing operation record time falls within the preset operation and data response deviation threshold... If the boundary is within the range, the candidate boundary is determined to be an illusory boundary of urine volume data caused by nursing operation and is excluded, and the reverse traversal continues; Next, query the patient's current medication records. When the candidate boundary time falls within the pharmacological onset delay range of any diuretic drug... If the boundary value is within the specified range, the candidate boundary is determined to be a pharmacologically driven boundary. Instead of using the candidate boundary time as the trend state boundary, the process continues to traverse backwards to the drug administration time. and will Marked as the source point for trend state triggering; When the candidate boundary time falls within the preset window of urine physiological rhythm changes, Temporary attenuation is:

[0010] in, The preset rhythm attenuation coefficient, and , with attenuation Reassess whether the candidate boundary is a valid trend state boundary.

[0011] Furthermore, the derivation method for the clinical failure time in the time-sensitivity topology module is as follows: For handover information items describing urine trend direction and the risk of crossing warning thresholds, their clinical expiration time is... for:

[0012] Where d is the absolute distance between the current urine volume and the warning threshold, and k is the slope of the trend segment element within the transition window. For nursing shift handover time; when hour, To preset the minimum effective slope threshold,

[0013] in, The preset steady-state default validity period applies when the direction of k is away from the warning threshold. Take an infinite value and mark it as valid indefinitely.

[0014] Furthermore, the cascading conflict prediction in the time-sensitive topology module includes: After recalculating the global failure time, patient pairs with temporal conflicts are detected in ascending order of failure time. For patient pairs with conflicts, the simulation nurse prioritizes the treatment of patients ranked higher, causing patients ranked lower to time out. The urine output trend of patients ranked lower deteriorates during the period without intervention. That is, the new slope after deterioration is predicted based on the current slope. At the same time, the next failure time of patients ranked lower is recalculated based on the new slope after deterioration. The system detects whether the next failure time will conflict with the failure times of other patients in the network. When the conflict propagation depth reaches two or more levels, a cascading conflict warning is generated and the complete propagation path containing the patient identifiers and corresponding failure times of each propagation level is recorded.

[0015] Furthermore, in the dual-state synchronization module, the current fluctuation range of the urine measurement time series is obtained by calculating the variance and standard deviation of the urine volume data within a preset sliding window. The preset stable threshold interval is the statistical distribution confidence interval of the fluctuation range of the patient within a stable historical period. When the fluctuation range of N consecutive sliding windows falls back to the stable threshold interval, a snapshot freeze is triggered. Clinical delay costs in adaptive interruption decision-making under on-call status Based on the absolute value of the current urine volume trend slope Calculated based on the remaining distance d between the urine volume value and the warning threshold:

[0016] The nurse's attention state is inferred based on the nurse's current operating state on the terminal. An interruption push is only executed when the clinical delay cost exceeds the interruption threshold mapped by the nurse's attention state; otherwise, the warning is added to the delayed push queue.

[0017] Furthermore, the root cause parameters for deviations traced along the step chain in the closed-loop calibration module include the following: For each deviation record, trace back upstream from the module from which the deviation record originated, and check the contribution of the intermediate values ​​output by each upstream module to the deviation in turn. The upstream module parameter with the largest contribution is identified as the principal cause parameter of the deviation. Targeted calibration is performed on the principal cause parameter. The calibration magnitude is positively correlated with the contribution weight of the upstream module parameter in the deviation and the amount of deviation. After calibration, the value of the upstream module parameter does not exceed the preset clinically reasonable value range.

[0018] Furthermore, in the dual-state synchronization module, the handover state completes the closed loop after both parties complete the confirmation and awareness operations on their respective terminals and then switches to the on-call state. If either party fails to complete the corresponding operation within the preset time limit, the system executes a step-by-step escalation: the first level pushes a closed loop reminder to the terminal of the party that did not complete the operation; the second level pushes a non-closed loop prompt and a list of the party's role and patients who exceeded the time limit to the nursing team leader's terminal; and the third level records the non-closed loop event in the handover exception log and forcibly terminates the handover state and enters the on-call state.

[0019] Furthermore, the time-sensitivity topology module is also used to: after recalculating the global failure time, detect whether there are multiple patients’ clinical failure times clustered in the same preset density time window in the time-sensitivity topology network. When the number of clustered patients exceeds the preset density threshold, it is determined as a decision density peak. For each patient within the decision density peak window, the clinical safety advance amount is calculated for each patient. The clinical safety advance amount is the ratio of the remaining distance between the patient's current urine output value and the warning threshold to the absolute value of the urine trend slope, minus the preset safety margin duration. It represents the maximum allowable time to advance the intervention time from the clinical failure time without causing additional clinical risks to the patient.

[0020] Furthermore, based on the clinical safety advance amount for each patient, a suggested sequence for advancing the intervention timing is generated. The suggested intervention time for patients with positive clinical safety advance amount is shifted forward from the original failure time, so that the number of instantaneous decisions within the aggregation window is reduced to below the density threshold.

[0021] The present invention has the following beneficial effects: This invention is based on adaptive trend segmentation of the patient's postoperative days and diuretic medication status. Combined with reverse traversal with the handover time as the anchor point and filtering of urine volume data artifacts caused by nursing operations, it can automatically extract urine trend semantics and generate handover information items in nursing handover scenarios. This avoids the problem of trend recognition loss caused by the transmission of raw numerical data, as well as the problem of operation artifacts and pathological trends interfering with nursing judgment. This improves the completeness and clinical accuracy of handover information transmission.

[0022] This invention derives the clinical failure time by measuring the distance between the urine trend slope and the urine volume warning threshold, constructs a multi-patient time-sensitive topology network, and performs cascade conflict prediction. This enables the incoming nurse to obtain dynamic patrol priorities and cascade risk predictions based on the urgency of urine trends in multi-patient scenarios, reducing the clinical risks caused by decision delays.

[0023] This invention monitors the real-time fluctuation of urine measurement data and uses the data's own stable window as a trigger condition to freeze the handover snapshot, ensuring the temporal consistency of the data views of both parties' terminals and avoiding inconsistencies caused by data refresh time differences. At the same time, through prediction and actual item-by-item reconciliation and root cause tracing calibration after the end of the shift, it has the ability to adaptively optimize for different patient types. Attached Figure Description

[0024] 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.

[0025] Figure 1 This is a block diagram of the overall architecture of the system of the present invention.

[0026] Figure 2 This is a schematic diagram of the processing flow of the trend segmentation module and the boundary detection module of the present invention.

[0027] Figure 3 This is a schematic diagram of the multi-patient time-response topology network and cascade conflict prediction in the time-response topology module of the present invention.

[0028] Figure 4 This is a schematic diagram of the dual-state switching between shift handover and duty states and the feedback process of the closed-loop calibration module of the present invention. Detailed Implementation

[0029] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0030] Example 1 See Figure 1-4 This invention is a synchronous nursing management system for urine measurement data applicable to closed-loop pipelines. It is interconnected with hospital information systems, nursing information systems, and medical order systems through data interfaces to acquire patient clinical data in real time. Each module operates collaboratively according to the data flow link. The urine measurement time series first enters the trend segmentation module for adaptive trend segmentation. The segmentation results are transmitted to the boundary detection module for artifact filtering and handover window definition. The handover information items enter the timeliness topology module for multi-patient timeliness analysis and conflict prediction. The early warning information enters the dual-state synchronization module for shift handover synchronization and shift interruption push. The judgment records of each module are summarized to the closed-loop calibration module for reconciliation and calibration after the shift ends.

[0031] In this embodiment, the processed urine measurement time series is processed by a digital urine measurement device connected in series at the end of a closed urine measurement pipeline. The measurement device adopts the differential pressure or weighing measurement principle, and collects the cumulative volume value of urine with device number, timestamp and patient identification at preset sampling intervals, and uploads it to the nursing data server through the ward wired network or wireless network.

[0032] The nursing data server establishes real-time data connections with the following external systems through data interfaces: after connecting with the electronic medical record module of the hospital information system, it obtains the patient's surgery type and postoperative days; after connecting with the medical order system, it obtains the medication records of diuretics, including the drug name, administration time, and the lower and upper limits of the pharmacological onset delay interval; and after connecting with the nursing information system, it obtains nursing operation records, including the operation type and time of turning over, tubing flushing, and body position adjustment.

[0033] The specific contents of each module in this embodiment include: The trend segmentation module is used to calculate the segmentation sensitivity coefficient of the patient's urine data based on the number of days after surgery and the diuretic medication status, and to perform trend segmentation on the urine measurement time series collected by the closed tubing using the segmentation sensitivity coefficient. In this embodiment, in the trend segmentation module, the segmentation sensitivity coefficient The calculation method is as follows:

[0034] in, The baseline sensitivity is set based on clinical experience; in this embodiment, it is set to 1.0. The decay function is based on the patient's disease type and the number of days after surgery; the shorter the number of days after surgery, the more... The closer the value is to 1.0, the more critical the postoperative observation period is, and even small changes in urine output may have significant clinical implications. Therefore, this trend segmentation needs to be more sensitive, and as the number of postoperative days increases... The rate of decline gradually varies depending on the type of surgery: after kidney-related surgery, patients' renal function is in a compensatory phase, and fluctuations in urine output still have significant value in assessment. The rate of decline is lower than in patients who did not undergo kidney surgery. Drug sensitivity factors based on diuretic use status, when a patient uses a diuretic This increases the overall segmented sensitivity coefficient and improves segmented sensitivity, thereby capturing the details of urine volume changes under the influence of drugs, when diuretics are not used. ; This is a nursing risk level mapping factor, derived from the nursing level assessment results in the nursing information system. The higher the nursing risk level, the higher the risk level. The larger; To extract the normal fluctuation range of the patient's historical urine output based on the normalization factor, the normal fluctuation range was extracted from the patient's historical urine output data during the stable period over the past seven days. The larger the historical fluctuation range, the better. The smaller the value, the less likely it is to cause over-segmentation of patients with already large fluctuations in their condition; The values ​​of the above factors are obtained in real time through the electronic medical records, medical order system, and nursing record system in the hospital information system.

[0035] The specific values ​​of the segmented sensitivity are determined through the following mapping logic: when the patient's attribute combination is kidney surgery, 0-2 days post-surgery, and diuretics in use, the values ​​of each factor result in a higher segmented sensitivity coefficient; when the patient's attribute combination is non-kidney surgery, post-surgery time greater than 10 days, and no diuretics used, the segmented sensitivity coefficient is in a lower range; the value range of the segmented sensitivity coefficient is 0.1-1.5.

[0036] Trend segmentation is based on the change in the first derivative between adjacent time windows. and The comparison results were segmented as follows: the sliding window width was set to 30 minutes, the sliding step size was 5 minutes, and the urine flow rate was calculated window by window. The change in flow rate between two adjacent windows was recorded. At the corresponding time, segmentation is performed; when At that time, no segmentation is performed and the sliding window continues along the time series; in the adaptive trend segment set output after segmentation, each trend segment contains the start and end times of the segment, direction label, average slope, instantaneous urine volume at the end of the segment, and coefficient of variation within the segment.

[0037] The boundary detection module is used to reversely traverse the segmented results with the nursing handover time as the anchor point, combine the urine volume data illusion filtering rules caused by nursing operations to determine the starting boundary of the current urine trend state, define the interval from the anchor point to the boundary time as the handover window and generate urine trend handover information items. During the reverse traversal process, the boundary detection module starts from the nursing shift handover time and traverses the adaptive trend segment set segment by segment along the time axis. The specific execution flow of the urine data artifact filtering rule is as follows: First, the set of nursing operation records for the current shift is queried. These records are retrieved from the nursing information system and include the operation type, operation time, and patient identifier. Operation types must include at least turning over, tubing flushing, and patient repositioning. The system then checks if the time difference between the candidate boundary time and any nursing operation record time falls within a preset operation-data response deviation threshold. If the value is within the specified range, the candidate boundary is determined to be an illusory boundary of urine volume data caused by the nursing operation and is excluded, and the reverse traversal continues. The deviation threshold is set according to the operation type. If the turning operation involves changes in body position and causes fluctuations in the urine tubing level, the instantaneous impact window on the urine volume reading is large, and the deviation threshold is set to ±5 minutes. If the data disturbance caused by the tubing flushing operation is short-lived, the deviation threshold is set to ±3 minutes. If the impact of the body position adjustment operation is between the two, the deviation threshold is set to ±4 minutes.

[0038] Next, the patient's medication records are queried. These records are retrieved from the medical order system and include the drug name, administration time, and the lower and upper limits of the onset delay interval. The candidate boundary time falls within the pharmacological onset delay interval of any diuretic drug. If the boundary value is within the specified range, the candidate boundary is determined to be a pharmacologically driven boundary. Instead of using the candidate boundary time as the trend state boundary, the process continues to traverse backwards to the drug administration time. and will Marked as the trend state trigger source point; In this embodiment, the pharmacological onset delay range of the diuretic furosemide is usually 30 to 60 minutes after administration. If the candidate boundary time is within this range, then continue to backtrack to the administration time.

[0039] When the candidate boundary time falls within the preset urine physiological rhythm change window, the physiological rhythm change window is the preset diurnal natural change period of urine. Temporary attenuation is:

[0040] in, The preset rhythm attenuation coefficient, and In this embodiment, a value of 0.7 is used. The urine physiological rhythm change window is a preset period of natural diurnal urine variation, typically from 22:00 to 6:00 the next day. Because the rate of urine production naturally decreases during this period, the sensitivity of urine volume changes to the segment boundaries needs to be correspondingly reduced. The attenuated value is used as the reference value. Reassess whether the candidate boundary is a valid trend state boundary.

[0041] After the reverse traversal finally determines the starting boundary time of the trend state, the interval from that time to the nursing shift handover time is defined as the handover window of the current shift. The adaptive trend segment subset within the handover window, the spurious event records filtered out by the reverse traversal, and the trend state trigger source point marker after pharmacological delay compensation together constitute the urine trend handover information item, which is called by subsequent modules.

[0042] The time-effect topology module is used to construct the time-effect topology network of all patients under the care of the nurse taking over the shift. It derives the clinical failure time of each handover information item by the distance between the urine trend slope and the urine volume warning threshold, and performs global failure time recalculation and cascading conflict prediction driven by the deterioration of urine volume trend when urine volume trend changes or nursing behavior events are triggered. The construction method of the time-sensitive topology network is as follows: the nurse on duty is the central node of the network, and each patient under her care is a patient node. The central node and each patient node are connected by a directed time edge. The weight of the time edge is the estimated time taken for the nurse to move from the current position to the corresponding patient and complete the inspection and intervention. In this embodiment, the weight is pre-calibrated according to the ward layout and the distance between beds, and the value range is 1 to 5 minutes.

[0043] The derivation method for the clinical failure time in the time-sensitivity topology module is as follows: For handover information items describing urine trend direction and the risk of crossing warning thresholds, their clinical expiration time is... for:

[0044] Where d is the absolute distance between the current urine output and the warning threshold, which is the minimum safe urine output preset based on the patient's weight and clinical guidelines. For adult patients, it is usually taken as 0.5 mL / kg / h. For example, for a 70 kg patient, the warning threshold is approximately 35 mL / h; k is the slope of the trend segment element within the transition window. For nursing shift handover time; when hour, To preset the minimum effective slope threshold,

[0045] in, As a preset steady-state default validity period, this embodiment takes... For two hours, when the direction of k is away from the warning threshold, Take an infinite value and mark it as valid indefinitely.

[0046] The triggering condition for a urine volume trend change event is when the absolute change in the current trend slope of any patient caused by new urine volume data exceeds the segmentation sensitivity coefficient calculated for that patient in the trend segmentation module; the nursing behavior event is when the shift nurse completes a round intervention operation and performs a confirmation operation on the terminal; after either of the above two types of events occurs, the time-sensitivity topology module performs a global recalculation of the failure time of all patient nodes.

[0047] Among them, the cascading conflict prediction in the time-sensitive topology module includes: After recalculating the global failure time, patient pairs with time conflicts are detected in ascending order of failure time. The criteria for determining time conflicts are: for two patients with adjacent failure times, if the clinical failure time of the patient ranked later is earlier than the sum of the clinical failure time of the patient ranked earlier and the time edge weights of the two patients, then a time conflict is determined. After recalculating the global failure time, patient pairs with time conflicts are detected in ascending order of failure time. For patient pairs with conflicts, the simulation nurse prioritizes the treatment of the patient ranked earlier, causing the patient ranked later to exceed the time limit. The urine output trend of the patient ranked later deteriorates during the period without intervention. That is, based on its current slope, the new slope after deterioration is predicted. In the deterioration scenario, the absolute value of the new slope is the absolute value of the current slope multiplied by a deterioration acceleration factor greater than one. The deterioration acceleration factor is set to 1.5. At the same time, the next failure time of the patient ranked later is recalculated based on the new slope after deterioration. The system detects whether the next failure time will conflict with the failure times of other patients in the network. When the conflict propagation depth reaches two or more levels, a cascading conflict warning is generated and the complete propagation path containing the patient identifiers and corresponding failure times of each propagation level is recorded.

[0048] In this embodiment, the cascade conflict prediction results are used to generate the next optimal nursing action recommendations. If a cascade conflict is detected, the action recommendations include a description of the cascade propagation path and priority treatment for patients whose cascade can be blocked. If only a single-layer conflict exists, the next action patient is recommended according to the principle of minimizing the total clinical delay cost. If there is no conflict, the next action patient is recommended in ascending order of failure time.

[0049] The dual-state synchronization module is used to monitor the current fluctuation range of urine measurement time series during shift handover. When the fluctuation range falls back to the preset stable threshold range, it triggers the freezing of urine measurement data snapshot and pushes the frozen data synchronously to the terminals of both parties during shift handover. During on-call state, it performs adaptive interruption decision based on the clinical delay cost of urine volume trend warning and the nurse's attention status. In the dual-state synchronization module, the current fluctuation amplitude of the urine measurement time series is obtained by calculating the variance and standard deviation of the urine volume data within a preset sliding window. The sliding window width is set to 15 minutes, the sliding step size is 5 minutes, and the fluctuation amplitude is quantified using exponentially weighted moving variance to reduce the weight decay of historical data.

[0050] in, This is the attenuation coefficient, with a value of 0.9. The current urine flow rate. The current exponentially weighted moving average; The update method is as follows:

[0051] The preset stability threshold range is the statistical confidence interval of the fluctuation range of the patient within a stable historical period. In the patient's data over the past 7 days, continuous periods within the same postoperative day range without nursing operation records and without changes in diuretic medication are selected. The coefficient of variation (CV) of the urine flow rate is < 15% as the selection criterion, and the upper limit of the 95% confidence interval of the fluctuation range of the period that meets the criteria is taken as the stability threshold.

[0052] When the fluctuation amplitude of N consecutive sliding windows falls back to the stable threshold range, a snapshot freeze is triggered; in this embodiment, N is 5, that is, it is triggered when 5 consecutive sliding windows meet the stable condition; the frozen data slice is pushed to the terminals of both the shift handover parties along with a unique version number. The unique version number is the system version prefix plus the freeze trigger time, plus the first 8 digits of the patient identifier hash; after both terminals receive the frozen data slices with the same version number, they each render the data view on their respective interfaces, and the urine measurement data viewed by both parties is completely consistent.

[0053] In the dual-state synchronization module, the handover state completes the closed loop and switches to the on-call state after both parties complete the confirmation and awareness operations on their respective terminals. The operation time limits for both parties are as follows: 5 minutes for the handover party to confirm and 8 minutes for the receiving party to be aware. If either party fails to complete the corresponding operation within the preset time limit, the system will perform a step-by-step upgrade: the first level pushes a closed loop reminder to the terminal of the party that has not completed the operation; the second level pushes a non-closed loop prompt and a list of the party's role and patients who have exceeded the time limit to the nursing team leader's terminal. There is a 3-minute interval between each upgrade level; the third level records the non-closed loop event in the handover exception log and forcibly cancels the handover state and enters the on-call state.

[0054] Clinical delay costs in adaptive interruption decision-making under on-call status Based on the absolute value of the current urine volume trend slope Calculation of the remaining distance d between urine volume and the warning threshold: This formula represents the remaining time required for urine volume to deteriorate from the current level to the warning threshold at the current rate of deterioration;

[0055] The nurse's attention status is inferred based on the current operation status of the nurse's terminal and is divided into three levels. When the terminal is continuously viewing patient details, the interruptibility of attention is set to a lower value; when the terminal is idle or switching views, a higher value is set; and within a preset time window after the terminal has just completed a round confirmation operation, the highest value is set and considered the best push window. Interruption-based push is only executed when the clinical delay cost exceeds the interruption threshold mapped by the nurse's attention status; otherwise, the warning is added to the delayed push queue. The delayed push queue triggers centralized push when there is a break in the nurse's terminal operation or when the remaining effective duration of any warning in the queue is less than the preset emergency time limit.

[0056] The closed-loop calibration module is used to reconcile the urine trend judgment records of each module with the actual urine volume data one by one after the end of the shift, trace the root cause parameters of deviation along the step chain and perform targeted calibration.

[0057] In this embodiment, during reconciliation, all urine trend judgment records output by each module within the shift are extracted, including the start and end times and direction labels of trend segments output by the trend segmentation module, the start and end times of trend states output by the boundary detection module, the clinical failure times and cascade conflict prediction results output by the timeliness topology module, and the interruption push times output by the dual-state synchronization module. The predicted value in each judgment record is compared with the corresponding actual value in the complete urine measurement time series that can be obtained after the shift ends, and the deviation is calculated. The deviation is calculated as the percentage of the absolute difference between the predicted value and the actual value to the actual value.

[0058] The root cause parameters for deviations traced along the step chain in the closed-loop calibration module include the following: For each deviation record, trace back upstream from the module from which the deviation record originates, and check the contribution of the intermediate value output by each upstream module to the deviation in turn. When calculating the contribution, substitute the intermediate value output by the upstream module into the calculation logic of the deviation generating module, and evaluate the proportion of the deviation change caused by the individual change of the intermediate value to the original deviation. The higher the proportion, the greater the contribution. The upstream module parameter with the largest contribution is identified as the principal cause parameter of the deviation. Targeted calibration is performed on the principal cause parameter. The calibration magnitude is positively correlated with the contribution weight of the upstream module parameter in the deviation and the amount of deviation. The calibration direction is the direction that reduces this type of deviation. After calibration, the value of the upstream module parameter does not exceed the preset clinically reasonable value range.

[0059] After calibration, the parameter values ​​are associated with the clinical profile type of the patient that triggered the deviation, including disease type, postoperative day range, diuretic use status, and nursing risk level, and stored in the patient type parameter configuration library. When a patient with the same clinical profile type appears in a subsequent shift, the system will first load the latest calibrated parameter value associated with that type from the parameter configuration library to replace the initial preset value as the system operating parameter.

[0060] Example 2 In the above embodiments, the time-sensitivity topology module provides the assuming nurse with a priority ranking of multiple patients' rounds through clinical failure time calculation and cascading conflict prediction. However, cascading conflict prediction only focuses on the temporal contradictions between patients caused by the nurse's processing order, and does not consider that even if there is no cascading conflict, the clinical failure times of multiple patients may still be densely clustered in the same short period of time on the time axis, forming a nursing decision density peak. That is, the assuming nurse needs to continuously perform multiple clinical judgments and intervention operations within a compressed time window. Excessive instantaneous cognitive load will increase the risk of clinical decision-making errors. Therefore, it is impossible to identify such decision density peaks, nor can it shift some intervention opportunities forward to reduce decision density while maintaining the clinical safety of each patient. In order to solve this problem, in this embodiment, the time-sensitivity topology module is also used to: after the global failure time is recalculated, detect whether there are multiple patients' clinical failure times clustered in the same preset density time window in the time-sensitivity topology network. When the number of clustered patients exceeds the preset density threshold, it is determined to be a decision density peak. For each patient within the decision density peak window, the clinical safety advance measure is calculated individually. The clinical safety advance measure is the ratio of the remaining distance between the patient's current urine output and the warning threshold to the absolute value of the urine trend slope, minus the preset safety margin duration. It represents the maximum allowable time to advance the intervention time from the clinical failure time without causing additional clinical risks to the patient. When the calculated clinical safety advance measure is less than or equal to zero, it indicates that the patient does not meet the conditions for safe advance and must be treated at its original failure time.

[0061] Furthermore, based on the clinical safety advance shift amount for each patient, a sequence of intervention timing advance shift suggestions is generated. Patients with positive clinical safety advance shift amounts are arranged from largest to smallest, and the suggested intervention time for each patient is shifted forward from the original failure time by the corresponding clinical safety advance shift amount. After each patient's advance shift is completed, the remaining number of instantaneous decisions in the aggregation window is recalculated. When the remaining number drops below the preset density threshold, the advance shift operation is stopped, and the suggested intervention time for patients with positive clinical safety advance shift amounts is shifted forward from the original failure time, so that the number of instantaneous decisions in the aggregation window drops below the density threshold.

[0062] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A synchronous nursing management system for urine metering data applicable to closed-loop tubing systems, characterized in that, include: The trend segmentation module is used to calculate the segmentation sensitivity coefficient of the patient's urine data based on the number of days after surgery and the diuretic medication status, and to perform trend segmentation on the urine measurement time series collected by the closed tubing using the segmentation sensitivity coefficient. The boundary detection module is used to reversely traverse the segmented results with the nursing handover time as the anchor point, combine the urine volume data illusion filtering rules caused by nursing operations to determine the starting boundary of the current urine trend state, define the interval from the anchor point to the boundary time as the handover window and generate urine trend handover information items. The time-effect topology module is used to construct the time-effect topology network of all patients under the care of the nurse taking over the shift. It derives the clinical failure time of each handover information item by the distance between the urine trend slope and the urine volume warning threshold, and performs global failure time recalculation and cascading conflict prediction driven by the deterioration of urine volume trend when urine volume trend changes or nursing behavior events are triggered. The dual-state synchronization module is used to monitor the current fluctuation range of urine measurement time series during shift handover. When the fluctuation range falls back to the preset stable threshold range, it triggers the freezing of urine measurement data snapshot and pushes the frozen data synchronously to the terminals of both parties during shift handover. During on-call state, it performs adaptive interruption decision based on the clinical delay cost of urine volume trend warning and the nurse's attention status. The closed-loop calibration module is used to reconcile the urine trend judgment records of each module with the actual urine volume data one by one after the end of the shift, trace the root cause parameters of deviation along the step chain and perform targeted calibration.

2. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, In the trend segmentation module, the segmentation sensitivity coefficient The calculation method is as follows: in, Based on sensitivity, This is a decay function based on the patient's disease type and the number of days post-surgery; the shorter the number of days post-surgery, the more... The larger the kidney, the more likely it is to be post-kidney surgery. The rate of decline is lower than that of non-kidney surgery. As a drug sensitivity factor based on diuretic use status, when using diuretics When not in use , Based on nursing risk level mapping factors, This is a normalization factor based on the normal fluctuation range of the patient's historical urine output; the greater the historical fluctuation range, the better. The smaller; The trend segmentation is based on the change in the first derivative between adjacent time windows. and The comparison results are used to perform the splitting, when At that time, segmentation is performed at the corresponding moment.

3. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, During the reverse traversal process, the boundary detection module further includes the following rules for filtering artifacts in urine volume data: First, query the set of nursing operation records within the current shift. If the time difference between the candidate boundary time and any nursing operation record time falls within the preset operation and data response deviation threshold... If the boundary is within the range, the candidate boundary is determined to be an illusory boundary of urine volume data caused by nursing operation and is excluded, and the reverse traversal continues; Next, query the patient's current medication records. When the candidate boundary time falls within the pharmacological onset delay range of any diuretic drug... If the boundary value is within the specified range, the candidate boundary is determined to be a pharmacologically driven boundary. Instead of using the candidate boundary time as the trend state boundary, the process continues to traverse backwards to the drug administration time. and will Marked as the source point for trend state triggering; When the candidate boundary time falls within the preset window of urine physiological rhythm changes, Temporary attenuation is: in, The preset rhythm attenuation coefficient, and , with attenuation Reassess whether the candidate boundary is a valid trend state boundary.

4. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, The derivation method for the clinical failure time in the time-sensitivity topology module is as follows: For handover information items describing urine trend direction and the risk of crossing warning thresholds, their clinical expiration time is... for: Where d is the absolute distance between the current urine volume and the warning threshold, and k is the slope of the trend segment element within the transition window. For nursing shift handover time; when hour, To preset the minimum effective slope threshold, in, The preset steady-state default validity period applies when the direction of k is away from the warning threshold. Take an infinite value and mark it as valid indefinitely.

5. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, The cascading conflict prediction in the time-sensitive topology module includes: After recalculating the global failure time, patient pairs with temporal conflicts are detected in ascending order of failure time. For patient pairs with conflicts, the simulation nurse prioritizes the treatment of patients ranked higher, causing patients ranked lower to time out. The urine output trend of patients ranked lower deteriorates during the period without intervention. That is, the new slope after deterioration is predicted based on the current slope. At the same time, the next failure time of patients ranked lower is recalculated based on the new slope after deterioration. The system detects whether the next failure time will conflict with the failure times of other patients in the network. When the conflict propagation depth reaches two or more levels, a cascading conflict warning is generated and the complete propagation path containing the patient identifiers and corresponding failure times of each propagation level is recorded.

6. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, In the dual-state synchronization module, the current fluctuation range of the urine measurement time series is obtained by calculating the variance and standard deviation of the urine volume data within a preset sliding window. The preset stable threshold interval is the statistical distribution confidence interval of the fluctuation range of the patient within a stable historical period. When the fluctuation range of N consecutive sliding windows falls back to the stable threshold interval, the snapshot freeze is triggered. In the adaptive interruption decision-making under the on-call status, the clinical delay cost Based on the absolute value of the current urine volume trend slope Calculated based on the remaining distance d between the urine volume value and the warning threshold: The nurse's attention state is inferred based on the nurse's current operating state on the terminal. An interruption push is only executed when the clinical delay cost exceeds the interruption threshold mapped by the nurse's attention state; otherwise, the warning is added to the delayed push queue.

7. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, The root cause parameters of the deviation traced along the step chain in the closed-loop calibration module include the following: For each deviation record, trace back upstream from the module from which the deviation record originated, and check the contribution of the intermediate values ​​output by each upstream module to the deviation in turn. The upstream module parameter with the largest contribution is identified as the principal cause parameter of the deviation. Targeted calibration is performed on the principal cause parameter. The calibration magnitude is positively correlated with the contribution weight of the upstream module parameter in the deviation and the amount of deviation. After calibration, the value of the upstream module parameter does not exceed the preset clinically reasonable value range.

8. The urine metering data synchronous nursing management system applicable to closed-loop pipelines according to claim 1, characterized in that, In the dual-state synchronization module, the handover state is closed after both parties complete the confirmation and awareness operations on their respective terminals, and then switches to the on-call state. If either party fails to complete the corresponding operation within the preset time limit, the system performs a step-by-step upgrade: the first level pushes a closed-loop reminder to the terminal of the party that did not complete the operation; the second level pushes a non-closed-loop prompt and a list of the party's role and patients who exceeded the time limit to the nursing team leader's terminal; and the third level records the non-closed-loop event in the handover exception log and forcibly cancels the handover state and enters the on-call state.

9. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 1, characterized in that, The time-sensitive topology module is also used to: after recalculating the global failure time, detect whether there are multiple patients’ clinical failure times clustered in the same preset density time window in the time-sensitive topology network, and when the number of clustered patients exceeds the preset density threshold, it is determined to be a decision density peak. For each patient within the decision density peak window, the clinical safety advance amount is calculated for each patient. The clinical safety advance amount is the ratio of the remaining distance between the patient's current urine output value and the warning threshold to the absolute value of the urine trend slope, minus the preset safety margin duration. It represents the maximum allowable duration for advancing the intervention time from the clinical failure time without causing additional clinical risks to the patient.

10. The urine metering data synchronous nursing management system suitable for closed-loop pipelines according to claim 9, characterized in that, Based on the clinical safety advance measure of each patient, a sequence of intervention timing advance suggestions is generated. The suggested intervention time for patients with positive clinical safety advance measures is shifted forward from the original failure time, so that the number of instantaneous decisions within the aggregation window is reduced to below the density threshold.

Images