Early-stage automatic identification method and system for acute kidney injury related to cardiac surgery

Abstract

The invention provides an early-stage automatic identification method and system for acute kidney injury related to cardiac surgery, and relates to the technical field of physiological parameter detection. The method comprises the following steps: acquiring average arterial pressure, cardiac blood discharge volume, central venous pressure and blood oxygen saturation, regularly recording and extracting data of adjacent time points, calculating a variation range, screening a mutation region, judging whether a reference range is exceeded or not, analyzing a multi-parameter synchronous variation condition, and marking a hemodynamics abnormal state. According to the method, abnormal changes of hemodynamics and urine metabolism are accurately recognized through real-time monitoring and multi-parameter analysis, the urine flow rate and the metabolite excretion rate are combined, misjudgment of a single index is avoided, the fineness of renal function change analysis is improved, the high-risk time period is accurately positioned through time cross comparison, the diagnosis accuracy is improved, and the diagnosis accuracy is improved. Trend analysis and risk grade division enable the change of the illness state to be more visual, and a real-time early warning system effectively monitors the fluctuation trend, optimizes clinical management and improves the timeliness of medical response.

Description

Technical Field

[0001] The present invention relates to the technical field of physiological parameter detection, and particularly to an early automatic recognition method and system for acute kidney injury related to cardiac surgery. Background Art

[0002] The technical field of physiological parameter detection includes real-time monitoring, data acquisition and analysis of human vital signs for medical diagnosis, condition assessment and health management. This technical field mainly involves core contents such as sensor technology, biological signal processing, data analysis and medical imaging technology. The application scope of physiological parameter detection covers multiple medical fields such as the cardiovascular system, respiratory system, and nervous system. Common detection indicators include electrocardiogram, blood oxygen saturation, blood pressure, body temperature, respiratory rate, etc. In recent years, with the development of medical intelligence, artificial intelligence and big data analysis technologies have been gradually introduced into this field to achieve comprehensive analysis from single parameter measurement to multi-parameter, improving the accuracy and efficiency of early disease recognition.

[0003] Among them, the early automatic recognition method and system for acute kidney injury related to cardiac surgery refer to monitoring the changes in the renal function of patients during or after cardiac surgery, and automatically recognizing possible acute kidney injury situations based on the dynamic data of physiological parameters. This technical theme covers the acquisition, feature extraction, index calculation and automatic recognition of physiological signal data. The main methods include using non-invasive or minimally invasive sensing devices to obtain real-time physiological data of patients, such as hemodynamic parameters, urine volume, serum creatinine concentration, etc.; using specific data processing methods to analyze these physiological parameters to identify key biomarkers that may lead to acute kidney injury; combining rule algorithms or models trained based on historical data to calculate and match the collected data, automatically determining whether there is a risk of acute kidney injury, and providing corresponding classification results.

[0004] The existing technologies mainly rely on single physiological parameter monitoring, lack in-depth analysis of the correlation between multiple parameters, and are prone to one-sidedness in abnormal recognition. The data processing method relies on fixed thresholds and cannot adapt to individual differences and dynamic changes in physiological states, which may lead to some risk signals not being captured in time. The urine metabolism assessment method is relatively single, relying only on a certain metabolite index, which may not accurately reflect the overall renal function state and affect the judgment of early lesions. The correlation analysis between hemodynamic abnormalities and urine metabolism abnormalities is weak, making it difficult to identify the overlapping relationship between the two in the time dimension and reducing the accurate judgment of renal function deterioration. The risk assessment lacks a dynamic grading mechanism, resulting in a lack of clear stratification basis for clinical intervention and affecting the pertinence of treatment strategies. The warning mechanism is lagging, fails to make full use of the change trend of physiological parameters, is difficult to provide effective intervention before the condition worsens, may delay the treatment time, and affect the quality of postoperative recovery of patients. Summary of the Invention

[0005] The object of the present invention is to solve the drawbacks existing in the prior art, and to propose an early automatic recognition method and system for acute kidney injury related to cardiac surgery.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] An early automatic recognition method for acute kidney injury related to cardiac surgery, comprising:

[0008] S1: Obtain the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, regularly record and extract data at adjacent time points, calculate the change amplitude, screen the mutation region, determine whether it exceeds the reference range, analyze the synchronous change of multiple parameters, and then mark the hemodynamic abnormal state;

[0009] S2: Collect postoperative urine samples, detect the urine flow rate and metabolite content, calculate the excretion rate, compare with the reference range, extract the change trend, screen the data with a large decrease in a short time, analyze the relationship between the urine flow rate and metabolite concentration, and mark the urine metabolism abnormal state;

[0010] S3: Screen the hemodynamic abnormal state, calculate the abnormal duration, obtain the urine metabolism abnormal state, analyze the decrease amplitude of the urine flow rate and metabolism rate, screen the continuously abnormal interval, and compare the overlap of hemodynamic mutations, and then mark it as the risk state of renal function deterioration;

[0011] S4: Invoke the risk state of renal function deterioration, calculate the hemodynamic abnormal duration and mutation degree, compare the range of decrease in urine metabolism rate, screen the high-risk time interval, analyze the trends of the two, judge the stability of risk changes, and generate the risk level of renal function deterioration according to parameter fluctuations.

[0012] Optionally, the hemodynamic abnormal state includes the abnormal parameter category, abnormal fluctuation amplitude, abnormal duration, and mutation occurrence period; the urine metabolism abnormal state includes the deviation degree of urine excretion rate, the change trend of metabolite concentration, the change of urine flow rate, and the abnormal maintenance duration; the risk state of renal function deterioration includes the hemodynamic abnormal interval, the urine metabolism abnormal interval, the cross ratio of abnormal states, and the functional deterioration trend characteristics; the risk level of renal function deterioration includes the hemodynamic abnormal level, the urine metabolism fluctuation degree, the risk category division standard, and the deterioration progress judgment basis.

[0013] Optionally, the specific steps of obtaining the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, regularly recording and extracting data at adjacent time points, calculating the change amplitude, screening the mutation region, determining whether it exceeds the reference range, analyzing the synchronous change of multiple parameters, and then marking the hemodynamic abnormal state are as follows:

[0014] S101: Obtain the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, record them at fixed time intervals, extract data at adjacent time points, calculate the variation amplitude index, classify the fluctuation characteristics of each parameter, screen the data points with obvious variations, and obtain the parameter variation amplitude range;

[0015] S102: Based on the parameter variation amplitude range, determine whether the single parameter variation exceeds the preoperative baseline range, screen the time points outside the range, record the parameter category and variation value, calculate the multi-parameter variation amplitude ratio, analyze the fluctuation trend of the ratio, identify the deviation degree of the synchronously varying parameters, and screen the data points conforming to the synchronous characteristics according to the deviation degree to obtain the synchronous variation parameter deviation rate;

[0016] S103: Invoke the synchronous variation parameter deviation rate, screen the time periods meeting the mutation conditions, calculate the variation rate index of each parameter, determine whether the variation rate conforms to the hemodynamic abnormality characteristics, screen the time intervals in the abnormal state, and obtain the hemodynamic abnormal state.

[0017] Optionally, collect postoperative urine samples, detect the urine flow rate and metabolite content, calculate the excretion rate, compare with the baseline range, extract the change trend, screen the data with a large decrease in a short time, and analyze the relationship between the urine flow rate and metabolite concentration. The specific steps for marking the abnormal state of urine metabolism are as follows:

[0018] S201: Obtain the urine samples collected at postoperative intervals, detect the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite content, calculate the excretion rate value of each metabolite per unit time, compare with the preoperative stable range, extract the change trend, and screen the data with an obvious decrease in a short time to obtain the change trend of the urine metabolite excretion rate;

[0019] S202: Based on the change trend of the urine metabolite excretion rate, determine whether the decrease amplitude is continuous, screen the continuously decreasing parameters, calculate the duration of the decrease and the cumulative change amplitude, analyze the corresponding relationship between the urine flow rate and metabolite concentration, calculate the influence ratio of the urine flow rate fluctuation on the excretion rate of each metabolite, and screen the data points conforming to the characteristics according to the ratio fluctuation characteristics to obtain the corresponding ratio of the urine flow rate to metabolite concentration;

[0020] S203: Invoke the corresponding ratio of the urine flow rate to metabolite concentration, screen the time periods meeting the abnormal criteria, analyze the abnormal characteristics of the urine flow rate and metabolite excretion rate, mark the time intervals in the abnormal state, and obtain the abnormal state of urine metabolism.

[0021] Optionally, the specific calculation formula for the excretion rate value of the metabolite per unit time is:

[0022]

[0023] Rd represents the excretion rate value of metabolite d per unit time of the metabolite, C d represents the concentration of metabolite d in urine, V f represents the urine flow rate represents the total sum of urine flow rates at all monitoring time points, M represents the total number of monitoring time points, j represents the index of different monitoring time points, and d represents different metabolite categories

[0024] Optionally, screening the hemodynamic abnormal state, calculating the abnormal duration, obtaining the urine metabolic abnormal state, analyzing the urine flow rate and the decline range of the metabolic rate, screening the continuously abnormal interval, and comparing the overlap of hemodynamic mutations, the specific steps for marking as the risk state of renal function deterioration are as follows:

[0025] S301: Based on the hemodynamic abnormal state, obtain hemodynamic data, screen the mutation time period of the hemodynamic abnormal state, extract the change range of hemodynamic parameters within the corresponding time period, calculate the duration of the mutation time period, and obtain the hemodynamic mutation time period range

[0026] S302: Based on the hemodynamic mutation time period range, call the urine metabolic abnormal state, screen the time interval of continuous abnormal decline, and analyze the average decline trend of the urine metabolic rate within the interval to obtain the decline range of the urine metabolic rate

[0027] S303: According to the hemodynamic mutation time period range and the decline range of the urine metabolic rate, compare and analyze the overlap of the two, calculate the cross - time period length index, and determine whether the cross - time period is long enough. If it meets the standard, mark the time period as the risk state of renal function deterioration

[0028] Optionally, the specific calculation formula for the cross - time period length index is as follows:

[0029]

[0030] where, L c represents the cross - time period length index, Q represents the number of time periods meeting the overlap condition represents the end time of the q - th hemodynamic abnormal mutation time period represents the start time of the q - th hemodynamic abnormal mutation time period represents the end time of the q - th urine metabolic rate decline time period represents the start time of the q - th urine metabolic rate decline time period

[0031] Optionally, the specific steps for calling the risk status of renal function deterioration, calculating the duration and mutation degree of hemodynamic abnormalities, comparing the decline range of urine metabolic rate, screening high-risk time intervals, analyzing the trends of the two, judging the stability of risk changes, and generating the risk level of renal function deterioration according to parameter fluctuations are as follows:

[0032] S401: Obtain the risk status of renal function deterioration, extract the duration of hemodynamic abnormalities, calculate the mutation degree, and analyze the mutation amplitude of hemodynamic abnormalities to obtain the hemodynamic mutation degree;

[0033] S402: According to the hemodynamic mutation degree, call urine metabolic data, extract the decline range of urine metabolic rate, and make a comparison to obtain the risk time interval;

[0034] S403: Based on the hemodynamic mutation degree and the risk time interval, analyze the change trends of hemodynamics and urine metabolic rate, judge the stability of the risk trend, and generate the risk level of renal function deterioration according to the parameter fluctuation situation.

[0035] Optionally, the method further includes S5: Call the risk level of renal function deterioration, analyze the abnormal fluctuations of hemodynamics and the decline of urine metabolism, calculate the abnormal duration index, screen the time points with large abnormal fluctuations, judge whether the hemodynamic abnormalities expand, analyze whether the urine metabolism continues to decline, and if the warning conditions are met, mark the warning status of renal function deterioration;

[0036] The warning status of renal function deterioration includes hemodynamic fluctuation characteristics, urine metabolic rate change pattern, abnormal duration status identification, and high-risk patient determination conditions;

[0037] S501: Obtain the risk level of renal function deterioration, call the abnormal fluctuation data of hemodynamics, calculate the fluctuation amplitude of the time point, and make a judgment based on the fluctuation range, extract the abnormal fluctuation value, and analyze the relationship between the fluctuation amplitude and time distribution to obtain the hemodynamic fluctuation amplitude;

[0038] S502: According to the hemodynamic fluctuation amplitude, call the decline data of urine metabolic rate, calculate the decline speed, evaluate the stability of the change of urine metabolic rate, screen the relatively stable intervals and perform iterative comparative analysis to obtain the stability of urine metabolic rate;

[0039] S503: Based on the hemodynamic fluctuation amplitude and the stability of urine metabolic rate, call the stability data, make a comparison in combination with the time series, and analyze whether the urine metabolic rate continues to decline. If the set warning standard is met, mark it as the warning status of renal function deterioration.

[0040] An early automatic recognition system for cardiac surgery-related acute kidney injury, including:

[0041] The hemodynamic abnormality monitoring module acquires the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, records them at fixed time intervals, extracts the data at adjacent time points, calculates the variation range of each parameter, screens the mutation intervals, determines whether the variation of a single parameter exceeds the preoperative baseline range, analyzes the synchronous variation of multiple parameters, calculates the duration of the mutation of each parameter, determines whether it meets the hemodynamic abnormality determination conditions, and generates the time interval of the hemodynamic abnormality state;

[0042] The urine metabolism abnormality detection module calls the time interval of the hemodynamic abnormality state, acquires the urine samples collected at postoperative intervals, detects the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite contents, calculates the excretion rate per unit time, compares it with the preoperative stable range, extracts the change trend, screens the data with a large decline in a short period of time, determines the continuous decline situation, analyzes the corresponding relationship between the urine flow rate and the metabolite concentration, and obtains the urine metabolism abnormality state interval;

[0043] The renal function deterioration risk analysis module calls the time interval of the hemodynamic abnormality state and the time interval of the urine metabolism abnormality state, screens the overlapping part of the two time intervals, extracts the change range of the hemodynamic parameters, calculates the abnormal duration, analyzes the decline range of the urine metabolism rate, screens the time interval of the abnormal duration, determines the consistency of the abnormal area, and obtains the renal function deterioration risk state interval;

[0044] The renal function deterioration risk grading module calls the renal function deterioration risk state interval, extracts the duration of the hemodynamic abnormality, calculates the mutation degree, compares the decline range of the urine metabolism rate, screens the time interval with a greater risk, analyzes the change trend of the hemodynamics and urine metabolism, and calculates the abnormal fluctuation degree to obtain the renal function deterioration risk grade;

[0045] The renal function deterioration warning module calls the renal function deterioration risk grade, extracts the fluctuation trend of the hemodynamic abnormality, analyzes the stability of the decline of the urine metabolism rate, calculates the abnormal duration index of the patient, screens the time points with large abnormal fluctuations, determines whether the hemodynamic abnormality expands, analyzes whether the urine metabolism rate continues to decline, and obtains the renal function deterioration warning state.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] In the present invention, through real-time monitoring and multi-parameter analysis, abnormal changes in hemodynamics and urine metabolism are accurately identified. By combining urine flow rate and metabolite excretion rate, misjudgment of a single indicator is avoided, the precision of analyzing renal function changes is improved, high-risk periods are accurately located through time cross-comparison, the accuracy of diagnosis is enhanced, trend analysis and risk level classification make the changes in the condition more visual, and the real-time warning system effectively monitors the fluctuation trend, optimizing clinical management and improving the timeliness of medical response. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0049] Figure 1 It is a schematic flowchart of the steps of the present invention;

[0050] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The following will describe the technical solutions in the present invention in conjunction with the drawings.

[0052] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, the use of the word "example" aims to present concepts in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0053] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, the meanings they express are the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, the meanings they express are the same.

[0054] In the embodiments of the present invention, sometimes subscripts such as W1 may be written in a non-subscript form such as W1. When their differences are not emphasized, the meanings they express are the same.

[0055] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail in conjunction with the drawings and specific embodiments.

[0056] Please refer to Figure 1 , an embodiment of the present invention provides an early automatic recognition method for acute kidney injury related to cardiac surgery, including:

[0057] S1: Obtain the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, record them at fixed time intervals, extract data at adjacent time points, calculate the parameter variation range, identify the mutation region, screen the abnormal time period, determine whether the single parameter variation exceeds the preoperative baseline range, analyze the synchronous variation of multiple parameters. If the mutation condition is met, mark it as a hemodynamic abnormal state;

[0058] S2: Obtain the urine samples collected at postoperative intervals, detect the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite contents, calculate the excretion rate per unit time, compare with the preoperative stable range, extract the change trend, screen the data with a large decline in a short time, determine the continuous decline situation, analyze the corresponding relationship between the urine flow rate and the metabolite concentration. If it meets the abnormal standard, mark it as a urine metabolism abnormal state;

[0059] S3: Screen the mutation time period of the hemodynamic abnormal state, extract the change range of hemodynamic parameters, calculate the abnormal duration, obtain the urine metabolism abnormal state, analyze the decline range of the urine metabolism rate, screen the time interval of continuous abnormality, compare the overlap of hemodynamic mutations, and determine the consistency of the abnormal region. If the cross time period is long, mark it as a renal function deterioration risk state;

[0060] S4: Call the renal function deterioration risk state, extract the duration of hemodynamic abnormality, calculate the mutation degree, compare the decline range of the urine metabolism rate, screen the time interval with a greater risk, analyze the change trends of hemodynamics and urine metabolism, determine the stability of the risk trend, and generate a renal function deterioration risk level based on the parameter fluctuation situation;

[0061] S5: Call the renal function deterioration risk level, extract the fluctuation trend of hemodynamic abnormality, analyze the stability of the decline of the urine metabolism rate, calculate the patient's abnormal duration index, screen the time points with large abnormal fluctuations, determine whether the hemodynamic abnormality expands, and analyze whether the urine metabolism rate continues to decline. If the warning condition is met, mark it as a renal function deterioration warning state.

[0062] The abnormal hemodynamic states include abnormal parameter categories, abnormal fluctuation amplitudes, abnormal duration, and mutation occurrence time periods; the abnormal urine metabolism states include the deviation degree of urine excretion rate, the change trend of metabolite concentration, the change of urine flow rate, and the abnormal maintenance duration; the risk state of renal function deterioration includes the abnormal hemodynamic interval, the abnormal urine metabolism interval, the cross-ratio of abnormal states, and the functional deterioration trend characteristics; the risk level of renal function deterioration includes the abnormal hemodynamic level, the degree of urine metabolism fluctuation, the risk category classification standard, and the basis for judging the deterioration progress; the early warning state of renal function deterioration includes the hemodynamic fluctuation characteristics, the urine metabolism rate change pattern, the abnormal continuous state identifier, and the determination conditions for high-risk patients.

[0063] The specific steps of S1 are as follows:

[0064] S101: Obtain the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, record them at fixed time intervals, extract the data at adjacent time points, calculate the change amplitude index, classify the fluctuation characteristics of each parameter, screen the data points with obvious changes, and obtain the parameter change amplitude interval.

[0065] Obtain the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, call the corresponding monitoring devices, collect data at fixed time intervals, store the same parameters obtained at different time points in a data table, construct a continuous time series data, extract the values at adjacent time points, calculate the change amplitude index, and the change amplitude is calculated using the absolute value difference method, that is, for a certain parameter P i , at time point t k and t k+1 The change amplitude between them is defined as Perform the same calculation steps for all parameters to obtain a complete change data sequence. Subsequently, classify the change amplitudes of each parameter, determine the fluctuation characteristics of each parameter at different time points, and the statistical distribution method can be used to judge the fluctuation trend of the parameter. By calculating statistical quantities such as the mean, standard deviation, and coefficient of variation of the data, evaluate the fluctuation range of the parameter, and screen the values exceeding a certain specific variation threshold. The threshold can be set according to the baseline data before the operation. For example, set the normal fluctuation range of the mean arterial pressure as μ MAP ± 2σ MAP , where μ MAP is the preoperative baseline mean, and σ MAP is the standard deviation. Then the change data points outside this range will be screened out. In addition, the percentile method can be used to set the change amplitude interval. For example, set the 90% percentile as the high fluctuation threshold, mark the time points exceeding this threshold as abnormal change points. After performing the above steps for all parameters, obtain the parameter change amplitude interval.

[0066] S102: Based on the parameter change amplitude range, determine whether the change in a single parameter exceeds the preoperative baseline range, screen the time points that exceed the range, record the parameter category and the changed value, calculate the multi-parameter change amplitude ratio, analyze the fluctuation trend of the ratio, identify the deviation degree of the synchronously changing parameters, and screen the data points that meet the synchronous characteristics according to the deviation degree to obtain the deviation rate of the synchronously changing parameters;

[0067] Compare the changes of a single parameter at different time points to determine whether it exceeds the preoperative baseline range. The calculation method of the baseline range can be based on the statistical characteristics of the preoperative continuous monitoring data. For example, use μ P ±1.5σ P as the normal range. If the parameter value k at a certain moment t meets or then it is determined to exceed the range. Record the parameter category and the specific changed value at this time point, and then calculate the multi-parameter change amplitude ratio to measure the relative change degree between different parameters. For example, for two parameters P1 and P2, their change ratio can be calculated as And calculate the fluctuation trend of this ratio at consecutive time points, and use the moving average or difference method to analyze the ratio change. If the ratio shows abnormal fluctuations (such as exceeding a specific fluctuation threshold for n consecutive time points), then it is judged as a synchronous change trend. Further calculate the deviation degree of the synchronously changing parameters, that is, calculate the deviation ratio of each parameter relative to the preoperative baseline value. For example, for a certain parameter P, the deviation rate is defined as And screen the data points whose deviation rate exceeds a specific threshold. For example, if B P > 2, it is regarded as a significant deviation. Finally, obtain the deviation rate data of the synchronously changing parameters.

[0068] S103: Invoke the deviation rate of the synchronously changing parameters, screen the time periods that meet the mutation conditions, calculate the change rate index of each parameter, judge whether the change rate meets the characteristics of hemodynamic abnormalities, screen the time intervals of abnormal states, and obtain the hemodynamic abnormal state;

[0069] Invoke the deviation rate of the synchronously changing parameters, analyze its changes in different time periods, and screen the time periods that meet the mutation conditions. The mutation conditions can be defined as a sharp change in the deviation rate within consecutive multiple time points. For example, within three consecutive time points t k , t k+1 , t k+2 inside, the change of the deviation rate of parameter P satisfies τ and where τ is the mutation threshold, which can be set to 0.5 - 1.0 according to experience. For all parameters within the mutation time period, calculate the change rate index. The change rate can be defined as where Δt is the time interval. If the change rate exceeds a certain critical value, for example, if the judgment threshold for hemodynamic abnormality characteristics is set to V P > α, where the value of α can be determined based on preoperative data statistics. If it is the 95th percentile before surgery, it can be determined as the hemodynamic abnormal state. Finally, the time interval of the abnormal state is screened to obtain the hemodynamic abnormal state.

[0070] The specific steps of S2 are as follows:

[0071] S201: Obtain the urine samples collected at postoperative intervals, detect the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite contents, calculate the excretion rate value per unit time of each metabolite, compare with the preoperative stable range, extract the change trend, and screen the data with a significant decrease in a short time to obtain the change trend of the urine metabolite excretion rate;

[0072] The specific calculation formula for the excretion rate value per unit time of the metabolite is:

[0073]

[0074] R d represents the excretion rate value per unit time of metabolite d, C d represents the concentration of metabolite d in urine, V f represents the urine flow rate, represents the total urine flow rate at all monitoring time points, M represents the total number of monitoring time points, j represents the index of different monitoring time points, and d represents different metabolite categories:

[0075] This formula is used to calculate the normalized excretion rate R d per unit time of each metabolite d, which is achieved by multiplying the concentration C d of each metabolite by the urine flow rate V f , and then dividing by the total urine flow rate at all time points during the monitoring period. This calculation method not only reflects the excretion amount of metabolites but also takes into account the change in urine flow, thus more accurately evaluating the excretion dynamics of metabolites.

[0076] Specific calculation example: Assume that during a monitoring period (for example, within one day), the urine samples are collected four times, and the urine flow rate V f measured each time is 100 mL / h, 120 mL / h, 90 mL / h, and 110 mL / h respectively. At the same time, the concentration C d of a certain metabolite (such as gluconic acid) detected at the four time points is 2 mg / mL, 2.5 mg / mL, 1.8 mg / mL, and 2.2 mg / mL respectively.

[0077] First, calculate the total urine flow rate:

[0078]

[0079] Next, calculate the metabolite excretion amount at each time point:

[0080] Excretion amount at the first time point: 2 mg / mL × 100 mL / h = 200 mg / h;

[0081] Excretion amount at the second time point: 2.5 mg / mL × 120 mL / h = 300 mg / h;

[0082] Excretion amount at the third time point: 1.8 mg / mL × 90 mL / h = 162 mg / h;

[0083] Excretion amount at the fourth time point: 2.2 mg / mL × 110 mL / h = 242 mg / h;

[0084] Calculate the normalized value R of the excretion rate per unit time d :

[0085]

[0086] This result indicates that considering the change in urine flow rate, the average hourly excretion amount of gluconic acid is approximately 2.171 mg. This result reflects that during the monitoring period, the excretion rate of gluconic acid is relatively stable. Considering the fluctuation of urine flow rate, it provides a more accurate data basis for diagnosis or further health assessment.

[0087] S202: Based on the change trend of urine metabolite excretion rate, determine whether the decline amplitude continues, screen the continuously declining parameters, calculate the decline duration and the cumulative change amplitude, analyze the corresponding relationship between urine flow rate and metabolite concentration, calculate the influence ratio of urine flow rate fluctuation on the excretion rate of each metabolite, and screen the data points that meet the characteristics according to the ratio fluctuation characteristics to obtain the corresponding ratio of urine flow rate and metabolite concentration;

[0088] Analyze the change of the excretion rate of each metabolite in different time periods one by one, determine whether the decline amplitude continues, and determine whether the decline trend satisfies ΔS m within consecutive N m > τ m by calculating the change amplitude at consecutive multiple time points. If this condition is met, screen this metabolite and record the corresponding time points, and calculate the decline duration T d and the cumulative change amplitude A d , where T d is determined by the length of the time point sequence that meets the decline condition, and the calculation method of the cumulative change amplitude is Subsequently, analyze the corresponding relationship between urine flow rate and metabolite concentration, and calculate the influence ratio Q of urine flow rate fluctuation on the excretion rate of each metabolitem , define the calculation method as By recording the Q at different time points m Change situation, calculate the ratio fluctuation range, screen the data points that meet the fluctuation characteristics, and set the fluctuation threshold (that is, the ratio change exceeding one standard deviation from the mean), screen out the time points with large fluctuations, and obtain the corresponding ratio of urine flow rate to metabolite concentration.

[0089] S203: Call the corresponding ratio of urine flow rate to metabolite concentration, screen the time periods that meet the abnormal criteria, analyze the abnormal characteristics of urine flow rate and metabolite excretion rate, mark the time intervals of abnormal states, and obtain the urine metabolism abnormal state;

[0090] Call the corresponding ratio of urine flow rate to metabolite concentration, analyze the data at all time points, screen the time periods that meet the abnormal criteria, and define the abnormal criteria as the ratio Q m Within consecutive M m time points exceeds the set threshold λ m , if this condition is met, mark this time period as an abnormal state, further analyze the abnormal characteristics of urine flow rate and metabolite excretion rate, and calculate the change rate of urine flow rate and the change rate of metabolite excretion rate Set the abnormal threshold and If the change rate of any parameter exceeds the corresponding threshold, then determine that this time point is an abnormal state, and finally record the time interval of the abnormal state to obtain the urine metabolism abnormal state.

[0091] The specific steps of S3 are as follows:

[0092] S301: Based on the hemodynamic abnormal state, obtain hemodynamic data, screen the mutation time periods of the hemodynamic abnormal state, extract the change range of hemodynamic parameters within the corresponding time periods, and calculate the duration of the mutation time periods to obtain the hemodynamic mutation time period range;

[0093] Call the hemodynamic data sequence to obtain the time-series data of all relevant parameters, including mean arterial pressure, cardiac output, central venous pressure, etc. Organize the data at different time points, extract the continuous change situation at fixed time intervals, and screen the time periods when abnormal states occur one by one. Define the mutation time period as the interval in which the hemodynamic parameters fluctuate violently in a short time. Specifically, when screening, compare the hemodynamic data at adjacent time points, calculate the change amplitude between adjacent time points, and count the situations where the change amplitude is significantly higher than the preoperative fluctuation range within multiple consecutive time points. If the hemodynamic parameters all exceed the preoperative fluctuation range within a fixed time interval, mark this time period as the mutation time period. Subsequently, extract the range of hemodynamic parameters within this time period, compare the parameter values before and after the mutation, count the maximum change value and the change trends of different parameters, and calculate the duration of this mutation time period to obtain the range of the hemodynamic mutation time period.

[0094] S302: Based on the range of the hemodynamic mutation time period, call the abnormal state of urine metabolism, screen the time intervals of continuous abnormal decline, analyze the average decline trend of the urine metabolism rate within the interval, and obtain the decline amplitude of the urine metabolism rate;

[0095] Call the data of the abnormal state of urine metabolism, screen the abnormal situations of urine metabolism parameters within the mutation time period. During the screening process, compare the changes in the urine metabolism rate before and after the mutation time period, count the metabolism rate data at all time points, judge which time points the metabolism rate decreases, and record the duration of the decline. Screen out the intervals in which the metabolism rate continuously decreases within multiple consecutive time points, calculate the average decline trend within this interval, compare the change amplitude of the metabolism rate at each time point, determine whether the decline trend is stable enough, screen out the time periods that meet the characteristics of continuous decline, and then calculate the decline amplitude of the entire interval. Compare the metabolism rate data before the start of the mutation and after the end of the mutation to determine the maximum change value of the metabolism rate, and evaluate whether the decline amplitude reaches the abnormal standard. Finally, obtain the decline amplitude of the urine metabolism rate.

[0096] S303: According to the range of the hemodynamic mutation time period and the decline amplitude of the urine metabolism rate, compare and analyze the overlapping situation between the two, calculate the cross-time period length index, and judge whether the cross period is long enough. If it meets the standard, mark the time period as the risk state of renal function deterioration;

[0097] The specific calculation formula for the cross-time period length index is:

[0098]

[0099] Among them, L c represents the cross-time period length index, Q represents the number of time periods that meet the overlapping conditions, represents the end time of the q-th hemodynamic abnormality mutation time period, represents the start time of the q-th hemodynamic abnormality mutation time period, represents the end time of the q-th urine metabolic rate decline time period, represents the start time of the q-th urine metabolic rate decline time period:

[0100] This formula calculates the total overlapping length L of the hemodynamic abnormality time period and the urine metabolic rate decline time period c . The parameter Q represents the number of time periods that meet the overlapping condition. The start and end times of each pair of time periods are defined by and respectively, representing the start and end times of hemodynamics and urine metabolic rate.

[0101] Specific calculation example: Set two time periods in the actual situation, where: Hemodynamic abnormality time period: starts at 3:00 and ends at 4:00. Urine metabolic rate decline time period: starts at 3:30 and ends at 5:00.

[0102] Convert the time to the number of hours relative to the start of the day, that is:

[0103] Calculate the overlapping length of each time period:

[0104] Overlap|min(4,5)-max(3,3.5)|=|4 3.5|=0.5 hours;

[0105] Therefore, the length L of the cross time period c is 0.5 hours.

[0106] This result indicates that there is a 0.5-hour time overlap between the hemodynamic abnormality state and the urine metabolic rate decline state. The identification of this time overlap helps to determine whether the two states occur at the same time, thus providing data support for further analysis of the risk of renal function deterioration.

[0107] The specific steps of S4 are as follows:

[0108] S401: Obtain the risk state of renal function deterioration, extract the duration of hemodynamic abnormality, calculate the mutation degree, analyze the mutation amplitude of hemodynamic abnormality, and obtain the hemodynamic mutation degree;

[0109] Call hemodynamic abnormality data, screen the duration of abnormalities, extract key time points in the time series, gradually calculate the degree of mutation, evaluate the deviation of hemodynamic parameters at each time point from the preoperative baseline value, compare the parameter change amplitudes at each time point, extract the maximum change range, compare the parameter change rates within the change range, screen data points with prominent change amplitudes in a short period of time, calculate the mutation amplitude based on the numerical changes at adjacent time points, count the mutation degrees at all time points, screen out time periods exceeding the preset threshold, and calculate the fluctuation amplitude within this time period, and finally obtain the hemodynamic mutation degree.

[0110] S402: According to the hemodynamic mutation degree, call urine metabolism data, extract the decline range of urine metabolism rate, make a comparison, and obtain the risk time interval;

[0111] Call urine metabolism data, screen the decline of urine metabolism rate within the mutation time period, extract the time range of the decline of metabolism rate, analyze the metabolism rate values at each time point, compare the change of metabolism rate at adjacent time points, calculate the average decline rate within the decline range, screen out time points with a decline amplitude exceeding the set baseline, and count whether the decline trend continues, screen out time periods with a stable decline trend, and calculate the duration of this time period, compare it with the hemodynamic mutation time period, screen out the time area where the two overlap, and calculate the length of the risk time interval, and finally obtain the risk time interval.

[0112] S403: Based on the hemodynamic mutation degree and the risk time interval, analyze the change trends of hemodynamics and urine metabolism rate, judge the stability of the risk trend, and generate the risk level of renal function deterioration according to the parameter fluctuation situation;

[0113] Gradually analyze the change trends of hemodynamics and urine metabolism rate, extract the change trend data, compare the trend characteristics before and after the mutation occurs, calculate the fluctuation of hemodynamic parameters, and compare it with the decline trend of urine metabolism rate, count the fluctuation range of parameters, screen whether there is a synchronous change of parameters within the same time period, and compare the correlation between hemodynamic parameters and urine metabolism rate changes, calculate the stability of the risk trend, extract the data at high-risk time points, and screen different levels of risk status according to the stability of parameter fluctuations, compare whether the risk trend continues, and generate the risk level of renal function deterioration according to the fluctuation amplitude and trend characteristics.

[0114] The specific steps of S5 are as follows:

[0115] S501: Obtain the risk level of renal function deterioration, call the abnormal fluctuation data of hemodynamics, calculate the fluctuation amplitude at the time point, and make a judgment based on the fluctuation range, extract the abnormal fluctuation value, analyze the relationship between the fluctuation amplitude and time distribution, and obtain the hemodynamic fluctuation amplitude;

[0116] Call the hemodynamic abnormal fluctuation data, extract the hemodynamic parameters at different time points, compare the numerical changes at adjacent time points, calculate the fluctuation amplitude, gradually screen out the time periods with prominent fluctuations, analyze the change characteristics of the fluctuation range, judge the occurrence rule of the abnormal fluctuation value, extract the fluctuation data beyond the normal physiological range, count the fluctuation values and time point distributions of different parameters, calculate the abnormal fluctuation conditions at different time points, screen out the time periods with large changes in hemodynamic parameters within a short time, and count all the abnormal fluctuation values within this time period, calculate the distribution of the fluctuation amplitude in the entire time series, and analyze whether the fluctuation amplitude is concentrated at specific time points or evenly distributed in the entire time interval, and finally obtain the hemodynamic fluctuation amplitude.

[0117] S502: According to the hemodynamic fluctuation amplitude, call the data of the decline in urine metabolic rate, calculate the decline rate, evaluate the stability of the change in urine metabolic rate, screen out relatively stable intervals and perform iterative comparative analysis to obtain the urine metabolic rate stability;

[0118] Call the data of the decline in urine metabolic rate, screen out the changes in urine metabolic rate corresponding to the hemodynamic fluctuation time periods, calculate the decline rate of urine metabolic rate at different time points, record the change values at each time point, screen out the time periods with a large decline rate, and analyze multiple time points, calculate the stability of the change in urine metabolic rate, screen out the time periods with a relatively stable decline trend, and compare the data at the hemodynamic fluctuation time points, screen out the intervals with a relatively stable decline trend in urine metabolic rate, and perform iterative calculations for the screened intervals, compare the decline trends at different time points, analyze whether the change trend continues, and screen out the time periods that meet the trend stability, and finally obtain the urine metabolic rate stability.

[0119] S503: Based on the hemodynamic fluctuation amplitude and the urine metabolic rate stability, call the stability data, compare it with the time series, and analyze whether the urine metabolic rate continues to decline. If the set warning standard is met, mark it as the renal function deterioration warning state;

[0120] Call the stability data, compare it with the time series one by one, screen out the time points where the hemodynamic fluctuation matches the change trend of the urine metabolic rate, calculate whether the urine metabolic rate continues to decline after the hemodynamic fluctuation, screen out the data where the urine metabolic rate maintains a declining trend within multiple consecutive time points, and analyze whether the decline trend continues. At the same time, calculate whether the decline rate exceeds the set warning threshold, screen out the time points exceeding the set threshold, and count the distribution of these time points in the entire time series, calculate the stability of the change trend of the renal function risk. If the decline trend of the urine metabolic rate continues and meets the set warning standard, mark this time period as the renal function deterioration warning state.

[0121] Please refer to Figure 2 , an early automatic recognition system for acute kidney injury related to cardiac surgery provided by an embodiment of the present invention includes:

[0122] The hemodynamic abnormality monitoring module obtains the mean arterial pressure, cardiac output, central venous pressure, and oxygen saturation, records them at fixed time intervals, extracts the data at adjacent time points, calculates the variation range of each parameter, screens the mutation intervals, determines whether the variation of a single parameter exceeds the preoperative baseline range, analyzes the synchronous variation of multiple parameters, calculates the duration of the mutation of each parameter, determines whether it meets the hemodynamic abnormality determination conditions, and generates the time interval of the hemodynamic abnormality state;

[0123] The urine metabolism abnormality detection module calls the time interval of the hemodynamic abnormality state, obtains the urine samples collected at intervals after the operation, detects the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite contents, calculates the excretion rate per unit time, compares with the preoperative stable range, extracts the change trend, screens the data with a large decline in a short time, determines the continuous decline situation, analyzes the corresponding relationship between the urine flow rate and the metabolite concentration, and obtains the urine metabolism abnormality state interval;

[0124] The renal function deterioration risk analysis module calls the time interval of the hemodynamic abnormality state and the time interval of the urine metabolism abnormality state, screens the overlapping part of the two time intervals, extracts the change range of the hemodynamic parameters, calculates the abnormal duration, analyzes the decline range of the urine metabolism rate, screens the time interval of the abnormal duration, determines the consistency of the abnormal area, and obtains the renal function deterioration risk state interval;

[0125] The renal function deterioration risk grading module calls the renal function deterioration risk state interval, extracts the duration of the hemodynamic abnormality, calculates the mutation degree, compares with the decline range of the urine metabolism rate, screens the time interval with a greater risk, analyzes the change trend of the hemodynamic and urine metabolism, calculates the abnormal fluctuation degree, and obtains the renal function deterioration risk level;

[0126] The renal function deterioration warning module calls the renal function deterioration risk level, extracts the fluctuation trend of the hemodynamic abnormality, analyzes the stability of the decline of the urine metabolism rate, calculates the abnormal duration index of the patient, screens the time points with large abnormal fluctuations, determines whether the hemodynamic abnormality expands, analyzes whether the urine metabolism rate continues to decline, and obtains the renal function deterioration warning state.

[0127] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims described above.

Claims

1. An early automatic recognition method for acute kidney injury related to cardiac surgery, characterized in that, Including: S1: Obtain the mean arterial pressure, cardiac output, central venous pressure, and blood oxygen saturation, regularly record and extract data at adjacent time points, calculate the change amplitude, screen the mutation regions, determine whether it exceeds the baseline range, analyze the synchronous change of multiple parameters, and mark the hemodynamic abnormal state; S2: Collect postoperative urine samples, detect the urine flow rate and metabolite content, calculate the excretion rate, compare with the baseline range, extract the change trend, screen the data with a large decrease in a short time, analyze the relationship between the urine flow rate and metabolite concentration, and mark the urine metabolism abnormal state; S3: Screen the hemodynamic abnormal state, calculate the abnormal duration, obtain the urine metabolism abnormal state, analyze the decrease amplitude of the urine flow rate and metabolism rate, screen the continuously abnormal intervals, compare the overlapping situation of hemodynamic mutations, and mark it as the risk state of renal function deterioration; S4: Call the risk state of renal function deterioration, calculate the hemodynamic abnormal duration and mutation degree, compare the range of the decrease in urine metabolism rate, screen the high-risk time intervals, analyze the trends of the two, judge the stability of risk changes, and generate the risk level of renal function deterioration.

2. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, characterized in that The hemodynamic abnormal state includes the abnormal parameter category, abnormal fluctuation amplitude, abnormal duration, and mutation occurrence time period; the urine metabolism abnormal state includes the deviation degree of urine excretion rate, the change trend of metabolite concentration, the change of urine flow rate, and the abnormal maintenance duration; the risk state of renal function deterioration includes the hemodynamic abnormal interval, the urine metabolism abnormal interval, the cross ratio of abnormal states, and the functional deterioration trend characteristics; the risk level of renal function deterioration includes the hemodynamic abnormal level, the urine metabolism fluctuation degree, the risk category classification standard, and the deterioration progress judgment basis.

3. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, wherein The specific steps for obtaining the mean arterial pressure, cardiac output, central venous pressure, blood oxygen saturation, regularly recording and extracting data at adjacent time points, calculating the change amplitude, screening the mutation regions, determining whether it exceeds the baseline range, analyzing the synchronous change of multiple parameters, and then marking the hemodynamic abnormal state are as follows: S101: Obtain the mean arterial pressure, cardiac output, central venous pressure, and blood oxygen saturation, record at fixed time intervals, extract the data at adjacent time points, calculate the change amplitude index, classify the fluctuation characteristics of each parameter, screen the data points with obvious changes, and obtain the parameter change amplitude interval; S102: Based on the parameter change amplitude interval, determine whether the single parameter change exceeds the preoperative baseline range, screen the time points outside the range, record the parameter category and change value, calculate the multiple parameter change amplitude ratio, analyze the fluctuation trend of the ratio, identify the deviation degree of the synchronous change parameters, and screen the data points that meet the synchronous characteristics according to the deviation degree to obtain the synchronous change parameter deviation rate; S103: Call the synchronous change parameter deviation rate, screen the time periods that meet the mutation conditions, calculate the change rate index of each parameter, determine whether the change rate meets the hemodynamic abnormal characteristics, screen the time intervals of the abnormal state, and obtain the hemodynamic abnormal state.

4. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, wherein Collect urine samples after surgery, detect urine flow rate and metabolite content, calculate the excretion rate, compare with the baseline range, extract the change trend, screen data with a large decrease in a short time, analyze the relationship between urine flow rate and metabolite concentration, and mark the specific steps of abnormal urine metabolism status as follows: S201: Obtain urine samples collected at intervals after surgery, detect urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite content, calculate the excretion rate value of each metabolite per unit time, compare with the preoperative stable range, extract the change trend, screen data with an obvious decrease in a short time, and obtain the change trend of urine metabolite excretion rate; S202: Based on the change trend of the urine metabolite excretion rate, judge whether the decrease amplitude is continuous, screen the continuously decreasing parameters, calculate the duration of the decrease and the cumulative change amplitude, analyze the corresponding relationship between urine flow rate and metabolite concentration, calculate the influence ratio of urine flow rate fluctuation on the excretion rate of each metabolite, and screen data points that meet the characteristic according to the ratio fluctuation characteristic to obtain the corresponding ratio of urine flow rate and metabolite concentration; S203: Call the corresponding ratio of urine flow rate and metabolite concentration, screen the time periods that meet the abnormal criteria, analyze the abnormal characteristics of urine flow rate and metabolite excretion rate, mark the time interval of abnormal status, and obtain the abnormal urine metabolism status.

5. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 4, wherein The specific calculation formula for the excretion rate value of each metabolite per unit time is as follows: R d represents the excretion rate value of metabolite d per unit time of the metabolite, C d represents the concentration of metabolite d in urine, V f represents the urine flow rate, represents the sum of urine flow rates at all monitoring time points, M represents the total number of monitoring time points, j represents the index of different monitoring time points, and d represents different metabolite categories.

6. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, wherein The specific steps for screening the hemodynamic abnormal status, calculating the duration of the abnormality, obtaining the abnormal urine metabolism status, analyzing the decrease amplitude of urine flow rate and metabolic rate, screening the continuously abnormal interval, and comparing the overlapping situation of hemodynamic mutations, and then marking it as the risk status of renal function deterioration are as follows: S301: Based on the hemodynamic abnormal status, obtain hemodynamic data, screen the mutation time periods of the hemodynamic abnormal status, extract the change range of hemodynamic parameters in the corresponding time periods, and calculate the duration of the mutation time period to obtain the range of hemodynamic mutation time periods; S302: Based on the range of the hemodynamic mutation time period, call the abnormal urine metabolism status, screen the time intervals of continuous abnormal decrease, and analyze the average decrease trend of urine metabolism rate in the interval to obtain the decrease amplitude of urine metabolism rate; S303: According to the range of the hemodynamic mutation time period and the decrease amplitude of the urine metabolism rate, compare and analyze the overlapping situation of the two, calculate the cross-time period length index, and judge whether the cross-time period is long enough. If it meets the standard, mark the time period as the risk status of renal function deterioration.

7. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 6, characterized in that, The specific calculation formula for the cross-time period length index is as follows: Among them, L c represents the cross - time - period length index, Q represents the number of time periods meeting the overlap condition, represents the end time of the q - th hemodynamic abnormal mutation time period, represents the start time of the q - th hemodynamic abnormal mutation time period, represents the end time of the q - th urine metabolic rate decline time period, represents the start time of the q - th urine metabolic rate decline time period.

8. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, wherein The specific steps for calling the risk status of renal function deterioration, calculating the duration of hemodynamic abnormality and the degree of mutation, comparing the decrease range of urine metabolism rate, screening the high-risk time intervals, analyzing the trends of the two, judging the stability of risk changes, and generating the risk level of renal function deterioration according to parameter fluctuations are as follows: S401: Obtain the risk status of renal function deterioration, extract the duration of hemodynamic abnormality, calculate the degree of mutation, and analyze the mutation amplitude of hemodynamic abnormality to obtain the degree of hemodynamic mutation; S402: According to the degree of hemodynamic mutation, call the urine metabolism data, extract the decline range of the urine metabolism rate, make a comparison, and obtain the risk time interval; S403: Based on the degree of hemodynamic mutation and the risk time interval, analyze the change trends of hemodynamics and urine metabolism rate, judge the stability of the risk trend, and generate the risk level of renal function deterioration according to the parameter fluctuation situation.

9. The early automatic recognition method for acute kidney injury related to cardiac surgery according to claim 1, wherein The method further includes S5: Call the risk level of renal function deterioration, analyze the abnormal fluctuation of hemodynamics and the decline of urine metabolism, calculate the abnormal persistence index, screen the time points with relatively large abnormal fluctuations, judge whether the hemodynamic abnormality expands, and analyze whether the urine metabolism continues to decline. If the warning conditions are met, mark the warning state of renal function deterioration; The warning state of renal function deterioration includes hemodynamic fluctuation characteristics, urine metabolism rate change pattern, abnormal persistence state identifier, and high-risk patient determination conditions; S501: Obtain the risk level of renal function deterioration, call the abnormal fluctuation data of hemodynamics, calculate the fluctuation amplitude of the time point, and make a judgment based on the fluctuation range, extract the abnormal fluctuation value, analyze the relationship between the fluctuation amplitude and the time distribution, and obtain the hemodynamic fluctuation amplitude; S502: According to the hemodynamic fluctuation amplitude, call the decline data of the urine metabolism rate, calculate the decline speed, evaluate the stability of the change of the urine metabolism rate, screen the relatively stable intervals and perform iterative comparative analysis to obtain the stability of the urine metabolism rate; S503: Based on the hemodynamic fluctuation amplitude and the stability of the urine metabolism rate, call the stability data, make a comparison in combination with the time series, and analyze whether the urine metabolism rate continues to decline. If the set warning standard is met, mark it as the warning state of renal function deterioration.

10. An early automatic recognition system for acute kidney injury related to cardiac surgery, characterized in that, According to the early automatic recognition method for acute kidney injury related to cardiac surgery according to any one of claims 1-9, the system includes: The hemodynamic abnormality monitoring module obtains the mean arterial pressure, cardiac output, central venous pressure, and blood oxygen saturation, records them at fixed time intervals, extracts the data of adjacent time points, calculates the change amplitude of each parameter, screens the mutation interval, judges whether the change of a single parameter exceeds the preoperative baseline range, analyzes the synchronous change situation of multiple parameters, calculates the duration of the mutation of each parameter, judges whether it meets the hemodynamic abnormality determination conditions, and generates the time interval of the hemodynamic abnormality state; The urine metabolism abnormality detection module calls the time interval of the hemodynamic abnormality state, obtains the urine samples collected at postoperative intervals, detects the urine flow rate, gluconic acid, fumaric acid, and pseudouridine metabolite contents, calculates the excretion rate per unit time, compares with the preoperative stable range, extracts the change trend, screens the data with a relatively large decline amplitude in a short time, judges the continuous decline situation, and analyzes the corresponding relationship between the urine flow rate and the metabolite concentration to obtain the urine metabolism abnormality state interval; The renal function deterioration risk analysis module calls the hemodynamic abnormal state time interval and the urine metabolism abnormal state interval, screens the overlapping part of the two time intervals, extracts the change range of hemodynamic parameters, calculates the abnormal duration, analyzes the decline range of urine metabolism rate, screens the time interval of abnormal persistence, judges the consistency of the abnormal area, and obtains the renal function deterioration risk state interval; The renal function deterioration risk grading module calls the renal function deterioration risk state interval, extracts the duration of hemodynamic abnormality, calculates the mutation degree, compares the decline range of urine metabolism rate, screens the time interval with greater risk, analyzes the change trends of hemodynamics and urine metabolism, calculates the abnormal fluctuation degree, and obtains the renal function deterioration risk grade; The renal function deterioration warning module calls the renal function deterioration risk grade, extracts the fluctuation trend of hemodynamic abnormality, analyzes the stability of the decline of urine metabolism rate, calculates the patient's abnormal persistence index, screens the time points with greater abnormal fluctuations, judges whether the hemodynamic abnormality expands, analyzes whether the urine metabolism rate continues to decline, and obtains the renal function deterioration warning state.

Images