A CRRT unplanned discontinuation warning and decision support information system

Abstract

The present invention discloses a CRRT unplanned discontinuation warning and decision support information system, which relates to the field of data mining technology. The system includes a medical record database, a data analysis server, and a medical workstation device. The medical record database is used to record the medical records of all historical CRRT patients. The data analysis server is used to obtain a prediction model for determining whether an unplanned CRRT discontinuation event has occurred based on the medical records of successful and failed CRRT discontinuations and artificial intelligence algorithm calibration and verification modeling. The prediction model is then imported into the prediction model using patient attribute information of the patient to be discontinued, the onboarding plan attribute information of the onboarding plan formulated for the patient to be discontinued, and the nursing staff attribute information of the nursing staff assigned to the patient to be discontinued. The prediction result is obtained and fed back to the workstation device so that the warning level is determined based on the prediction result and output and displayed. In this way, the risk of unplanned CRRT discontinuation can be accurately identified as early as possible, enriching the amount of reference information.

Description

Technical Field

[0001] The present invention belongs to the technical field of data mining, and in particular relates to a CRRT unplanned discontinuation warning and decision support information system. Background Art

[0002] Continuous renal replacement therapy (CRRT) refers to a blood purification technique performed continuously for 24 hours or nearly 24 hours per day. It continuously and slowly exchanges solutes through diffusion and / or convection, maximally simulating the renal glomerular clearance pattern of solutes and water and the reabsorption function of the renal tubules. It removes excess water, large, medium, and small molecule uremic toxins, and metabolic waste from the blood. Simultaneously, it replenishes an equal amount of replacement fluid to maintain cardiovascular stability, thereby replacing damaged kidney function and creating favorable conditions for the recovery of body functions. CRRT has the advantages of maintaining good hemodynamic stability, maintaining stable cerebral perfusion pressure, high solute clearance, and preventing the occurrence of imbalance syndrome. As CRRT technology becomes increasingly mature, its clinical application scope is also expanding, playing a particularly important role in the treatment of critical illnesses such as severe infection, trauma, poisoning, and multiple organ failure.

[0003] Currently, on-machine time is a key indicator for evaluating the efficacy of CRRT therapy. However, unplanned CRRT treatment interruptions, resulting from a variety of factors, have become a common clinical problem. Unplanned CRRT treatment interruptions refer to unplanned CRRT treatment interruptions that fail to achieve treatment goals or duration. The primary indicator of treatment goal is the therapeutic dose, which is the sum of dehydration and replacement fluid volume. Studies have shown that the incidence of unplanned CRRT treatment interruptions in clinical practice is as high as 55% to 66.6%. For critically ill patients receiving CRRT, unplanned CRRT treatment interruptions not only directly impact treatment efficacy, increase the financial and psychological burden on patients and their families, prolong recovery time, and even affect survival, but also indirectly increase the workload of medical staff and waste medical resources. Therefore, how to accurately and early identify the risk of unplanned CRRT treatment interruptions, enrich the available information, assist clinicians in making accurate CRRT treatment decisions, and ultimately reduce the incidence of unplanned CRRT treatment interruptions and safeguard CRRT treatment efficacy, is a topic that requires urgent research by those skilled in the art. Summary of the Invention

[0004] The purpose of the present invention is to provide a CRRT unplanned discontinuation warning and decision support information system to solve the problem that the existing technology is unable to achieve early identification of the risk of unplanned CRRT discontinuation and timely warning to effectively assist clinicians in making accurate CRRT machine decisions.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] The present invention provides a CRRT unplanned discontinuation warning and decision support information system, comprising a medical record database, a data analysis server and a medical workstation device, wherein the data analysis server is respectively connected to the medical record database and the medical workstation device;

[0007] The medical record database is used to record the medical records of all patients who have been treated with CRRT. The medical records include patient attribute information, treatment plan attribute information, nursing staff attribute information, and reasons for treatment. The reasons for treatment are divided into planned treatment and unplanned treatment.

[0008] A data analysis server is used to randomly select an equal number of successful onboarding medical records and failed onboarding medical records from the medical record database, then organize the successful onboarding medical records to obtain negative sample data, and organize the failed onboarding medical records to obtain positive sample data, and finally import all positive sample data and all negative sample data into a machine learning model based on an artificial intelligence algorithm for calibration and validation modeling, thereby obtaining a prediction model for determining whether an unplanned CRRT discontinuation event has occurred, wherein a successful onboarding medical record refers to a medical record in which the reason for discontinuation is a planned discontinuation, and a failed onboarding medical record refers to a medical record in which the reason for discontinuation is an unplanned discontinuation;

[0009] A medical workstation device is used to input the patient attribute information of the patient to be connected to the ventilator, the attribute information of the ventilator plan formulated for the patient to be connected to the ventilator, and the attribute information of the nurse assigned to the patient to be connected to the ventilator, and transmit the input data to the data analysis server;

[0010] The data analysis server is also used to import the input data into the prediction model, output the prediction results and feed them back to the medical workstation equipment;

[0011] The medical workstation equipment is also used to determine the warning level of unplanned CRRT discontinuation based on the prediction results and output it for display, so as to provide reference auxiliary information for medical staff when making decisions on whether to put patients on the CRRT machine.

[0012] Based on the above invention, a new solution for providing early warning of the risk of unplanned CRRT discontinuation based on an artificial intelligence algorithm to support decision-making is provided. The solution includes a medical record database, a data analysis server, and a medical workstation device. The medical record database is used to record the medical records of all historical CRRT patients. The data analysis server is used to obtain a prediction model for determining whether an unplanned CRRT discontinuation event has occurred based on the medical records of successful and failed CRRT sessions and artificial intelligence algorithm calibration and verification modeling. The prediction model is then imported into the prediction model to obtain a prediction result, which is fed back to the medical workstation device so that the medical workstation device can determine the warning level of unplanned CRRT discontinuation based on the prediction result and output it for display. In this way, the risk of unplanned CRRT discontinuation can be accurately identified as early as possible, the amount of reference information can be enriched, and clinicians can be assisted in making accurate CRRT session decisions, thereby effectively reducing the incidence of unplanned CRRT discontinuation, ensuring the effectiveness of CRRT treatment, and facilitating practical application and promotion.

[0013] In a possible design, the patient attribute information includes the label values ​​of the patient's age and gender, the patient's disease-related parameters and / or the patient's coagulation-related parameters, wherein the patient's disease-related parameters include the label value of the body position, diastolic blood pressure, systolic blood pressure and / or APACHE II score. The body position is divided into left lateral position, supine position and right lateral position. The patient's coagulation-related parameters include prothrombin time, activated partial thromboplastin time, fibrinogen and / or platelet count.

[0014] In one possible design, the on-line plan attribute information includes the planned on-line duration, on-line anticoagulation related parameters, on-line blood purification related parameters and / or on-line catheter related parameters, wherein the on-line anticoagulation related parameters include the label values ​​of the number of blood products transfused and / or the type of anticoagulant, the on-line blood purification related parameters include the label values ​​of the treatment mode, the label values ​​of the filter type, the pre-replacement fluid volume, the post-replacement fluid volume, the blood flow rate, the arterial pressure and / or the transmembrane pressure, and the on-line catheter related parameters include the label values ​​of the catheterization site, the catheterization duration and / or the label values ​​of the catheter specifications.

[0015] In one possible design, the nursing staff attribute information includes label values ​​of the nursing staff's years of experience, the nursing staff's gender, and / or the nursing staff's job grade.

[0016] In a possible design, negative sample data is obtained based on the medical records of successful computer use, and positive sample data is obtained based on the medical records of failed computer use, including:

[0017] Randomly sort all selected medical records to obtain a medical record sequence;

[0018] For each medical record file in the medical record file sequence, if the corresponding disembarkation reason is planned disembarkation, the occurrence tag value of the corresponding CRRT unplanned disembarkation event is determined to be zero; otherwise, the occurrence tag value of the corresponding CRRT unplanned disembarkation event is determined to be one;

[0019] For each parameter in the medical record except the reason for CRRT withdrawal, the correlation coefficient between the corresponding parameter and the unplanned CRRT withdrawal event was calculated based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record sequence.

[0020] Arrange all parameters in the medical records except the reasons for leaving the machine in descending order according to the correlation coefficient to obtain a parameter sequence;

[0021] Select the first N parameters from the parameter sequence as key parameters, where N represents a positive integer greater than or equal to 5 and less than or equal to 15;

[0022] Based on the successful on-boarding medical records, negative sample data is obtained by organizing the values ​​of all key parameters in the successful on-boarding medical records as the first model input items, and the occurrence label value of the CRRT unplanned off-boarding event corresponding to the successful on-boarding medical records as the first model output item, and then the first model input item and the first model output item are used together as negative sample data;

[0023] Based on the medical records of failed machine access, positive sample data are obtained by organizing them in the following manner: the parameter values ​​of all key parameters in the medical records of failed machine access are used as the second model input items, and the occurrence label value of the CRRT unplanned discontinuation event corresponding to the medical records of failed machine access is used as the second model output item, and then the second model input item and the second model output item are used together as the positive sample data.

[0024] In a possible design, for each parameter in the medical record except the reason for discharge, the correlation coefficient between the corresponding parameter and the CRRT unplanned discharge event is calculated based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record sequence, including:

[0025] For any parameter in the medical record file except the reason for the medical record file, extract the corresponding parameter value sequence S1 that corresponds one-to-one to the medical record file sequence based on all the selected medical record files, and perform a normal distribution KS check on the corresponding parameter value sequence S1 to calculate a first check statistic p value p1;

[0026] Perform a normal distribution KS check on the occurrence label value sequence S2 that corresponds one-to-one to the medical record file sequence to calculate a second check statistic p value p2;

[0027] If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the preset threshold, the correlation coefficient r between any of the parameters and the CRRT unplanned discontinuation event is calculated according to the following formula: q :

[0028]

[0029] In the formula, i represents a positive integer, x i Indicates the value of the i-th parameter in S1, y i represents the i-th occurrence label value in S2, represents the mean value of S1, represents the mean of S2.

[0030] In one possible design, after selecting the first N parameters from the parameter sequence as key parameters, the input data is fed into the prediction model, including:

[0031] Extract the parameter values ​​of all key parameters from the input data and import the extracted results into the prediction model.

[0032] In one possible design, the data analysis server is further configured to optimize the onboarding plan formulated for the patient to be onboarded based on patient attribute information of the patient to be onboarded in the input data and nursing attribute information of the nursing staff assigned to the patient to be onboarded, using a prediction model in combination with an optimization algorithm, to obtain an optimized search result for the onboarding plan and for minimizing the incidence of unplanned CRRT weaning. When the second incidence of unplanned CRRT weaning corresponding to the optimized search result is lower than the first incidence of unplanned CRRT weaning obtained based on the prediction result, the optimized search result is used as a recommended onboarding plan formulated for the patient to be onboarded. Finally, the recommended onboarding plan and the second incidence of unplanned CRRT weaning are transmitted together to the medical workstation device.

[0033] The medical workstation equipment is also used to output and display recommended machine-on plans, and to determine and output a new CRRT unplanned discontinuation warning level based on the second CRRT unplanned discontinuation rate, so as to provide reference auxiliary information for medical staff when making decisions on whether to perform CRRT on patients on the machine.

[0034] In one possible design, based on the patient attribute information of the patient to be connected to the CRRT in the input data and the nursing attribute information of the nursing staff assigned to the patient to be connected to the CRRT, a prediction model is applied in combination with an optimization algorithm to optimize the CRRT plan formulated for the patient to be connected to the CRRT, thereby obtaining the optimal search result for the CRRT plan and minimizing the incidence of unplanned CRRT disconnections, including the following steps S301 to S309:

[0035] S301. Initialize the optimization algorithm parameters including the maximum number of iterations T, and randomly generate an initial search value array of the parameter set to be optimized, and then execute step S302, wherein the parameter set to be optimized includes all parameters in the on-line plan attribute information of the on-line plan formulated for the patient to be on-line, and the initial search value array y of the parameter set to be optimized 0 It is expressed as follows:

[0036]

[0037] Where d″ represents a positive integer less than or equal to D″, D″ represents the total number of parameters in the parameter set to be optimized, represents the initial search value corresponding to the d″th parameter in the parameter set to be optimized, u c,d″ represents the upper limit of the parameter search space corresponding to the d″th parameter, l c,d″ represents the lower limit of the parameter search space corresponding to the d″th parameter, and rand(0,1) represents a pure decimal random generator function;

[0038] S302. The patient attribute information of the patient to be admitted and the nursing attribute information of the nursing staff assigned to the patient to be admitted, along with the initial search value array of the parameter set to be optimized, are imported into the prediction model. A first confidence level for determining the occurrence of an unplanned CRRT discharge event is output, and the first confidence level is used as the fitness level corresponding to the initial search value array. Then, step S303 is executed.

[0039] S303. The initial search value array of the parameter set to be optimized is used as the current optimal search value array, and the current iteration number t′ is initialized to 0. The countdown value of each parameter in the parameter set to be optimized is also initialized to zero, and then step S304 is executed;

[0040] S304. Based on the current optimal search value array, each parameter in the parameter set to be optimized and whose current prohibited change countdown value is zero is independently incremented and decremented, resulting in 2 × D″′ new search value arrays for the parameter set to be optimized. Then, step S305 is executed, where D″′ represents the total number of parameters in the parameter set to be optimized and whose current prohibited change countdown value is zero, and 1 ≤ D″′ ≤ D″.

[0041] S305. For each array in the 2×D″′ new search value arrays, the patient attribute information of the patient to be admitted and the nursing staff attribute information of the nursing staff assigned to the patient to be admitted, along with the corresponding array, are imported into the prediction model. A second confidence level for determining the occurrence of an unplanned CRRT discharge event is output, and this second confidence level is used as the fitness level of the corresponding array. Then, step S306 is executed.

[0042] S306. For each array, the fitness of the corresponding array is subtracted from the fitness corresponding to the current optimal search value array to obtain a fitness difference value for the corresponding array. If the fitness difference value is greater than zero and is the current maximum fitness difference value, the prohibited change countdown value of the corresponding unique variable parameter is updated to a positive integer value positively correlated with the fitness difference value, and then step S307 is executed. The unique variable parameter is a parameter in the set of parameters to be optimized that is used to obtain the corresponding array through addition or subtraction processing.

[0043] S307. Determine whether there is any parameter in the set of parameters to be optimized whose current prohibited change countdown value is zero. If not, decrement the current prohibited change countdown value of each parameter in the set of parameters to be optimized by 1, and then return to step S307. Otherwise, execute step S308.

[0044] S308. Determine whether any fitness difference value among the fitness difference values ​​of each array is greater than zero. If so, update the current optimal search value array to the array corresponding to the minimum fitness in the 2×D″′ new search value arrays, and then execute step S309. Otherwise, directly execute step S309;

[0045] S309. Increment the current number of iterations t′ by 1, and determine whether the current number of iterations t′ reaches the maximum number of iterations T. If so, use the current optimal search value array as the on-machine plan formulated for the patients to be on-machine and the optimized search result for minimizing the incidence of unplanned CRRT discontinuation. Otherwise, return to step S304.

[0046] In a possible design, the warning level for unplanned CRRT removal is determined based on the prediction results, including:

[0047] The confidence level of the input data and the occurrence of the CRRT unplanned discontinuation event is extracted based on the prediction results, and the confidence level is used as the first CRRT unplanned discontinuation rate;

[0048] According to all preset probability ranges corresponding to all CRRT unplanned disembarkation warning levels, if it is found that the incidence rate of the first CRRT unplanned disembarkation is within the preset probability range corresponding to a certain CRRT unplanned disembarkation warning level, then the certain CRRT unplanned disembarkation warning level will be used as the CRRT unplanned disembarkation warning level determined according to the prediction results.

[0049] Beneficial effects of the above scheme:

[0050] (1) The present invention provides a new solution for early warning of the risk of unplanned CRRT discontinuation based on an artificial intelligence algorithm to achieve decision support, including a medical record database, a data analysis server and a medical workstation device, wherein the medical record database is used to record the medical records of all historical CRRT patients, and the data analysis server is used to obtain a prediction model for judging whether an unplanned CRRT discontinuation event occurs based on the successful and failed medical records and the artificial intelligence algorithm calibration and verification modeling, and the patient attribute information of the patient to be discontinued from the medical workstation device, the on-boarding plan attribute information of the on-boarding plan formulated for the patient to be discontinued, and the nursing staff attribute information of the nursing staff assigned to the patient to be discontinued are imported into the prediction model, and the prediction result is obtained and fed back to the medical workstation device so that the CRRT unplanned discontinuation warning level can be determined according to the prediction result and output and displayed. In this way, the risk of unplanned CRRT discontinuation can be accurately identified as early as possible, the reference information volume can be enriched, and clinicians can be assisted in making accurate CRRT on-boarding decisions, thereby effectively promoting the reduction of the incidence of unplanned CRRT discontinuation, ensuring the treatment effect of CRRT, and facilitating practical application and promotion;

[0051] (2) It is also possible to combine the prediction model and optimization algorithm to optimize the best patient computerization plan and nursing staff allocation plan, further enriching the reference information;

[0052] (3) A new heuristic algorithm step is also provided to optimize the on-machine solution. During the optimization process, the search results in each optimal neighborhood search direction can be temporarily locked for different numbers of iterations based on different fitness difference values, thereby facilitating the rapid search for the optimal on-machine solution attribute information. Furthermore, by first quantitatively increasing / decreasing the parameters to be adjusted and then performing an unquantified random increase / decrease in the parameters adjustment method, it can also avoid falling into the local optimal solution, further facilitating the rapid and accurate acquisition of the optimal search results. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 A schematic diagram of the structure of the CRRT unplanned discontinuation warning and decision support information system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly introduced below in conjunction with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0056] It should be understood that although the terms first, second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are merely used to distinguish one object from another. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0057] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0058] Example

[0059] like Figure 1As shown, the CRRT unplanned discontinuation warning and decision support information system provided in this embodiment includes but is not limited to a medical record database, a data analysis server and a medical workstation device, wherein the data analysis server is communicatively connected to the medical record database and the medical workstation device respectively.

[0060] The medical record database is used to record the medical records of all historical CRRT patients, wherein the medical records include, but are not limited to, patient attribute information, treatment plan attribute information, nursing staff attribute information, and reasons for treatment, wherein the reasons for treatment are divided into planned treatment and unplanned treatment. The aforementioned historical CRRT patients are patients who have historically undergone CRRT treatment, and therefore the corresponding medical records can be routinely recorded and stored in the medical record database. Specifically, the patient attribute information includes but is not limited to the label values ​​of the patient's age and patient's gender (for example, the value "1" represents male, and the value "0" represents female), patient disease-related parameters and / or patient coagulation-related parameters, etc., wherein the patient disease-related parameters include but are not limited to the label value of the body position, diastolic blood pressure, systolic blood pressure and / or APACHE II score (APACHE's full name is Acute Physiology and Chronic Health Evaluation, translated into Chinese as Acute Physiology and Chronic Health Score), etc. The body position is divided into left lateral decubitus position (the corresponding label value can be, for example, the value "-1"), supine position (the corresponding label value can be, for example, the value "0") and right lateral decubitus position (the corresponding label value can be, for example, the value "1"), etc. The patient coagulation-related parameters include but are not limited to prothrombin time, activated partial thromboplastin time, fibrinogen and / or platelet count, etc. The on-line plan attribute information includes but is not limited to the planned on-line duration, on-line anticoagulation related parameters, on-line blood purification related parameters and / or on-line catheter related parameters, wherein the on-line anticoagulation related parameters include but are not limited to the label values ​​of the number of blood products to be transfused and / or the type of anticoagulant, the on-line blood purification related parameters include but are not limited to the label values ​​of the treatment mode, the label value of the filter type, the pre-replacement fluid volume, the post-replacement fluid volume, the blood flow rate, the arterial pressure and / or the transmembrane pressure, etc., and the on-line catheter related parameters include but are not limited to the label values ​​of the catheterization site, the catheterization duration and / or the catheter specification, etc. The nursing staff attribute information includes but is not limited to the label values ​​of the nursing staff's years of work, the nursing staff's gender (for example, the value "1" represents male, and the value "0" represents female) and / or the label value of the nursing staff's job level, etc.The aforementioned types of anticoagulants can be specifically divided into, but not limited to, heparin anticoagulants, low molecular weight heparin anticoagulants, and no anticoagulants; the aforementioned treatment modes can be specifically divided into, but not limited to, continuous venovenous hemofiltration mode (CVVH), continuous venovenous hemodialysis mode (CVVHD), continuous venovenous hemodiafiltration mode (CVVHDF), slow continuous ultrafiltration mode (SCUF), continuous plasma filtration adsorption mode (CPFA), endotoxin adsorption mode, plasma exchange mode (PE) and double membrane plasma exchange mode (DFPP); the aforementioned filter types can be specifically divided into, but not limited to, AV600S filter. , AV600S+HA330 filter, HF1200 filter and HF1200+HA330 filter, etc.; the aforementioned catheter placement site can be specifically but not limited to the right femoral vein site, left femoral vein site and other sites, etc.; the aforementioned catheter specifications can be specifically but not limited to the 11.5F×16cm specification, 11.5F×20cm specification, 7F×20cm specification and three-chamber 12F×16cm specification, etc.; the aforementioned nursing staff job level can be specifically but not limited to the junior, intermediate and senior levels, each level is further divided into different grades, for example, the senior level is further divided into N3 and N4. Each specific term in the medical record file is a common term in the medical field, and the specific meaning is not described in detail here. In addition, the medical record file database can be specifically but not limited to conventionally implemented using a relational database.

[0061] The data analysis server is used to randomly select an equal number of successful medical records and failed medical records from the medical record database, and then obtain negative sample data based on the successful medical records, and obtain positive sample data based on the failed medical records, and finally import all positive sample data and all negative sample data into a machine learning model based on an artificial intelligence algorithm for calibration and verification modeling to obtain a prediction model for determining whether an unplanned CRRT disembarkation event has occurred, wherein the successful medical records refer to the medical records whose disembarkation reason is a planned disembarkation, and the failed medical records refer to the medical records whose disembarkation reason is an unplanned disembarkation. The specific process of randomly selecting successful medical records and failed medical records can be conventionally implemented in combination with the existing pseudo-random algorithm. For example: suppose the total number of successful medical records in the medical record database is N. su , a pseudo-random algorithm can be used to generate 100 different numbers in the interval [1, N su], and then select the medical records of successful computer access with serial numbers of these random positive integers to obtain 100 medical records of successful computer access; similarly, 100 medical records of failed computer access can also be obtained. The purpose of the aforementioned requirement that the number of medical records of successful computer access and medical records of failed computer access is equal (for example, 100 each) is to make the distribution of each result relatively uniform in all subsequent sample data, otherwise the trained model cannot achieve basic prediction functions. Considering that there are many model input parameters available in the medical records, and different parameters must have different effects on unplanned computer access, in order to ensure that a prediction model with high accuracy can be trained in the future, and to achieve the purpose of minimizing the demand for computing resources by reducing dimensionality, preferably, negative sample data is obtained according to the medical records of successful computer access, and positive sample data is obtained according to the medical records of failed computer access, including but not limited to the following steps S101 to S107.

[0062] S101. Randomly sort all selected medical records to obtain a medical record sequence.

[0063] In step S101 , all the medical records selected above include all successful medical records and all failed medical records randomly selected from the medical record database (the two types of cases are equal).

[0064] S102. For each medical record file in the medical record file sequence, if the corresponding disembarkation reason is planned disembarkation, determine that the occurrence tag value of the corresponding CRRT unplanned disembarkation event is zero; otherwise, determine that the occurrence tag value of the corresponding CRRT unplanned disembarkation event is one.

[0065] S103. For each parameter in the medical record file except the reason for the CRRT unplanned CRRT discontinuation, the correlation coefficient between the corresponding parameter and the CRRT unplanned discontinuation event is calculated based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record file sequence.

[0066] In step S103, the various parameters are respectively the label values ​​of the patient's age, the patient's gender, the patient's disease-related parameters, the patient's coagulation-related parameters, the planned duration of the CRRT machine, the anticoagulation-related parameters, the blood purification-related parameters, the catheter-related parameters, the nursing staff's years of experience, the label values ​​of the nursing staff's gender, and / or the label values ​​of the nursing staff's job grade, etc. Specifically, for each parameter in the medical record file except the reason for the CRRT machine removal, based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record file sequence, the correlation coefficient between the corresponding parameter and the CRRT unplanned machine removal event is calculated, including but not limited to the following steps S1031 to S1033.

[0067] S1031. For any parameter in the medical record file except the reason for disconnection, based on all the selected medical record files, extract the corresponding parameter value sequence S1 that corresponds one-to-one to the medical record file sequence, and perform a normal distribution KS check on the corresponding parameter value sequence S1 to calculate the first check statistic p value p1.

[0068] In step S1031, the normal distribution KS check is the Kolmogorov-Smirnov test, which is an existing statistical test method used to determine whether the observed values ​​of a set of data conform to a specific distribution. Therefore, it can be conventionally used to obtain the first check statistic p value p1.

[0069] S1032. Perform the normal distribution KS check on the occurrence label value sequence S2 corresponding one-to-one to the medical record file sequence to calculate a second check statistic p value p2.

[0070] S1033. If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the preset threshold, the correlation coefficient r between any of the parameters and the CRRT unplanned discontinuation event is calculated according to the following formula: q :

[0071]

[0072] In the formula, i represents a positive integer, x i Indicates the value of the i-th parameter in S1, y i represents the i-th occurrence label value in S2, represents the mean value of S1, represents the mean of S2.

[0073] In step S1033 , the preset threshold may be 0.05, for example.

[0074] S104. Arrange all parameters in the medical record file except the reason for leaving the machine in descending order according to the correlation coefficient to obtain a parameter sequence.

[0075] S105. Select the first N parameters from the parameter sequence as key parameters, where N represents a positive integer greater than or equal to 5 and less than or equal to 15.

[0076] In step S105 , N may be set to 10, for example.

[0077] S106. Based on the successful on-boarding medical record file, negative sample data is obtained by organizing in the following manner: the parameter values ​​of all key parameters in the successful on-boarding medical record file are used as the first model input item, and the occurrence label value of the CRRT unplanned off-boarding event corresponding to the successful on-boarding medical record file is used as the first model output item, and then the first model input item and the first model output item are used together as negative sample data.

[0078] S107. Based on the failed medical record file, positive sample data is obtained by sorting in the following manner: the parameter values ​​of all key parameters in the failed medical record file are used as the second model input items, and the occurrence label value of the CRRT unplanned discontinuation event corresponding to the failed medical record file is used as the second model output item, and then the second model input item and the second model output item are used together as the positive sample data.

[0079] The aforementioned artificial intelligence algorithm is a core artificial intelligence algorithm that specifically studies how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their own performance. It is the fundamental way to make computers intelligent. Specifically, the artificial intelligence algorithm can be, but is not limited to, machine learning algorithms based on support vector machines, stochastic gradient descent, multivariate linear regression, multilayer perceptrons, decision trees, back propagation neural networks, or radial basis function networks, so as to quickly and accurately find patterns in the data. Therefore, based on all the positive sample data and all the negative sample data, through a conventional calibration and verification modeling process (specifically including a calibration process and a verification process of the model, that is, first by comparing the model simulation results with the measured data, and then adjusting the model parameters according to the comparison results so that the simulation results are consistent with the actual process), the prediction model for determining whether an unplanned CRRT disembarkation event occurs can be trained. In addition, in the calibration and verification modeling process of the prediction model, it is preferred to use a tree-structured Bayesian optimization algorithm to tune the model parameters; the data analysis server can specifically use an existing server structure.

[0080] The medical workstation device is used to input the patient attribute information of the patient to be connected to the machine, the attribute information of the connection plan formulated for the patient to be connected to the machine, and the attribute information of the nursing staff assigned to the patient to be connected to the machine, and transmit the input data to the data analysis server. The aforementioned connection plan can be manually formulated by the attending physician of the patient to be connected to the machine; the aforementioned nursing staff can be automatically assigned based on existing technology or can also be assigned by the attending physician; the aforementioned input data can be manually entered or automatically entered. In addition, the medical workstation device can be specifically implemented using existing software and hardware structures.

[0081] The data analysis server is also used to import the input data into the prediction model, output the prediction results and feed them back to the medical workstation equipment. The aforementioned prediction results will include the following two situations: (1) judging that an unplanned CRRT discontinuation event will occur based on the input data (in this case, the output value of one is used to indicate this) and giving a confidence level; (2) judging that an unplanned CRRT discontinuation event will not occur based on the input data (in this case, the output value of zero is used to indicate this) and giving a confidence level. In addition, if the first N parameters have been selected from the parameter sequence as key parameters, in order to ensure the consistency of the model input items in the model training stage and the model application stage, preferably, after the first N parameters have been selected from the parameter sequence as key parameters, the input data is imported into the prediction model, including but not limited to: extracting the parameter values ​​of all key parameters from the input data, and importing the extracted results into the prediction model.

[0082] The medical workstation device is also used to determine the CRRT unplanned disembarkation warning level based on the prediction result and output it for display, so as to provide reference auxiliary information for medical staff when making a decision on whether to put the patient on the machine for CRRT. Since the prediction result will give the confidence level of determining the occurrence of a CRRT unplanned disembarkation event or the confidence level of determining that the CRRT unplanned disembarkation event will not occur, the former confidence level or the difference between 1 and the latter confidence level can be used as the estimated CRRT unplanned disembarkation incidence rate for the input data, and then the CRRT unplanned disembarkation warning level can be determined based on the CRRT unplanned disembarkation incidence rate. Specifically, determining the CRRT unplanned disembarkation warning level based on the prediction result includes but is not limited to the following steps S201 to S202.

[0083] S201. Extracting the confidence level of the input data and determining the occurrence of an unplanned CRRT discontinuation event based on the prediction result, and using the confidence level as a first unplanned CRRT discontinuation incidence rate.

[0084] S202. Based on all preset probability ranges corresponding to all CRRT unplanned disembarkation warning levels, if it is found that the first CRRT unplanned disembarkation incidence rate is within the preset probability range corresponding to a certain CRRT unplanned disembarkation warning level, then the certain CRRT unplanned disembarkation warning level is used as the CRRT unplanned disembarkation warning level determined according to the prediction result.

[0085] In step S202, all CRRT unplanned discontinuation warning levels specifically include but are not limited to risk warning (the corresponding preset probability range is, for example, 5% or less), low risk warning level (the corresponding preset probability range is, for example, 5% to 15%), medium risk warning level (the corresponding preset probability range is, for example, 15% to 30%), and high risk warning level (the corresponding preset probability range is, for example, 30% or more). For example, if the first CRRT unplanned discontinuation rate is 45%, the high risk warning level can be determined as the CRRT unplanned discontinuation warning level determined according to the prediction result and output and displayed, so as to provide reference auxiliary information for medical staff when making a decision on whether to put the patient on the machine. In this way, the risk of CRRT unplanned discontinuation can be accurately identified as early as possible, so as to enrich the amount of reference information, assist clinicians in making accurate CRRT machine decisions, and thus effectively promote the reduction of the CRRT unplanned discontinuation rate, ensure the treatment effect of CRRT, and facilitate practical application and promotion.

[0086] Preferably, the data analysis server is further configured to apply the prediction model and, in combination with the optimization algorithm, optimize the onboarding plan formulated for the patient to be onboarded based on the patient attribute information of the patient to be onboarded and the nursing staff attribute information of the nursing staff assigned to the patient to be onboarded in the input data, to obtain the onboarding plan and the optimized search result for minimizing the incidence rate of unplanned CRRT discontinuation; and then, when the second unplanned CRRT discontinuation incidence rate corresponding to the optimized search result is lower than the first unplanned CRRT discontinuation incidence rate obtained according to the prediction result, the optimized search result is used as the recommended onboarding plan formulated for the patient to be onboarded; and finally, the recommended onboarding plan and the second unplanned CRRT discontinuation incidence rate are transmitted together to the medical workstation device; the medical workstation device is further configured to output and display the recommended onboarding plan, and to determine a new CRRT unplanned discontinuation warning level based on the second CRRT unplanned discontinuation incidence rate and output and display it, so as to provide reference auxiliary information for medical staff when making a decision on whether to perform CRRT on the patient to be onboarded. The aforementioned optimization algorithm may be, but is not limited to, a particle swarm optimization algorithm, a Newton optimization algorithm, a genetic optimization algorithm, a Grey Wolf Algorithm, a whale optimization algorithm, or a tuna swarm optimization algorithm. In this way, the best patient on-machine plan can be optimized to further enrich the amount of reference information. In order to quickly and accurately obtain the optimal search result, preferably, based on the patient attribute information of the patient to be on the machine in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be on the machine, the prediction model is applied in combination with the optimization algorithm to optimize the on-machine plan formulated for the patient to be on the machine, and the on-machine plan is obtained and the optimal search result for minimizing the incidence of unplanned CRRT discontinuation is obtained, including the following steps S301 to S309.

[0087] S301. Initialize the optimization algorithm parameters including the maximum number of iterations T, and randomly generate an initial search value array of the parameter set to be optimized, and then execute step S302, wherein the parameter set to be optimized includes all parameters in the on-line program attribute information of the on-line program formulated for the patient to be on-line, and the initial search value array y of the parameter set to be optimized is 0 It is expressed as follows:

[0088]

[0089] Wherein, d″ represents a positive integer less than or equal to D″, D″ represents the total number of parameters in the parameter set to be optimized, represents the initial search value corresponding to the d″th parameter in the parameter set to be optimized, u c,d″represents the upper limit of the parameter search space corresponding to the d″th parameter, l c,d″ represents the lower limit of the parameter search space corresponding to the d″th parameter, and rand(0,1) represents a pure decimal random generation function.

[0090] In step S301 , the parameter search space upper limit and the parameter search space lower limit can be obtained based on conventional statistics of all selected medical records.

[0091] S302. Import the patient attribute information of the patient to be boarded in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be boarded into the prediction model together with the initial search value array of the parameter set to be optimized, output a first confidence level for determining the occurrence of an unplanned CRRT discontinuation event, and use the first confidence level as the fitness level corresponding to the initial search value array, and then execute step S303.

[0092] In step S302, considering that some parameters need to be integers (such as the label value of the treatment mode or the label value of the filter type, etc.), before importing the initial search value array into the model as one of the model input items, it is necessary to round the corresponding parameter values, such as rounding the label value of the treatment mode. Similarly, if the first N parameters have been selected from the parameter sequence as key parameters, in order to ensure the consistency of the model input items in the model training stage and the model application stage, preferably, after the first N parameters are selected from the parameter sequence as key parameters, the patient attribute information of the patient to be on the machine in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be on the machine are imported into the prediction model together with the initial search value array of the parameter set to be optimized, including but not limited to: extracting the parameter values ​​of all key parameters from the first imported data, and importing the extracted results into the prediction model, wherein the first imported data includes the patient attribute information of the patient to be on the machine in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be on the machine and the initial search value array of the parameter set to be optimized. In addition, the first confidence level is the incidence rate of unplanned CRRT discontinuation predicted for the first imported data.

[0093] S303. The initial search value array of the parameter set to be optimized is used as the current optimal search value array, and the current number of iterations t′ is initialized to 0, and the prohibited change countdown value of each parameter in the parameter set to be optimized is also initialized to zero, and then step S304 is executed.

[0094] S304. Based on the current optimal search value array, each parameter in the parameter set to be optimized and whose current prohibited change countdown value is zero is independently increased and decreased to obtain 2×D″′ new search value arrays of the parameter set to be optimized, and then execute step S305, wherein D″′ represents the total number of parameters in the parameter set to be optimized and whose current prohibited change countdown value is zero, and 1≤D″′≤D″.

[0095] In step S304, the increase processing or the decrease processing needs to be performed within the corresponding parameter search space, and can be a quantitative step increase / decrease, or an unquantified random increase / decrease, and it is also possible to first perform a quantitative step increase / decrease on each parameter in the parameter set to be optimized and whose current prohibited change countdown value is zero, and then if it is found that the update of the current optimal search value array is not completed (that is, step S309 is directly executed in the subsequent step S308), then perform an unquantified random increase / decrease on each parameter in the parameter set to be optimized and whose current prohibited change countdown value is zero, so as to avoid falling into a local optimal solution. For example, if the parameter set to be optimized includes the following six parameters: parameter A, parameter B, parameter C, parameter D, parameter E, and parameter F, wherein the current prohibited change countdown values ​​of parameter A, parameter B, parameter E, and parameter F are respectively zero (i.e., the value of D″′ is 4), then, based on the current optimal search value array, parameter A can be independently increased and decreased to obtain two different new search value arrays of the parameter set to be optimized (one of which is obtained based on the increase process and the other is obtained based on the decrease process); based on the current optimal search value array, parameter B can be independently increased and decreased to obtain two different new search value arrays of the parameter set to be optimized; based on the current optimal search value array, parameter E can be independently increased and decreased to obtain two different new search value arrays of the parameter set to be optimized; based on the current optimal search value array, parameter F can be independently increased and decreased to obtain two different new search value arrays of the parameter set to be optimized; thereby, 2×4=8 new search value arrays can be obtained, each array representing a neighborhood search direction in the search space.

[0096] S305. For each array in the 2×D″′ new search value arrays, the patient attribute information of the patient to be boarded in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be boarded in the corresponding array are imported into the prediction model together with the corresponding array, and a second confidence level for determining the occurrence of an unplanned CRRT discontinuation event is output, and the second confidence level is used as the fitness of the corresponding array, and then step S306 is executed.

[0097] In step S305, specific technical details can be obtained by referring to the conventional derivation of step S302, which will not be repeated here.

[0098] S306. For each array, subtract the fitness of the corresponding array from the fitness corresponding to the current optimal search value array to obtain the fitness difference value of the corresponding array, and when the fitness difference value is greater than zero and is the maximum fitness difference value this time, update the prohibited change countdown value of the corresponding unique change parameter to a positive integer value positively correlated with the fitness difference value, and then execute step S307, wherein the unique change parameter refers to the parameter in the parameter set to be optimized and is used to obtain the parameter of the corresponding array by increasing or decreasing the processing.

[0099] In step S306, the fitness difference value is greater than zero and is the maximum fitness difference value this time, which means that the best search value array has been obtained in the neighborhood search direction corresponding to the corresponding parameter this time. Therefore, it is necessary to temporarily lock the search results in this neighborhood search direction by updating the prohibited change countdown value of the corresponding parameter to a positive integer positively correlated with the fitness difference value. In addition, continuing based on the example in step S304 above, among the eight new search value arrays corresponding to parameters A, B, E, and F, if the fitness difference value of a new search value array corresponding to parameter F is greater than zero and is the maximum fitness difference value this time (that is, the largest among the eight fitness difference values ​​this time), then the prohibited change countdown value of parameter F is updated from zero to a positive integer value positively correlated with the fitness difference value, while the prohibited change countdown values ​​of parameters A, B, and E remain zero.

[0100] S307. Determine whether there is any parameter in the parameter set to be optimized whose current prohibited change countdown value is zero. If not, decrement the current prohibited change countdown value of each parameter in the parameter set to be optimized by 1, and then return to step S307, otherwise execute step S308.

[0101] In step S307, if the current change prohibition countdown value of all parameters in the parameter set to be optimized is not zero, it means that it is impossible to return to execute step S304, so it needs to be decremented together until the current change prohibition countdown value of at least one parameter is zero.

[0102] S308. Determine whether there is any fitness difference value greater than zero among the fitness difference values ​​of each array. If so, update the current optimal search value array to an array in the 2×D″′ new search value arrays and corresponding to the minimum fitness, and then execute step S309; ​​otherwise, directly execute step S309.

[0103] S309. Increment the current number of iterations t′ by 1, and determine whether the current number of iterations t′ reaches the maximum number of iterations T. If so, use the current optimal search value array as the on-machine plan formulated for the patient to be on-machine and the optimized search result for minimizing the incidence of unplanned CRRT discontinuation. Otherwise, return to step S304.

[0104] Therefore, based on the aforementioned steps S301 to S309, a new heuristic algorithm step is provided to optimize the on-machine solution. During the optimization process, the search results in each optimal neighborhood search direction can be temporarily locked for different numbers of iterations based on different fitness difference values, thereby facilitating the rapid search for the optimal on-machine solution attribute information. By first quantitatively increasing / decreasing the parameters to be adjusted and then performing an unquantified random increase / decrease in the parameters adjustment method, it is also possible to avoid falling into a local optimal solution, which is further conducive to quickly and accurately obtaining the optimal search results.

[0105] In addition, when there are multiple nursing staff to be assigned, the data analysis server is further used to apply the prediction model and the optimization algorithm to optimize the on-boarding plan formulated for the patient to be on-boarded according to the patient attribute information of the patient to be on-boarded in the input data and the nursing staff attribute information of the corresponding staff for each nursing staff, so as to obtain the on-boarding plan and the corresponding optimized search result for minimizing the incidence rate of unplanned CRRT disembarkation, and then select any optimized search result with the lowest incidence rate of unplanned CRRT disembarkation from each optimized search result corresponding to each nursing staff, and then select the optimized search result when the third CRRT unplanned disembarkation incidence rate corresponding to any optimized search result is lower than the first one obtained according to the prediction result. When the incidence rate of unplanned CRRT weaning is determined, the optimized search result is used as another recommended weaning plan for the patient to be weaned, and finally the recommended weaning plan, the third incidence rate of unplanned CRRT weaning, and the nursing staff attribute information of any nursing staff corresponding to the optimized search result are transmitted to the medical workstation device; the medical workstation device is also used to output and display the recommended weaning plan and the nursing staff attribute information of any nursing staff, and to determine another new CRRT unplanned weaning warning level based on the third incidence rate of unplanned CRRT weaning and output it for display, so as to provide reference auxiliary information for medical staff when making a decision on whether to put the patient to be weaned on the machine. In this way, the best patient weaning plan and nursing staff allocation plan can be optimized to further enrich the amount of reference information.

[0106] In summary, the CRRT unplanned discontinuation warning and decision support information system provided by this embodiment has the following technical effects:

[0107] (1) This embodiment provides a new solution for early warning of the risk of unplanned CRRT discontinuation based on an artificial intelligence algorithm to achieve decision support, including a medical record database, a data analysis server, and a medical workstation device. The medical record database is used to record the medical records of all historical CRRT patients. The data analysis server is used to obtain a prediction model for determining whether an unplanned CRRT discontinuation event occurs based on the successful and failed medical records and the artificial intelligence algorithm calibration and verification modeling. The patient attribute information of the patient to be connected to the machine, the on-machine plan attribute information of the on-machine plan formulated for the patient to be connected to the machine, and the nursing staff attribute information of the nursing staff assigned to the patient to be connected to the machine are imported into the prediction model. The prediction results are obtained and fed back to the medical workstation device so that the medical workstation device can determine the CRRT unplanned discontinuation warning level according to the prediction results and output it for display. In this way, the risk of unplanned CRRT discontinuation can be accurately identified as early as possible, the amount of reference information can be enriched, and clinicians can be assisted in making accurate CRRT on-machine decisions, thereby effectively promoting the reduction of the incidence of unplanned CRRT discontinuation, ensuring the treatment effect of CRRT, and facilitating practical application and promotion.

[0108] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A CRRT unplanned discontinuation warning and decision support information system, characterized by: It includes a medical record database, a data analysis server and a medical workstation device, wherein the data analysis server is respectively connected to the medical record database and the medical workstation device; The medical record database is used to record the medical records of all patients who have been treated with CRRT. The medical records include patient attribute information, treatment plan attribute information, nursing staff attribute information, and reasons for treatment. The reasons for treatment are divided into planned treatment and unplanned treatment. A data analysis server is used to randomly select an equal number of successful onboarding medical records and failed onboarding medical records from the medical record database, then organize the successful onboarding medical records to obtain negative sample data, and organize the failed onboarding medical records to obtain positive sample data, and finally import all positive sample data and all negative sample data into a machine learning model based on an artificial intelligence algorithm for calibration and validation modeling, thereby obtaining a prediction model for determining whether an unplanned CRRT discontinuation event has occurred, wherein a successful onboarding medical record refers to a medical record in which the reason for discontinuation is a planned discontinuation, and a failed onboarding medical record refers to a medical record in which the reason for discontinuation is an unplanned discontinuation; A medical workstation device is used to input the patient attribute information of the patient to be connected to the ventilator, the attribute information of the ventilator plan formulated for the patient to be connected to the ventilator, and the attribute information of the nurse assigned to the patient to be connected to the ventilator, and transmit the input data to the data analysis server; The data analysis server is also used to import the input data into the prediction model, output the prediction results and feed them back to the medical workstation equipment; The medical workstation equipment is also used to determine the warning level of unplanned CRRT discontinuation based on the prediction results and output it for display, so as to provide reference auxiliary information for medical staff when making decisions on whether to put patients on the CRRT machine.

2. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: The patient attribute information includes the label values ​​of the patient's age and gender, the patient's disease-related parameters and / or the patient's coagulation-related parameters. Among them, the patient's disease-related parameters include the label value of the body position, diastolic blood pressure, systolic blood pressure and / or APACHE II score. The body position is divided into left lateral position, supine position and right lateral position. The patient's coagulation-related parameters include prothrombin time, activated partial thromboplastin time, fibrinogen and / or platelet count.

3. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: The on-line plan attribute information includes the planned on-line duration, on-line anticoagulation related parameters, on-line blood purification related parameters and / or on-line catheter related parameters, among which the on-line anticoagulation related parameters include the label values ​​of the number of blood products transfused and / or the type of anticoagulant, the on-line blood purification related parameters include the label values ​​of the treatment mode, the label value of the filter type, the pre-replacement fluid volume, the post-replacement fluid volume, the blood flow rate, the arterial pressure and / or the transmembrane pressure, and the on-line catheter related parameters include the label values ​​of the catheterization site, the catheterization duration and / or the label values ​​of the catheter specifications.

4. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: The nursing staff attribute information includes the label value of the nursing staff's working experience, the nursing staff's gender and / or the label value of the nursing staff's job grade.

5. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: Negative sample data is obtained based on the medical records of successful computer use, and positive sample data is obtained based on the medical records of failed computer use, including: Randomly sort all selected medical records to obtain a medical record sequence; For each medical record file in the medical record file sequence, if the corresponding disembarkation reason is planned disembarkation, the occurrence tag value of the corresponding CRRT unplanned disembarkation event is determined to be zero; otherwise, the occurrence tag value of the corresponding CRRT unplanned disembarkation event is determined to be one; For each parameter in the medical record except the reason for CRRT withdrawal, the correlation coefficient between the corresponding parameter and the unplanned CRRT withdrawal event was calculated based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record sequence. Arrange all parameters in the medical records except the reasons for leaving the machine in descending order according to the correlation coefficient to obtain a parameter sequence; Select the first N parameters from the parameter sequence as key parameters, where N represents a positive integer greater than or equal to 5 and less than or equal to 15; Based on the successful on-boarding medical records, negative sample data is obtained by organizing the values ​​of all key parameters in the successful on-boarding medical records as the first model input items, and the occurrence label value of the CRRT unplanned off-boarding event corresponding to the successful on-boarding medical records as the first model output item, and then the first model input item and the first model output item are used together as negative sample data; Based on the medical records of failed machine access, positive sample data are obtained by organizing them in the following manner: the parameter values ​​of all key parameters in the medical records of failed machine access are used as the second model input items, and the occurrence label value of the CRRT unplanned discontinuation event corresponding to the medical records of failed machine access is used as the second model output item, and then the second model input item and the second model output item are used together as the positive sample data.

6. The CRRT unplanned discontinuation warning and decision support information system according to claim 5, characterized in that: For each parameter in the medical record except the reason for CRRT withdrawal, the correlation coefficient between the corresponding parameter and the unplanned CRRT withdrawal event was calculated based on the occurrence label value sequence and the corresponding parameter value sequence corresponding to the medical record sequence, including: For any parameter in the medical record file except the reason for the medical record file, extract the corresponding parameter value sequence S1 that corresponds one-to-one to the medical record file sequence based on all the selected medical record files, and perform a normal distribution KS check on the corresponding parameter value sequence S1 to calculate a first check statistic p value p1; Perform a normal distribution KS check on the occurrence label value sequence S2 that corresponds one-to-one to the medical record file sequence to calculate a second check statistic p value p2; If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the preset threshold, the correlation coefficient r between any of the parameters and the CRRT unplanned discontinuation event is calculated according to the following formula: q : In the formula, i represents a positive integer, x i Indicates the value of the i-th parameter in S1, y i represents the i-th occurrence label value in S2, represents the mean value of S1, represents the mean of S2.

7. The CRRT unplanned discontinuation warning and decision support information system according to claim 5, characterized in that: After selecting the first N parameters from the parameter sequence as key parameters, the input data is imported into the prediction model, including: Extract the parameter values ​​of all key parameters from the input data and import the extracted results into the prediction model.

8. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: The data analysis server is further configured to optimize the onboarding plan formulated for the patient to be onboarded based on patient attribute information of the patient to be onboarded in the input data and nursing attribute information of the nursing staff assigned to the patient to be onboarded, using a prediction model in combination with an optimization algorithm, to obtain an optimized search result for the onboarding plan and for minimizing the incidence of unplanned CRRT offboarding; and when a second unplanned CRRT offboarding rate corresponding to the optimized search result is lower than a first unplanned CRRT offboarding rate obtained based on the prediction result, using the optimized search result as a recommended onboarding plan formulated for the patient to be onboarded; and finally transmitting the recommended onboarding plan and the second unplanned CRRT offboarding rate together to a medical workstation device; The medical workstation equipment is also used to output and display recommended machine-on plans, and to determine and output a new CRRT unplanned discontinuation warning level based on the second CRRT unplanned discontinuation rate, so as to provide reference auxiliary information for medical staff when making decisions on whether to perform CRRT on patients on the machine.

9. The CRRT unplanned discontinuation warning and decision support information system according to claim 8, characterized in that: Based on the patient attribute information of the patient to be connected to the CRRT in the input data and the nursing staff attribute information of the nursing staff assigned to the patient to be connected to the CRRT, a prediction model is applied in combination with an optimization algorithm to optimize the CRRT plan formulated for the patient to be connected to the CRRT, thereby obtaining the optimal search result for the CRRT plan and minimizing the incidence of unplanned CRRT disconnection, including the following steps S301 to S309: S301. Initialize the optimization algorithm parameters including the maximum number of iterations T, and randomly generate an initial search value array of the parameter set to be optimized, and then execute step S302, wherein the parameter set to be optimized includes all parameters in the on-line plan attribute information of the on-line plan formulated for the patient to be on-line, and the initial search value array y of the parameter set to be optimized 0 It is expressed as follows: Where d″ represents a positive integer less than or equal to D″, D″ represents the total number of parameters in the parameter set to be optimized, represents the initial search value corresponding to the d″th parameter in the parameter set to be optimized, u c,d″ represents the upper limit of the parameter search space corresponding to the d″th parameter, l c,d″ represents the lower limit of the parameter search space corresponding to the d″th parameter, and rand(0,1) represents a pure decimal random generator function; S302. The patient attribute information of the patient to be admitted and the nursing attribute information of the nursing staff assigned to the patient to be admitted, along with the initial search value array of the parameter set to be optimized, are imported into the prediction model. A first confidence level for determining the occurrence of an unplanned CRRT discharge event is output, and the first confidence level is used as the fitness level corresponding to the initial search value array. Then, step S303 is executed. S303. The initial search value array of the parameter set to be optimized is used as the current optimal search value array, and the current iteration number t′ is initialized to 0. The countdown value of each parameter in the parameter set to be optimized is also initialized to zero, and then step S304 is executed; S304. Based on the current optimal search value array, each parameter in the parameter set to be optimized and whose current prohibited change countdown value is zero is independently incremented and decremented, resulting in 2 × D″′ new search value arrays for the parameter set to be optimized. Then, step S305 is executed, where D″′ represents the total number of parameters in the parameter set to be optimized and whose current prohibited change countdown value is zero, and 1 ≤ D″′ ≤ D″. S305. For each array in the 2×D″′ new search value arrays, the patient attribute information of the patient to be admitted and the nursing staff attribute information of the nursing staff assigned to the patient to be admitted, along with the corresponding array, are imported into the prediction model. A second confidence level for determining the occurrence of an unplanned CRRT discharge event is output, and this second confidence level is used as the fitness level of the corresponding array. Then, step S306 is executed. S306. For each array, the fitness of the corresponding array is subtracted from the fitness corresponding to the current optimal search value array to obtain a fitness difference value for the corresponding array. If the fitness difference value is greater than zero and is the current maximum fitness difference value, the prohibited change countdown value of the corresponding unique variable parameter is updated to a positive integer value positively correlated with the fitness difference value, and then step S307 is executed. The unique variable parameter is a parameter in the set of parameters to be optimized that is used to obtain the corresponding array through addition or subtraction processing. S307. Determine whether there is any parameter in the set of parameters to be optimized whose current prohibited change countdown value is zero. If not, decrement the current prohibited change countdown value of each parameter in the set of parameters to be optimized by 1, and then return to step S307. Otherwise, execute step S308. S308. Determine whether any fitness difference value among the fitness difference values ​​of each array is greater than zero. If so, update the current optimal search value array to the array corresponding to the minimum fitness in the 2×D″′ new search value arrays, and then execute step S309. Otherwise, directly execute step S309; S309. Increment the current number of iterations t′ by 1, and determine whether the current number of iterations t′ reaches the maximum number of iterations T. If so, use the current optimal search value array as the on-machine plan formulated for the patients to be on-machine and the optimized search result for minimizing the incidence of unplanned CRRT discontinuation. Otherwise, return to step S304.

10. The CRRT unplanned discontinuation warning and decision support information system according to claim 1, characterized in that: Determine the warning level for unplanned CRRT removal based on the prediction results, including: The confidence level of the input data and the occurrence of the CRRT unplanned discontinuation event is extracted based on the prediction results, and the confidence level is used as the first CRRT unplanned discontinuation rate; According to all preset probability ranges corresponding to all CRRT unplanned disembarkation warning levels, if it is found that the incidence rate of the first CRRT unplanned disembarkation is within the preset probability range corresponding to a certain CRRT unplanned disembarkation warning level, then the certain CRRT unplanned disembarkation warning level will be used as the CRRT unplanned disembarkation warning level determined according to the prediction results.

Images