Intelligent management system for water inflow and outflow of nephrotic patient

Abstract

The invention discloses an intelligent management system for water inflow and outflow of a nephrotic patient. The intelligent management system comprises a data collecting and processing module, a water inflow and outflow prediction module, a water inflow and outflow balance optimization module, a health integrated management module and an intelligent water inflow and outflow management module. The invention belongs to the technical field of intelligent management of nephropathy, and particularly relates to an intelligent management system for water inflow and outflow of a nephropathy patient, which adopts an artificial intelligence technology to realize intelligent management of the water inflow and outflow of the nephropathy patient, and combines long-term and short-term water inflow and outflow prediction with balance optimization of the water inflow and outflow. Automatic and intelligent water inflow and outflow management is realized; performing short-term prediction by combining time sequence convolution with a bidirectional long-short-term memory network, and performing long-term prediction by combining empirical mode decomposition with a bidirectional gating cycle unit; a multi-objective Bayesian optimization evolutionary algorithm model is adopted, and dynamic adaptation and personalized water inflow and outflow balance improvement are achieved through multi-objective optimization.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent management of kidney disease, and specifically relates to an intelligent management system for water intake and output of kidney disease patients. Background Art

[0002] The intelligent management system for water intake and output of kidney patients is an innovative tool based on artificial intelligence technology, which is used to monitor and manage the body water balance of kidney patients. The system collects data such as the patient's water intake, urine output, and weight changes, and combines deep learning models and big data analysis to accurately predict and manage the patient's water intake and output. Its role is to help patients maintain an appropriate water balance in the body, prevent complications such as edema and dehydration, optimize the effect of dialysis treatment, and improve the patient's quality of life.

[0003] However, in the existing intelligent management system for water intake and output of kidney patients, the water intake and output of kidney patients is mostly managed comprehensively by manual observation, recording and feedback in actual clinical operations, which will inevitably lead to problems of untimely and inefficient communication between doctors and patients, and discharged kidney patients are unable to properly balance their own water intake and output in long-term health management, and it is difficult to obtain professional and reliable guidance, thus leading to technical problems of insufficient intelligence and automation in water intake and output management of kidney patients; in the existing water intake and output prediction methods for kidney patients, there is a technical problem that kidney patients need reliable, timely and rapid short-term water intake and output prediction, as well as long-term water intake and output trend prediction, and the existing technology is difficult to take both needs into account; in the existing water intake and output balance optimization methods, there is a technical problem that personalized balance optimization combined with patient personal information is a complex multi-objective optimization problem, which needs to consider both water intake and output, and also needs to be combined with the patient's personal situation and kidney disease situation, and therefore a reasonable multi-objective optimization model must be designed. Summary of the invention

[0004] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent management system for water intake and output of patients with kidney disease. In the existing intelligent management system for water intake and output of patients with kidney disease, the water intake and output of patients with kidney disease is mostly managed comprehensively by manual observation, recording and feedback in actual clinical operations, which will inevitably lead to the problem of untimely and inefficient communication between doctors and patients, and discharged kidney patients are unable to properly balance their own water intake and output in long-term health management, and it is difficult to obtain professional and reliable guidance, thus leading to the technical problem of insufficient intelligence and automation in the management of water intake and output of patients with kidney disease. This solution creatively adopts artificial intelligence technology to realize intelligent management of water intake and output of patients with kidney disease, and realizes automated and intelligent water intake and output management through long-term and short-term water intake and output prediction combined with water intake and output balance optimization, providing strong support for medical assistance for kidney disease; in the existing method for predicting water intake and output of patients with kidney disease, the kidney patients need reliable, timely and rapid short-term water intake and output prediction, and also need long-term water intake and output trend prediction, but the prior art is difficult to take both into account. In order to solve the technical problem of the needs of patients, this solution creatively adopts time series convolution combined with bidirectional long short-term memory network for short-term prediction, and adopts empirical mode decomposition combined with bidirectional gated recurrent unit for long-term prediction. By optimizing the time series relationship that changes sharply in the short term through time series convolution, and combining the bidirectional long short-term network optimization processing, the real-time and response speed of short-term prediction are improved. At the same time, empirical mode decomposition is used to process complex long-term trend data, and the stability of the prediction process is optimized in the bidirectional gated recurrent unit, which comprehensively improves the feasibility and availability of long-term prediction. In view of the technical problem that in the existing water input and output balance optimization methods, there is a personalized balance optimization combined with patient personal information, which is a complex multi-objective optimization problem. It is necessary to consider both water input and output and the patient's personal situation and kidney disease situation, so a reasonable multi-objective optimization model must be designed. This solution creatively adopts the evolutionary algorithm model of multi-objective Bayesian optimization. Through multi-objective optimization, dynamic adaptation and personalized water input and output balance improvement are achieved, while improving water input and output balance, patient comfort and kidney disease burden optimization, and improving the overall reliability of the system.

[0005] The technical solution adopted by the present invention is as follows: the present invention provides an intelligent water intake and output management system for patients with kidney disease, including a data collection and processing module, a water intake and output prediction module, a water intake and output balance optimization module, a health integrated management module and a water intake and output intelligent management module;

[0006] The data collection and processing module is used to collect and process the water inflow and outflow data of the kidney patients, obtain the optimized water inflow and outflow management data through data collection and processing, and send the optimized water inflow and outflow management data to the water inflow and outflow prediction module and the water inflow and outflow balance optimization module;

[0007] The water intake and output prediction module is used to predict the water intake and output of kidney patients, obtain water intake and output prediction data through water intake and output prediction, and send the water intake and output prediction to the water intake and output balance optimization module and the health integrated management module;

[0008] The water inflow and outflow balance optimization module is used to predict the water inflow and outflow balance optimization plan in combination with the disease information of the kidney patients, obtain the water inflow and outflow balance optimization plan through water inflow and outflow balance optimization, and send the water inflow and outflow balance optimization plan to the health integrated management module;

[0009] The health integrated management module is used to integrate and combine the water inflow and outflow balance optimization plan for kidney patients with the health management module, obtain a comprehensive water inflow and outflow management plan for kidney patients through health integrated management, and send the comprehensive water inflow and outflow management plan for kidney patients to the water inflow and outflow intelligent management module;

[0010] The water intake and output intelligent management module is used to perform comprehensive management of water intake and output in combination with health integrated management, and to perform intelligent management of water intake and output for kidney disease patients through intelligent management of water intake and output.

[0011] Furthermore, the data collection and processing module is used to collect water intake and output data and basic condition data of kidney disease patients, and optimize and pre-process the collected data, specifically, to obtain water intake and output prediction raw data by collecting from the medical system, medical records and patient self-recorded data, and to obtain optimized water intake and output management data by optimizing and pre-processing the water intake and output prediction raw data;

[0012] The water intake and output prediction raw data specifically includes sensor data, medical history data and self-water intake and output record data;

[0013] The optimization and preprocessing specifically include data cleaning, data labeling, data standardization and feature extraction.

[0014] Furthermore, the water intake and output prediction module is used to predict the long-term and short-term water intake and output of the kidney disease patient, specifically, based on the optimized water intake and output management data, long-term water intake and output prediction and short-term water intake and output prediction are performed to obtain water intake and output prediction data;

[0015] The water inflow and outflow forecast data specifically include long-term water inflow and outflow trend forecast data and short-term water inflow and outflow change forecast data;

[0016] The long-term water intake and output prediction is used to model and predict the long-term trend change law of water intake and output of kidney disease patients, specifically using a bidirectional gated recurrent unit combined with empirical mode decomposition to perform long-term water intake and output prediction to obtain long-term water intake and output trend prediction data;

[0017] The short-term water intake and output prediction is used to provide real-time and rapid short-term water drinking advice and urination prediction for kidney patients. Specifically, a bidirectional long short-term memory network combined with time series convolution is used to perform short-term water intake and output prediction to obtain short-term water intake and output change prediction data.

[0018] The steps of performing long-term water inflow and outflow prediction and short-term water inflow and outflow prediction to obtain water inflow and outflow prediction data include: full set empirical mode decomposition, adaptive noise addition, construction of bidirectional gated recurrent units, construction of temporal convolutional subnets, construction of bidirectional long short-term memory subnets, construction of attention residual layers and water inflow and outflow prediction;

[0019] The full set empirical mode decomposition is specifically to perform a full set empirical mode decomposition on the time series data in the optimized water inlet and outlet management data to obtain intrinsic mode function data;

[0020] The adaptive noise addition is specifically to apply adaptive noise in the full set empirical mode decomposition to assist time series decomposition;

[0021] The constructing of the bidirectional gated recurrent unit specifically involves constructing a bidirectional gated recurrent unit to perform bidirectional structural time series prediction on each subsequence in the intrinsic mode function data to perform long-term water inflow and outflow trend prediction;

[0022] The constructing of the time series convolution subnet specifically involves constructing a time series convolution network, extracting short-term dependency features of data in the optimized water inflow and outflow management data through a causal convolution layer, and obtaining short-term feature data;

[0023] The constructing of the bidirectional long short-term memory subnet is specifically to construct a standard bidirectional long short-term memory neural network, perform bidirectional structural time series prediction on the short-term feature data, and perform short-term water inflow and outflow change prediction;

[0024] The construction of the attention residual layer specifically introduces the attention mechanism and constructs a residual network to optimize the gradient disappearance problem;

[0025] The water inflow and outflow prediction is specifically to train a long-term water inflow and outflow prediction model through the full set empirical mode decomposition, the adaptive noise addition and the bidirectional gated recurrent unit to obtain a long-term water inflow and outflow prediction model Model L , and train the short-term water inflow and outflow prediction model through the temporal convolution subnet, the bidirectional long short-term memory subnet and the attention residual layer to obtain the short-term water inflow and outflow prediction model Model S , and by using the long-term water inflow and outflow prediction model L And the short-term water inflow and outflow prediction model S, predict water inflow and outflow, and obtain water inflow and outflow prediction data.

[0026] Furthermore, the water inflow and outflow balance optimization module is used to further predict the optimization of drinking water planning in combination with the water inflow and outflow prediction results of kidney patients, specifically, based on the water inflow and outflow prediction data, an evolutionary algorithm model combined with multi-objective Bayesian optimization is used to optimize the water inflow and outflow balance, and obtain a water inflow and outflow balance optimization plan, which specifically includes the following steps: balance optimization objective function setting, construction of a multi-objective Bayesian optimization model, evolutionary algorithm multi-objective optimization, Bayesian optimization sampling update and water inflow and outflow balance optimization decision;

[0027] The balance optimization objective function setting is used to define the objective function and set the constraint conditions, specifically to define an initial objective function group, the initial objective function group specifically includes a water inflow and outflow balance optimization objective function, a patient comfort optimization objective function and a kidney burden optimization objective function, and uses the maximum daily water intake and minimum urine output as constraint conditions;

[0028] The multi-objective Bayesian optimization model is constructed to train the multi-objective Bayesian optimization model, specifically using Gaussian process regression in Bayesian optimization as a proxy model to perform approximate calculations on the initial objective function group, and training the proxy model through an expected improvement strategy to construct the multi-objective Bayesian optimization model;

[0029] The multi-objective optimization of the evolutionary algorithm is specifically to obtain a decision optimization model by selecting a crossover operation and performing decision optimization of a multi-objective Bayesian optimization model based on a fitness function;

[0030] The Bayesian optimization sampling update is specifically to perform iterative training and optimization of the model through the multi-objective optimization of the evolutionary algorithm to obtain the iterative optimal water inflow and outflow balance model Model B ;

[0031] The water inflow and outflow balance optimization decision is specifically to use the iterative optimal water inflow and outflow balance model Model B , based on the water inlet and outlet prediction data, make a water inlet and outlet balance optimization decision to obtain a water inlet and outlet balance optimization plan;

[0032] The water intake and output balance optimization plan specifically includes water intake and output balance optimization plan data, patient comfort optimization plan data and kidney burden optimization plan data.

[0033] Furthermore, the health integrated management module is used to perform health integrated water intake and output in combination with the prediction of renal function deterioration in kidney disease, specifically, the water intake and output balance optimization program is combined with the renal function diagnosis auxiliary system for health integration, and through health integrated management, a comprehensive water intake and output management program for kidney disease patients is obtained.

[0034] Furthermore, the water intake and output intelligent management module is used to perform intelligent water intake and output management of kidney patients, specifically, to perform intelligent water intake and output management of kidney patients according to the comprehensive water intake and output management plan for kidney patients, and to obtain intelligent water intake and output feedback data of patients through the intelligent water intake and output management;

[0035] The patient's water intake and output intelligent feedback data is used for system maintenance and optimization, and specifically includes water intake and output balance feedback data, patient comfort feedback data, and kidney burden feedback data.

[0036] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0037] (1) In the existing intelligent management system for water intake and output of kidney patients, the water intake and output of kidney patients are mostly managed by manual observation, recording and feedback in actual clinical operations. This will inevitably lead to problems of untimely and inefficient communication between doctors and patients. Discharged kidney patients are also unable to properly balance their own water intake and output in long-term health management, and it is difficult for them to obtain professional and reliable guidance, which leads to technical problems such as insufficient intelligence and automation in the management of water intake and output of kidney patients. This solution creatively uses artificial intelligence technology to realize intelligent management of water intake and output of kidney patients. By combining long-term and short-term water intake and output prediction with water intake and output balance optimization, it realizes automated and intelligent water intake and output management, providing strong support for kidney disease medical assistance.

[0038] (2) In view of the technical problem that the existing methods for predicting the water intake and output of kidney patients require both reliable, timely and rapid short-term water intake and output predictions and long-term water intake and output trend predictions, and the existing technology is difficult to take both needs into account, this solution creatively uses time series convolution combined with bidirectional long short-term memory networks for short-term predictions, and uses empirical mode decomposition combined with bidirectional gated recurrent units for long-term predictions. By optimizing the time series relationship that changes sharply in the short term through time series convolution and combining it with bidirectional long short-term network optimization processing, the real-time and response speed of short-term predictions are improved. At the same time, empirical mode decomposition is used to process complex long-term trend data, and the stability of the prediction process is optimized in the bidirectional gated recurrent unit, which comprehensively improves the feasibility and availability of long-term predictions.

[0039] (3) In the existing water intake and output balance optimization methods, there is a problem that personalized water intake and output balance optimization combined with patient personal information is a complex multi-objective optimization problem. It needs to consider not only water intake and output, but also the patient's personal situation and kidney disease condition. Therefore, it is necessary to design a reasonable multi-objective optimization model. This solution creatively adopts the evolutionary algorithm model of multi-objective Bayesian optimization. Through multi-objective optimization, it realizes dynamic adaptation and personalized water intake and output balance improvement, while improving water intake and output balance, patient comfort and kidney disease burden optimization, and improving the overall reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of the structure of an intelligent management system for water intake and output of kidney patients provided by the present invention;

[0041] Figure 2 A flow chart of the steps performed by the water inflow and outflow prediction module;

[0042] Figure 3 This is a schematic diagram of the data visualization analysis results of the water inflow and outflow balance optimization module;

[0043] Figure 4 Flowchart of the steps performed by the water inlet and outlet balance optimization module.

[0044] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0046] Example 1, see Figure 1 The present invention provides an intelligent water intake and output management system for patients with kidney disease, comprising a data collection and processing module, a water intake and output prediction module, a water intake and output balance optimization module, a health integrated management module and a water intake and output intelligent management module;

[0047] The data collection and processing module is used to collect and process the water inflow and outflow data of the kidney patients, obtain the optimized water inflow and outflow management data through data collection and processing, and send the optimized water inflow and outflow management data to the water inflow and outflow prediction module and the water inflow and outflow balance optimization module;

[0048] The water intake and output prediction module is used to predict the water intake and output of kidney patients, obtain water intake and output prediction data through water intake and output prediction, and send the water intake and output prediction to the water intake and output balance optimization module and the health integrated management module;

[0049] The water inflow and outflow balance optimization module is used to predict the water inflow and outflow balance optimization plan in combination with the disease information of the kidney patients, obtain the water inflow and outflow balance optimization plan through water inflow and outflow balance optimization, and send the water inflow and outflow balance optimization plan to the health integrated management module;

[0050] The health integrated management module is used to integrate and combine the water inflow and outflow balance optimization plan for kidney patients with the health management module, obtain a comprehensive water inflow and outflow management plan for kidney patients through health integrated management, and send the comprehensive water inflow and outflow management plan for kidney patients to the water inflow and outflow intelligent management module;

[0051] The water intake and output intelligent management module is used to perform comprehensive management of water intake and output in combination with health integrated management, and to perform intelligent management of water intake and output for kidney disease patients through intelligent management of water intake and output.

[0052] By executing the above operations, in the existing intelligent management system for water intake and output of kidney patients, the water intake and output of kidney patients is mostly managed comprehensively by manual observation, recording and feedback in actual clinical operations, which will inevitably lead to untimely and inefficient communication between doctors and patients, and discharged kidney patients are unable to properly balance their own water intake and output in long-term health management, and it is difficult to obtain professional and reliable guidance, thus leading to technical problems such as insufficient intelligence and automation in the management of water intake and output of kidney patients. This solution creatively uses artificial intelligence technology to realize intelligent management of water intake and output of kidney patients, and realizes automated and intelligent water intake and output management through long-term and short-term water intake and output prediction combined with water intake and output balance optimization, providing strong support for medical assistance for kidney disease.

[0053] Embodiment 2, this embodiment is based on the above embodiment, see Figure 1 The data collection and processing module is used to collect water intake and output data and basic condition data of kidney patients, and optimize and pre-process the collected data, specifically, to obtain water intake and output prediction raw data by collecting from the medical system, medical records and patient self-record data, and to obtain optimized water intake and output management data by optimizing and pre-processing the water intake and output prediction raw data;

[0054] The water intake and output prediction raw data specifically includes sensor data, medical history data and self-water intake and output record data;

[0055] The optimization and preprocessing specifically include data cleaning, data labeling, data standardization and feature extraction.

[0056] Embodiment 3, this embodiment is based on the above embodiment, see Figure 2 The water intake and output prediction module is used to predict the long-term and short-term water intake and output of patients with kidney disease, specifically, to perform long-term water intake and output prediction and short-term water intake and output prediction based on the optimized water intake and output management data to obtain water intake and output prediction data;

[0057] The water inflow and outflow forecast data specifically include long-term water inflow and outflow trend forecast data and short-term water inflow and outflow change forecast data;

[0058] The long-term water intake and output prediction is used to model and predict the long-term trend change law of water intake and output of kidney disease patients, specifically using a bidirectional gated recurrent unit combined with empirical mode decomposition to perform long-term water intake and output prediction to obtain long-term water intake and output trend prediction data;

[0059] The short-term water intake and output prediction is used to provide real-time and rapid short-term water drinking advice and urination prediction for kidney patients. Specifically, a bidirectional long short-term memory network combined with time series convolution is used to perform short-term water intake and output prediction to obtain short-term water intake and output change prediction data.

[0060] The steps of performing long-term water inflow and outflow prediction and short-term water inflow and outflow prediction to obtain water inflow and outflow prediction data include: full set empirical mode decomposition, adaptive noise addition, construction of bidirectional gated recurrent units, construction of temporal convolutional subnets, construction of bidirectional long short-term memory subnets, construction of attention residual layers and water inflow and outflow prediction;

[0061] The full set empirical mode decomposition is specifically to perform a full set empirical mode decomposition on the time series data in the optimized water inlet and outlet management data to obtain intrinsic mode function data;

[0062] The adaptive noise addition is specifically to apply adaptive noise in the full set empirical mode decomposition to assist time series decomposition;

[0063] The constructing of the bidirectional gated recurrent unit specifically involves constructing a bidirectional gated recurrent unit to perform bidirectional structural time series prediction on each subsequence in the intrinsic mode function data to perform long-term water inflow and outflow trend prediction;

[0064] The constructing of the time series convolution subnet specifically involves constructing a time series convolution network, extracting short-term dependency features of data in the optimized water inflow and outflow management data through a causal convolution layer, and obtaining short-term feature data;

[0065] The constructing of the bidirectional long short-term memory subnet is specifically to construct a standard bidirectional long short-term memory neural network, perform bidirectional structural time series prediction on the short-term feature data, and perform short-term water inflow and outflow change prediction;

[0066] The construction of the attention residual layer specifically introduces the attention mechanism and constructs a residual network to optimize the gradient disappearance problem;

[0067] The water inflow and outflow prediction is specifically to train a long-term water inflow and outflow prediction model through the full set empirical mode decomposition, the adaptive noise addition and the bidirectional gated recurrent unit to obtain a long-term water inflow and outflow prediction model Model L , and train the short-term water inflow and outflow prediction model through the temporal convolution subnet, the bidirectional long short-term memory subnet and the attention residual layer to obtain the short-term water inflow and outflow prediction model Model S , and by using the long-term water inflow and outflow prediction model L And the short-term water inflow and outflow prediction model S , predict water inflow and outflow, and obtain water inflow and outflow prediction data.

[0068] By executing the above operations, in view of the technical problem that in the existing methods for predicting the water intake and output of kidney patients, kidney patients need both reliable, timely and rapid short-term water intake and output predictions, and long-term water intake and output trend predictions, and the existing technology is difficult to take both needs into account, this solution creatively adopts time series convolution combined with bidirectional long short-term memory network for short-term prediction, and adopts empirical mode decomposition combined with bidirectional gated recurrent unit for long-term prediction. By optimizing the time series relationship that changes sharply in the short term through time series convolution, and combining the bidirectional long short-term network optimization processing, the real-time and response speed of short-term prediction are improved. At the same time, empirical mode decomposition is used to process complex long-term trend data, and the stability of the prediction process is optimized in the bidirectional gated recurrent unit, which comprehensively improves the feasibility and availability of long-term prediction.

[0069] Embodiment 4, this embodiment is based on the above embodiment, see Figure 3 and Figure 4 The water inflow and outflow balance optimization module is used to further predict the optimization of drinking water planning in combination with the water inflow and outflow prediction results of kidney patients. Specifically, based on the water inflow and outflow prediction data, an evolutionary algorithm model combined with multi-objective Bayesian optimization is used to optimize the water inflow and outflow balance to obtain a water inflow and outflow balance optimization plan, which specifically includes the following steps: balance optimization objective function setting, construction of a multi-objective Bayesian optimization model, evolutionary algorithm multi-objective optimization, Bayesian optimization sampling update and water inflow and outflow balance optimization decision;

[0070] The balance optimization objective function setting is used to define the objective function and set the constraint conditions, specifically to define an initial objective function group, the initial objective function group specifically includes a water inflow and outflow balance optimization objective function, a patient comfort optimization objective function and a kidney burden optimization objective function, and uses the maximum daily water intake and minimum urine output as constraint conditions;

[0071] The multi-objective Bayesian optimization model is constructed to train the multi-objective Bayesian optimization model, specifically using Gaussian process regression in Bayesian optimization as a proxy model to perform approximate calculations on the initial objective function group, and training the proxy model through an expected improvement strategy to construct the multi-objective Bayesian optimization model;

[0072] The multi-objective optimization of the evolutionary algorithm is specifically to obtain a decision optimization model by selecting a crossover operation and performing decision optimization of a multi-objective Bayesian optimization model based on a fitness function;

[0073] The Bayesian optimization sampling update is specifically to perform iterative training and optimization of the model through the multi-objective optimization of the evolutionary algorithm to obtain the iterative optimal water inflow and outflow balance model Model B ;

[0074] The water inflow and outflow balance optimization decision is specifically to use the iterative optimal water inflow and outflow balance model Model B , based on the water inlet and outlet prediction data, make a water inlet and outlet balance optimization decision to obtain a water inlet and outlet balance optimization plan;

[0075] The water intake and output balance optimization plan specifically includes water intake and output balance optimization plan data, patient comfort optimization plan data and kidney burden optimization plan data.

[0076] See also Figure 3 ,In the figure, the horizontal direction is the water inlet and outlet balance ,optimization program data, the vertical direction is the patient comfort ,optimization program data, and the depth direction is the kidney burden ,optimization program data.

[0077] By performing the above operations, in the existing water intake and output balance optimization methods, there is a problem that personalized water intake and output balance optimization combined with patient personal information is a complex multi-objective optimization problem. It not only needs to consider water intake and output, but also needs to be combined with the patient's personal situation and kidney disease condition. Therefore, it is necessary to design a reasonable multi-objective optimization model. The technical problem, this solution creatively adopts the evolutionary algorithm model of multi-objective Bayesian optimization, and realizes dynamic adaptation and personalized water intake and output balance improvement through multi-objective optimization, while improving water intake and output balance, patient comfort and kidney disease burden optimization, and improving the overall reliability of the system.

[0078] Embodiment 5, this embodiment is based on the above embodiment, see Figure 1The health integrated management module is used to perform a health integrated management of water intake and output in combination with the prediction of renal function deterioration in kidney disease. Specifically, the water intake and output balance optimization program is combined with the renal function diagnosis auxiliary system for health integration. Through health integrated management, a comprehensive water intake and output management program for kidney disease patients is obtained.

[0079] Embodiment 6, this embodiment is based on the above embodiment, see Figure 1 The water intake and output intelligent management module is used to perform intelligent water intake and output management for patients with kidney disease, specifically to perform intelligent water intake and output management for patients with kidney disease according to the comprehensive water intake and output management plan for patients with kidney disease, and to obtain intelligent water intake and output feedback data for patients through the intelligent water intake and output management;

[0080] The patient's water intake and output intelligent feedback data is used for system maintenance and optimization, and specifically includes water intake and output balance feedback data, patient comfort feedback data, and kidney burden feedback data.

[0081] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process and method including a series of elements include not only those elements, but also other elements not explicitly listed, or also include elements inherent to such process and method.

[0082] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0083] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent management system for water intake and output of patients with kidney disease, characterized by: It includes data collection and processing module, water inflow and outflow prediction module, water inflow and outflow balance optimization module, health integrated management module and water inflow and outflow intelligent management module; The data collection and processing module obtains raw water inflow and outflow prediction data through collection, and obtains optimized water inflow and outflow management data through optimization and preprocessing; The water inflow and outflow prediction module specifically performs long-term water inflow and outflow prediction and short-term water inflow and outflow prediction based on the optimized water inflow and outflow management data to obtain water inflow and outflow prediction data; The steps of performing long-term water inflow and outflow prediction and short-term water inflow and outflow prediction to obtain water inflow and outflow prediction data include: full set empirical mode decomposition, adaptive noise addition, construction of bidirectional gated recurrent units, construction of temporal convolutional subnets, construction of bidirectional long short-term memory subnets, construction of attention residual layers and water inflow and outflow prediction; The water inflow and outflow balance optimization module is specifically based on the water inflow and outflow prediction data, using an evolutionary algorithm model combined with multi-objective Bayesian optimization to optimize the water inflow and outflow balance, and obtain a water inflow and outflow balance optimization plan, which specifically includes the following steps: balance optimization objective function setting, building a multi-objective Bayesian optimization model, evolutionary algorithm multi-objective optimization, Bayesian optimization sampling update and water inflow and outflow balance optimization decision; The health integrated management module obtains a comprehensive management plan for water intake and output of kidney patients through health integrated management; The water intake and output intelligent management module obtains the patient's water intake and output intelligent feedback data through the water intake and output intelligent management.

2. The intelligent water intake and output management system for kidney patients according to claim 1, characterized in that: The water inflow and outflow forecast data specifically include long-term water inflow and outflow trend forecast data and short-term water inflow and outflow change forecast data; The long-term water intake and output prediction is used to model and predict the long-term trend change law of water intake and output of kidney disease patients, specifically using a bidirectional gated recurrent unit combined with empirical mode decomposition to perform long-term water intake and output prediction to obtain long-term water intake and output trend prediction data; The short-term water intake and output prediction is used to provide real-time and rapid short-term water drinking advice and urination prediction for kidney patients. Specifically, a bidirectional long short-term memory network combined with time series convolution is used to perform short-term water intake and output prediction to obtain short-term water intake and output change prediction data.

3. The intelligent water intake and output management system for kidney patients according to claim 2 is characterized by: The full set empirical mode decomposition is specifically to perform a full set empirical mode decomposition on the time series data in the optimized water inlet and outlet management data to obtain intrinsic mode function data; The adaptive noise addition is specifically to apply adaptive noise in the full set empirical mode decomposition to assist time series decomposition; The constructing of the bidirectional gated recurrent unit specifically involves constructing a bidirectional gated recurrent unit to perform bidirectional structural time series prediction on each subsequence in the intrinsic mode function data and to perform long-term water inflow and outflow trend prediction.

4. The intelligent water intake and output management system for kidney patients according to claim 3 is characterized by: The constructing of the time series convolution subnet specifically involves constructing a time series convolution network, extracting short-term dependency features of data in the optimized water inflow and outflow management data through a causal convolution layer, and obtaining short-term feature data; The constructing of the bidirectional long short-term memory subnet is specifically to construct a standard bidirectional long short-term memory neural network, perform bidirectional structural time series prediction on the short-term feature data, and perform short-term water inflow and outflow change prediction; The construction of the attention residual layer specifically introduces the attention mechanism and constructs a residual network to optimize the gradient disappearance problem; The water inflow and outflow prediction is specifically to perform long-term water inflow and outflow prediction model training through the full set empirical mode decomposition, the adaptive noise addition and the bidirectional gated recurrent unit to obtain a long-term water inflow and outflow prediction model ModelL, and to perform short-term water inflow and outflow prediction model training through the temporal convolutional subnet, the bidirectional long short-term memory subnet and the attention residual layer to obtain a short-term water inflow and outflow prediction model ModelS, and to perform water inflow and outflow prediction by using the long-term water inflow and outflow prediction model ModelL and the short-term water inflow and outflow prediction model ModelS to obtain water inflow and outflow prediction data.

5. The intelligent water intake and output management system for kidney patients according to claim 4, characterized in that: The water inflow and outflow balance optimization module is used to further predict the optimization of drinking water planning in combination with the water inflow and outflow prediction results of kidney patients. Specifically, based on the water inflow and outflow prediction data, an evolutionary algorithm model combined with multi-objective Bayesian optimization is used to optimize the water inflow and outflow balance to obtain a water inflow and outflow balance optimization plan, which specifically includes the following steps: balance optimization objective function setting, construction of a multi-objective Bayesian optimization model, evolutionary algorithm multi-objective optimization, Bayesian optimization sampling update and water inflow and outflow balance optimization decision; The balance optimization objective function setting is used to define the objective function and set the constraint conditions, specifically to define an initial objective function group, the initial objective function group specifically includes a water inflow and outflow balance optimization objective function, a patient comfort optimization objective function and a kidney burden optimization objective function, and uses the maximum daily water intake and minimum urine output as constraint conditions; The multi-objective Bayesian optimization model is constructed to train the multi-objective Bayesian optimization model, specifically using Gaussian process regression in Bayesian optimization as a proxy model to perform approximate calculations on the initial objective function group, and training the proxy model through an expected improvement strategy to construct the multi-objective Bayesian optimization model; The multi-objective optimization of the evolutionary algorithm is specifically to obtain a decision optimization model by selecting a crossover operation and performing decision optimization of a multi-objective Bayesian optimization model based on a fitness function; The Bayesian optimization sampling update is specifically to perform iterative training and optimization of the model through the evolutionary algorithm multi-objective optimization to obtain the iterative optimal water inflow and outflow balance model Model B ; The water inflow and outflow balance optimization decision is specifically to use the iterative optimal water inflow and outflow balance model Model B , based on the water inlet and outlet prediction data, make a water inlet and outlet balance optimization decision to obtain a water inlet and outlet balance optimization plan; The water intake and output balance optimization plan specifically includes water intake and output balance optimization plan data, patient comfort optimization plan data and kidney burden optimization plan data.

6. The intelligent water intake and output management system for kidney patients according to claim 5, characterized in that: The health integrated management module is used to perform health integrated water intake and output in combination with the prediction of renal function deterioration in kidney disease. Specifically, the water intake and output balance optimization program is combined with the renal function diagnosis auxiliary system for health integration. Through health integrated management, a comprehensive water intake and output management program for kidney disease patients is obtained.

7. The intelligent water intake and output management system for kidney patients according to claim 6, characterized in that: The water intake and output intelligent management module is used to perform intelligent water intake and output management of kidney patients, specifically to perform intelligent water intake and output management of kidney patients according to the comprehensive water intake and output management plan for kidney patients, and to obtain intelligent water intake and output feedback data of patients through the intelligent water intake and output management; The patient's water intake and output intelligent feedback data is used for system maintenance and optimization, and specifically includes water intake and output balance feedback data, patient comfort feedback data, and kidney burden feedback data.

8. The intelligent water intake and output management system for kidney patients according to claim 7, characterized in that: The data collection and processing module is used to collect and process the water inflow and outflow data of the kidney patients, obtain the optimized water inflow and outflow management data through data collection and processing, and send the optimized water inflow and outflow management data to the water inflow and outflow prediction module and the water inflow and outflow balance optimization module; The water intake and output prediction module is used to predict the water intake and output of kidney patients, obtain water intake and output prediction data through water intake and output prediction, and send the water intake and output prediction to the water intake and output balance optimization module and the health integrated management module; The water inflow and outflow balance optimization module is used to predict the water inflow and outflow balance optimization plan in combination with the disease information of the kidney patients, obtain the water inflow and outflow balance optimization plan through water inflow and outflow balance optimization, and send the water inflow and outflow balance optimization plan to the health integrated management module; The health integrated management module is used to integrate and combine the water inflow and outflow balance optimization plan for kidney patients with the health management module, obtain a comprehensive water inflow and outflow management plan for kidney patients through health integrated management, and send the comprehensive water inflow and outflow management plan for kidney patients to the water inflow and outflow intelligent management module; The water intake and output intelligent management module is used to perform comprehensive management of water intake and output in combination with health integrated management, and to perform intelligent management of water intake and output for kidney disease patients through intelligent management of water intake and output.

Images