CN117809811B - Artificial intelligence-based weight-reduction operation postoperative management method and system - Google Patents
Artificial intelligence-based weight-reduction operation postoperative management method and system Download PDFInfo
- Publication number
- CN117809811B CN117809811B CN202410217422.8A CN202410217422A CN117809811B CN 117809811 B CN117809811 B CN 117809811B CN 202410217422 A CN202410217422 A CN 202410217422A CN 117809811 B CN117809811 B CN 117809811B
- Authority
- CN
- China
- Prior art keywords
- weight
- patient
- readmission
- model
- layer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000007726 management method Methods 0.000 title claims abstract description 81
- 230000002980 postoperative effect Effects 0.000 title claims abstract description 50
- 208000016261 weight loss Diseases 0.000 title claims abstract description 37
- 239000013585 weight reducing agent Substances 0.000 title claims abstract description 30
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 16
- 238000000034 method Methods 0.000 claims abstract description 33
- 238000013528 artificial neural network Methods 0.000 claims abstract description 28
- 230000036541 health Effects 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 13
- 239000011159 matrix material Substances 0.000 claims description 66
- 230000006870 function Effects 0.000 claims description 63
- 230000004913 activation Effects 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 21
- 238000007781 pre-processing Methods 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 12
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 238000003066 decision tree Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 7
- 230000004580 weight loss Effects 0.000 claims description 7
- 230000037213 diet Effects 0.000 claims description 6
- 235000005911 diet Nutrition 0.000 claims description 6
- 239000003814 drug Substances 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 230000037396 body weight Effects 0.000 claims description 3
- 229940079593 drug Drugs 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000012552 review Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 claims description 2
- 238000003062 neural network model Methods 0.000 abstract description 4
- 238000012545 processing Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000011084 recovery Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 2
- 238000001356 surgical procedure Methods 0.000 description 2
- 208000008589 Obesity Diseases 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 230000002779 inactivation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000020824 obesity Nutrition 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/40—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Epidemiology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Surgery (AREA)
- Urology & Nephrology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a weight-reduction operation postoperative management method and system based on artificial intelligence. The invention belongs to the technical field of medical health management, and particularly relates to a weight-reducing operation postoperative management method and system based on artificial intelligence, wherein an Elman circulating neural network is adopted for predicting the weight of a patient, so that the processing capacity of a model on dynamic information is improved, the adaptability and generalization capacity of the neural network model are improved, the training efficiency is improved while the accuracy is improved, and the overall usability of the method is enhanced; the characteristic selection is carried out by adopting the Bo Lu Da algorithm combined with the limit gradient lifting model, and the postoperative hospital admission prediction is further carried out, so that the randomness of the traditional limit gradient lifting model is reduced, the characteristics of high precision, high robustness and low randomness are achieved, and the practicability of the weight-reduction postoperative management system is improved.
Description
Technical Field
The invention belongs to the technical field of medical health management, and particularly relates to a weight-reduction operation postoperative management method and system based on artificial intelligence.
Background
The postoperative management of the weight reduction surgery aims at finding the potential health risk of the patients after the weight reduction surgery in advance by using an artificial intelligence technology, helping medical staff to make management strategies efficiently and take intervention measures timely, so that the readmission rate is reduced, the postoperative rehabilitation effect of the patients is improved, the requirement of personalized management of the patients is met, effective weight reduction is facilitated for people, and the health problem related to obesity is improved. However, in the existing management process after the weight-reducing operation, there is a technical problem that the weight of a patient required by the management after the weight-reducing operation is predicted, the requirement on the performance of a model is high, and a neural network with high accuracy, high adaptability and strong generalization capability is required; the existing readmission prediction model has low robustness, so that an actual prediction result is unreliable, and the practical technical problem of the weight-reduction postoperative management system is further affected.
Disclosure of Invention
Aiming at the problems that in the existing weight-reduction operation post-operation management process, the weight prediction of a patient required by the weight-reduction operation post-operation management is high in requirements on model performance and needs a neural network with high accuracy, high adaptability and high generalization capability, the technical scheme creatively adopts an Elman circulating neural network to predict the weight of the patient, and increases the processing capability of a model on dynamic information by adding a context layer behind a hidden layer, further improves the adaptability and generalization capability of the neural network model, improves the accuracy and the training efficiency, and enhances the overall usability of the method; aiming at the technical problems that in the existing management process after the weight-reducing operation, the existing readmission prediction model has low robustness and unreliable actual prediction results, thereby influencing the practicability of the management system after the weight-reducing operation; the scheme creatively adopts the Bo Lu Da algorithm combined with the limit gradient lifting model to perform feature selection, further performs postoperative readmission prediction, reduces the randomness of the traditional limit gradient lifting model by introducing the shadow feature matrix and the Z score, has the characteristics of high precision, high robustness and low randomness, improves the practicability of the weight-reduction postoperative management system, and further is beneficial to medical staff to efficiently formulate a management strategy according to the condition of a patient and promotes postoperative recovery of the patient.
The technical scheme adopted by the invention is as follows: the invention provides a weight-reduction operation postoperative management method based on artificial intelligence, which comprises the following steps:
step S1: patient data acquisition, in particular to acquisition of patient clinical data;
Step S2: preprocessing data;
Step S3: the method comprises the steps of predicting the weight of a patient, specifically, screening clinical initial data through a Pearson correlation coefficient to obtain weight prediction characteristics, constructing a patient weight prediction model through an Elman circulating neural network, and predicting the weight through the constructed patient weight prediction model to obtain the predicted weight of the patient;
Step S4: prediction of hospital readmission after operation;
Step S5: and generating a post-operation management strategy.
Further, in step S1, the patient clinical data includes patient basic information including sex, age, body mass index, specifically BMI index, a visit record describing a post-operation weight and a reason for readmission and readmission within 3 months after the operation of the patient, a medical history, operation information, and a follow-up record, specifically body weight after 3 months after the operation.
Further, in step S2, the data preprocessing, specifically, performing outlier removal, denoising, missing value filling and normalization on the clinical data of the patient, to obtain clinical initial data.
Further, in step S3, the patient weight prediction is used for predicting the weight of the patient after the weight-loss operation, and includes the following steps:
Step S31: the method comprises the steps of adopting pearson correlation coefficients to perform feature screening on clinical initial data, specifically calculating pearson correlation coefficients of each variable in the clinical initial data and postoperative weight, and selecting k variables with highest pearson correlation coefficients to obtain weight prediction features;
step S32: the patient weight prediction model construction, in particular to a patient weight prediction model construction by adopting an Elman circulating neural network, comprises the following steps:
Step S321: building an Elman circulating neural network structure, wherein the Elman circulating neural network structure comprises an input layer, a hidden layer, a context layer and an output layer;
The input layer receives the weight prediction feature as an input feature, and sets the number of input layer nodes according to the dimension of the weight prediction feature, so that the number of the input layer nodes is equal to the dimension of the weight prediction feature;
the hidden layer sets a Sigmoid activation function as a hidden layer activation function;
the output layer sets a linear activation function as an output layer activation function;
The context layer is used for memorizing the output value of the hidden layer at the previous moment and feeding back the output value of the hidden layer to the hidden layer again;
Step S322: calculating initial output of the hidden layer by adopting a Sigmoid activation function, taking the initial output of the hidden layer as input of a context layer, feeding back the initial output of the hidden layer to the hidden layer through the context layer, and calculating to obtain final output of the hidden layer, wherein a calculation formula is as follows:
hlout(t)＝σ(Wt·y(t)+Wt-1·hlout(t-1)+g1)；
wherein t is a time step index, h out2 (t) is a final output of the hidden layer, the final output of the hidden layer is used for representing the hidden layer output at the current moment, sigma (·) is a Sigmoid activation function, W t is a hidden layer weight matrix at the current moment, y (t) is an input feature at the current moment, W t-1 is a weight matrix of the hidden layer at the previous moment, hl out1 (t-1) is an initial output of the hidden layer, the initial output of the hidden layer is used for representing the hidden layer output at the previous moment, and g 1 is a hidden layer bias term;
step S323: calculating the output of the output layer, which is used for calculating the output result of the model, wherein the calculation formula is as follows:
olout(t)＝ε(W1t·hlout(t)+g2)；
Wherein, ol out (t) is the output of the output layer at the current moment, epsilon (·) is a linear activation function, W1 t is the output layer weight matrix at the current moment, and g 2 is the output layer bias term;
Step S324: calculating a loss function of the model, and updating model parameters by back-propagating the loss function;
Step S325: performing model training by repeatedly executing the steps S322, S323 and S324 for N times by adopting a gradient descent algorithm with momentum to obtain a patient weight prediction model;
Step S33: and predicting the weight by adopting a patient weight prediction model to obtain the predicted weight of the patient.
Further, in step S4, the post-operation readmission prediction is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, and includes the following steps:
Step S41: extracting features of clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, wherein the one-dimensional convolutional neural network comprises a one-dimensional convolutional layer, two maximum pooling layers, a flattening layer, a first full-connection layer, a second full-connection layer and a regularization layer;
Step S42: the limiting gradient lifting model is combined with the Bo Lu Da algorithm to obtain a Bo Lu Da algorithm combined with the limiting gradient lifting model;
step S43: the feature selection is carried out by adopting a Bo Lu Da algorithm combined with a limit gradient lifting model, and the method comprises the following steps:
Step S431: constructing an input feature matrix, namely randomly arranging the readmission feature matrix to obtain a shadow feature matrix, and splicing the readmission feature matrix and the shadow feature matrix to obtain the input feature matrix;
Step S432: training a limit gradient lifting model, specifically, training the model by minimizing a loss function and an objective function according to a gradient lifting principle, comprising the following steps:
Step S4321: calculating an objective function, which is used for determining an optimization target of the model, wherein a calculation formula is as follows:
Where K is the decision tree index, mf (K) is the objective function, j is the sample index, the samples are used to represent the samples in the input feature matrix, M is the number of samples, the number of samples are used to represent the number of samples in the input feature matrix, Is the loss function, x j is the actual value of the j-th sample,/>Is the predicted value of the jth sample, ψ (f K) is the regularization term for controlling the complexity of the model, f K is the predicted value of the kth decision tree;
step S4322: and calculating a loss function, wherein the loss function is used for measuring the difference between a model predicted value and an actual value, and the calculation formula is as follows:
step S4323: the performance of the K decision tree is evaluated by adopting a second-order Taylor expansion of the objective function, the performance is used for minimizing the objective function, and the calculation formula is as follows:
In the method, in the process of the invention, Is the second order Taylor expansion of the objective function, i is the leaf node index, lno is the number of leaf nodes, FD i is the first derivative of the loss function for the ith leaf node, SD i is the second derivative of the loss function for the ith leaf node, α is the regularization weight, β is the model hyper-parameter, for controlling the number of leaf nodes;
Step S433: and calculating a gain value of the leaf node for evaluating the importance of the feature, wherein the calculation formula is as follows:
Where leaf g is the gain value of the leaf node, FD L is the first derivative of the left node, SD L is the second derivative of the left node, FD R is the first derivative of the right node, SD R is the second derivative of the right node;
step S434: judging the importance of each feature in the readmission feature matrix according to the Z score, and removing the readmission feature from the readmission feature matrix if the Z score of the readmission feature is lower than the maximum Z score of the shadow feature, wherein the calculation formula of the Z score is as follows:
wherein Z (K) is the Z fraction, Is the average of the gain values of the leaf nodes,/>Is the standard deviation of the gain values of the leaf nodes;
Step S435: repeatedly executing the steps S431, S432, S433 and S434 for K max times to obtain a readmission critical feature matrix, wherein the readmission critical feature matrix is used for representing the readmission feature matrix after feature selection;
Step S44: according to the predicted weight of the patient and the readmission critical feature matrix, the postoperative readmission prediction is carried out by adopting an artificial neural network model, and the patient is divided into a readmission high risk group and a readmission non-high risk group, so that the postoperative readmission risk is obtained.
Further, in step S5, the post-operation management policy is generated, and the post-operation management policy is generated according to the predicted weight of the patient and the risk of readmission after the operation, specifically, a health management plan is formulated based on the predicted weight of the patient, the health management plan includes diet, exercise and medication management plans, and based on the risk of readmission after the operation, the post-operation rehabilitation monitoring is enhanced and the frequency of readmission is increased for the patient with high risk of readmission after the operation.
The invention provides an artificial intelligence-based weight-reduction postoperative management system, which comprises: the device comprises a patient data acquisition module, a data preprocessing module, a patient weight prediction module, a postoperative readmission prediction module and a postoperative management strategy generation module;
the patient data acquisition module is used for acquiring patient data, specifically acquiring related data after a weight-reduction operation to obtain patient clinical data, and sending the patient clinical data to the data preprocessing module;
The data preprocessing module is used for preprocessing data, specifically, clinical initial data is obtained by preprocessing clinical data of a patient, and the clinical initial data is sent to the patient weight prediction module and the postoperative hospital readmission prediction module;
The patient weight prediction module is used for predicting the weight of the patient after 3 months of the weight reduction operation, specifically, the clinical initial data is subjected to feature screening through a Pearson correlation coefficient to obtain weight prediction features, an Elman circulating neural network is adopted to construct a patient weight prediction model, a patient weight prediction model is obtained, the patient weight prediction model is adopted to conduct weight prediction to obtain the patient predicted weight, and the patient predicted weight is sent to the postoperative hospital-admission prediction module;
The post-operation readmission prediction module is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, specifically, extracting features of the clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, performing feature selection through a Bo Lu Da algorithm combined with a limit gradient lifting model to obtain a readmission key feature matrix, performing post-operation readmission prediction according to the predicted weight of the patient and the readmission key feature matrix through an artificial neural network model to obtain post-operation readmission risk, and sending the post-operation readmission risk to the post-operation management strategy generation module;
The post-operation management strategy generation module is used for generating a post-operation management strategy, and particularly for making a health management plan based on predicted weight of a patient, wherein the health management plan comprises diet, exercise and medicine management plans, and based on post-operation readmission risk, post-operation rehabilitation monitoring is enhanced and review frequency is increased for patients with high readmission risk.
By adopting the scheme, the beneficial effects obtained by the invention are as follows:
(1) Aiming at the technical problems that in the existing weight-reduction operation postoperative management process, the weight prediction of a patient required by the weight-reduction operation postoperative management is high in requirement on model performance, and a neural network with high accuracy, adaptability and generalization capability is required, the method creatively adopts the Elman circulating neural network to conduct the weight prediction of the patient, increases the processing capability of the model on dynamic information by adding a context layer behind a hidden layer, further improves the adaptability and generalization capability of the neural network model, improves the training efficiency while improving the accuracy, and enhances the overall usability of the method.
(2) Aiming at the technical problems that in the existing management process after the weight-reducing operation, the existing readmission prediction model has low robustness and unreliable actual prediction results, thereby influencing the practicability of the management system after the weight-reducing operation; the scheme creatively adopts the Bo Lu Da algorithm combined with the limit gradient lifting model to perform feature selection, further performs postoperative readmission prediction, reduces the randomness of the traditional limit gradient lifting model by introducing the shadow feature matrix and the Z score, has the characteristics of high precision, high robustness and low randomness, improves the practicability of the weight-reduction postoperative management system, and further is beneficial to medical staff to efficiently formulate a management strategy according to the condition of a patient and promotes postoperative recovery of the patient.
Drawings
FIG. 1 is a schematic flow chart of a weight-loss operation post-operation management method based on artificial intelligence;
FIG. 2 is a schematic diagram of a weight-loss operation post-operation management system based on artificial intelligence;
FIG. 3 is a flow chart of step S3;
fig. 4 is a flow chart of step S4.
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention; all other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," "outer," and the like indicate orientation or positional relationships based on those shown in the drawings, merely to facilitate description of the invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus should not be construed as limiting the invention.
First embodiment, referring to fig. 1, the present invention provides a weight-loss operation post-operation management method based on artificial intelligence, which includes the following steps:
step S1: patient data acquisition, in particular to acquisition of patient clinical data;
Step S2: preprocessing data;
Step S3: the method comprises the steps of predicting the weight of a patient, specifically, screening clinical initial data through a Pearson correlation coefficient to obtain weight prediction characteristics, constructing a patient weight prediction model through an Elman circulating neural network, and predicting the weight through the constructed patient weight prediction model to obtain the predicted weight of the patient;
Step S4: prediction of hospital readmission after operation;
Step S5: and generating a post-operation management strategy.
In step S1, the patient clinical data includes patient basic information including sex, age, body mass index, and specifically BMI index, the patient clinical data includes patient basic information including operation type, operation duration, and operation hospitalization duration, a medical history, operation information including patient basic information including sex, age, and body mass index, and a follow-up record describing a post-operation body mass and a patient' S post-operation readmission and a readmission reason within 3 months, and the post-operation body mass specifically refers to a post-operation body mass of 3 months.
In step S2, the data preprocessing, specifically, performing outlier removal, denoising, missing value filling and normalization operations on the clinical data of the patient, to obtain clinical initial data, with reference to fig. 1.
In a fourth embodiment, referring to fig. 1 and 3, the method is based on the above embodiment, in step S3, the patient weight prediction is used for predicting the weight of the patient after the weight-loss operation, and the method includes the following steps:
Step S31: the method comprises the steps of adopting pearson correlation coefficients to perform feature screening on clinical initial data, specifically calculating pearson correlation coefficients of each variable in the clinical initial data and postoperative weight, and selecting k variables with highest pearson correlation coefficients to obtain weight prediction features;
step S32: the patient weight prediction model construction, in particular to a patient weight prediction model construction by adopting an Elman circulating neural network, comprises the following steps:
step S321: building an Elman circulating neural network structure, wherein the Elman circulating neural network structure comprises an input layer, a hidden layer, a context layer and an output layer, and the learning rate is set to be 0.001;
the input layer receives the weight prediction features as input features, and sets the number of nodes of the input layer according to the dimension of the weight prediction features, so that the number of nodes of the input layer is equal to the dimension of the weight prediction features;
the hidden layer sets a Sigmoid activation function as a hidden layer activation function;
the output layer sets a linear activation function as an output layer activation function;
The context layer is used for memorizing the output value of the hidden layer at the previous moment and feeding back the output value of the hidden layer to the hidden layer again;
Step S322: calculating initial output of the hidden layer by adopting a Sigmoid activation function, taking the initial output of the hidden layer as input of a context layer, feeding back the initial output of the hidden layer to the hidden layer through the context layer, and calculating to obtain final output of the hidden layer, wherein a calculation formula is as follows:
hlout(t)＝σ(Wt·y(t)+Wt-1·hlout(t-1)+g1)；
wherein t is a time step index, h out2 (t) is a final output of the hidden layer, the final output of the hidden layer is used for representing the hidden layer output at the current moment, sigma (·) is a Sigmoid activation function, W t is a hidden layer weight matrix at the current moment, y (t) is an input feature at the current moment, W t-1 is a weight matrix of the hidden layer at the previous moment, hl out1 (t-1) is an initial output of the hidden layer, the initial output of the hidden layer is used for representing the hidden layer output at the previous moment, and g 1 is a hidden layer bias term;
step S323: calculating the output of the output layer, which is used for calculating the output result of the model, wherein the calculation formula is as follows:
olout(t)＝ε(W1t·hlout(t)+g2)；
Wherein, ol out (t) is the output of the output layer at the current moment, epsilon (·) is a linear activation function, W1 t is the output layer weight matrix at the current moment, and g 2 is the output layer bias term;
Step S324: calculating a loss function of the model, and updating model parameters by back-propagating the loss function;
Step S325: performing model training by repeatedly executing the steps S322, S323 and S324 for N times by adopting a gradient descent algorithm with momentum to obtain a patient weight prediction model;
Step S33: predicting the weight by adopting a patient weight prediction model to obtain the predicted weight of the patient;
Through executing the operation, aiming at the technical problems that in the existing weight-reduction operation postoperative management process, the weight prediction of a patient required by the weight-reduction operation postoperative management exists, the requirement on the model performance is high, and a neural network with high accuracy, high adaptability and strong generalization capability is required, the technical scheme creatively adopts the Elman circulating neural network to carry out the weight prediction of the patient, and increases the processing capability of the model to dynamic information by adding a context layer behind a hidden layer, and further improves the adaptability and generalization capability of the neural network model, improves the training efficiency and the overall usability of the method while improving the accuracy.
Embodiment five, referring to fig. 1 and 4, the embodiment is based on the above embodiment, in step S4, the post-operation readmission prediction is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, and includes the following steps:
Step S41: extracting features of clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, wherein the one-dimensional convolutional neural network comprises a one-dimensional convolutional layer, two maximum pooling layers, a flattening layer, a first full-connection layer, a second full-connection layer and a regularization layer;
the one-dimensional convolution layer is provided with 30 convolution kernels, the first full-connection layer is provided with 20 neurons, the second full-connection layer is provided with 1 neuron, and the random inactivation rate of the regularization layer is set to be 0.5;
Step S42: the limiting gradient lifting model is combined with the Bo Lu Da algorithm to obtain a Bo Lu Da algorithm combined with the limiting gradient lifting model;
step S43: the feature selection is carried out by adopting a Bo Lu Da algorithm combined with a limit gradient lifting model, and the method comprises the following steps:
Step S431: constructing an input feature matrix, namely randomly arranging the readmission feature matrix to obtain a shadow feature matrix, and splicing the readmission feature matrix and the shadow feature matrix to obtain the input feature matrix;
Step S432: training a limit gradient lifting model, specifically, training the model by minimizing a loss function and an objective function according to a gradient lifting principle, comprising the following steps:
Step S4321: calculating an objective function, which is used for determining an optimization target of the model, wherein a calculation formula is as follows:
Where K is the decision tree index, mf (K) is the objective function, j is the sample index, the samples are used to represent the samples in the input feature matrix, M is the number of samples, the number of samples are used to represent the number of samples in the input feature matrix, Is the loss function, x j is the actual value of the j-th sample,/>Is the predicted value of the jth sample, ψ (f K) is the regularization term for controlling the complexity of the model, f K is the predicted value of the kth decision tree;
step S4322: and calculating a loss function, wherein the loss function is used for measuring the difference between a model predicted value and an actual value, and the calculation formula is as follows:
step S4323: the performance of the K decision tree is evaluated by adopting a second-order Taylor expansion of the objective function, the performance is used for minimizing the objective function, and the calculation formula is as follows:
In the method, in the process of the invention, Is the second order Taylor expansion of the objective function, i is the leaf node index, lno is the number of leaf nodes, FD i is the first derivative of the loss function for the ith leaf node, SD i is the second derivative of the loss function for the ith leaf node, α is the regularization weight, β is the model hyper-parameter, for controlling the number of leaf nodes;
Step S433: and calculating a gain value of the leaf node for evaluating the importance of the feature, wherein the calculation formula is as follows:
Where leaf g is the gain value of the leaf node, FD L is the first derivative of the left node, SD L is the second derivative of the left node, FD R is the first derivative of the right node, SD R is the second derivative of the right node;
step S434: judging the importance of each feature in the readmission feature matrix according to the Z score, and removing the readmission feature from the readmission feature matrix if the Z score of the readmission feature is lower than the maximum Z score of the shadow feature, wherein the calculation formula of the Z score is as follows:
wherein Z (K) is the Z fraction, Is the average of the gain values of the leaf nodes,/>Is the standard deviation of the gain values of the leaf nodes;
Step S435: repeatedly executing the steps S431, S432, S433 and S434 for K max times to obtain a readmission critical feature matrix, wherein the readmission critical feature matrix is used for representing the readmission feature matrix after feature selection;
Step S44: according to the predicted weight of the patient and the readmission critical feature matrix, performing postoperative readmission prediction by adopting an artificial neural network model, dividing the patient into a readmission high risk group and a readmission non-high risk group, and obtaining postoperative readmission risk;
By executing the operation, the technical problems that in the existing management process after the weight-reducing operation, the robustness of the existing readmission prediction model is low, the actual prediction result is unreliable, and the practicability of the management system after the weight-reducing operation is further affected are solved; the scheme creatively adopts the Bo Lu Da algorithm combined with the limit gradient lifting model to perform feature selection, further performs postoperative readmission prediction, reduces the randomness of the traditional limit gradient lifting model by introducing the shadow feature matrix and the Z score, has the characteristics of high precision, high robustness and low randomness, improves the practicability of the weight-reduction postoperative management system, and further is beneficial to medical staff to efficiently formulate a management strategy according to the condition of a patient and promotes postoperative recovery of the patient.
In a sixth embodiment, referring to fig. 1, the embodiment is based on the foregoing embodiment, and in step S5, the post-operation management policy is generated, so as to generate the post-operation management policy according to the predicted weight of the patient and the risk of re-hospitalization after the operation, specifically, to formulate a health management plan based on the predicted weight of the patient, where the health management plan includes a diet, a exercise and a medication management plan, and based on the risk of re-hospitalization after the operation, to enhance the post-operation rehabilitation monitoring and increase the frequency of re-diagnosis for the patient with high risk of re-hospitalization.
Embodiment seven, referring to fig. 2, based on the above embodiment, the weight-reduction postoperative management system based on artificial intelligence provided by the present invention includes: the device comprises a patient data acquisition module, a data preprocessing module, a patient weight prediction module, a postoperative readmission prediction module and a postoperative management strategy generation module;
the patient data acquisition module is used for acquiring patient data, specifically acquiring related data after a weight-reduction operation to obtain patient clinical data, and sending the patient clinical data to the data preprocessing module;
The data preprocessing module is used for preprocessing data, specifically, clinical initial data is obtained by preprocessing clinical data of a patient, and the clinical initial data is sent to the patient weight prediction module and the postoperative hospital readmission prediction module;
The patient weight prediction module is used for predicting the weight of the patient after 3 months of the weight reduction operation, specifically, the clinical initial data is subjected to feature screening through a Pearson correlation coefficient to obtain weight prediction features, an Elman circulating neural network is adopted to construct a patient weight prediction model, a patient weight prediction model is obtained, the patient weight prediction model is adopted to conduct weight prediction to obtain the patient predicted weight, and the patient predicted weight is sent to the postoperative hospital-admission prediction module;
The post-operation readmission prediction module is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, specifically, extracting features of the clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, performing feature selection through a Bo Lu Da algorithm combined with a limit gradient lifting model to obtain a readmission key feature matrix, performing post-operation readmission prediction according to the predicted weight of the patient and the readmission key feature matrix through an artificial neural network model to obtain post-operation readmission risk, and sending the post-operation readmission risk to the post-operation management strategy generation module;
The post-operation management strategy generation module is used for generating a post-operation management strategy, and particularly for making a health management plan based on predicted weight of a patient, wherein the health management plan comprises diet, exercise and medicine management plans, and based on post-operation readmission risk, post-operation rehabilitation monitoring is enhanced and review frequency is increased for patients with high readmission risk.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
The invention and its embodiments have been described above with no limitation, and the actual construction is not limited to the embodiments of the invention as shown in the drawings. In summary, if one of ordinary skill in the art is informed by this disclosure, a structural manner and an embodiment similar to the technical solution should not be creatively devised without departing from the gist of the present invention.
Claims (3)
1. A weight-reduction operation postoperative management method based on artificial intelligence is characterized in that: the method comprises the following steps:
step S1: patient data acquisition, in particular to acquisition of patient clinical data;
Step S2: the data preprocessing comprises the steps of carrying out outlier removal, denoising, missing value filling and normalization on clinical data of a patient to obtain clinical initial data;
step S3: predicting the weight of a patient;
Step S4: prediction of hospital readmission after operation;
step S5: generating a post-operation management strategy;
in step S3, the patient weight prediction is used for predicting the weight of the patient after the weight-loss operation, and includes the following steps:
Step S31: the method comprises the steps of adopting pearson correlation coefficients to perform feature screening on clinical initial data, specifically calculating pearson correlation coefficients of each variable in the clinical initial data and postoperative weight, and selecting k variables with highest pearson correlation coefficients to obtain weight prediction features;
step S32: the patient weight prediction model construction, in particular to a patient weight prediction model construction by adopting an Elman circulating neural network, comprises the following steps:
Step S321: building an Elman circulating neural network structure, wherein the Elman circulating neural network structure comprises an input layer, a hidden layer, a context layer and an output layer;
the input layer receives the weight prediction features as input features, and sets the number of nodes of the input layer according to the dimension of the weight prediction features, so that the number of nodes of the input layer is equal to the dimension of the weight prediction features;
the hidden layer sets a Sigmoid activation function as a hidden layer activation function;
the output layer sets a linear activation function as an output layer activation function;
The context layer is used for memorizing the output value of the hidden layer at the previous moment and feeding back the output value of the hidden layer to the hidden layer again;
Step S322: calculating initial output of the hidden layer by adopting a Sigmoid activation function, taking the initial output of the hidden layer as input of a context layer, feeding back the initial output of the hidden layer to the hidden layer through the context layer, and calculating to obtain final output of the hidden layer, wherein a calculation formula is as follows:
hlout(t)＝o(Wt·y(t)+Wt-1·hlout(t-1)+g1)；
wherein t is a time step index, h out2 (t) is a final output of the hidden layer, the final output of the hidden layer is used for representing the hidden layer output at the current moment, sigma (·) is a Sigmoid activation function, W t is a hidden layer weight matrix at the current moment, y (t) is an input feature at the current moment, W t-1 is a weight matrix of the hidden layer at the previous moment, hl out1 (t-1) is an initial output of the hidden layer, the initial output of the hidden layer is used for representing the hidden layer output at the previous moment, and g 1 is a hidden layer bias term;
step S323: calculating the output of the output layer, which is used for calculating the output result of the model, wherein the calculation formula is as follows:
olout(t)＝ε(W1t·hlout(t)+g2)；
Wherein, ol out (t) is the output of the output layer at the current moment, epsilon (·) is a linear activation function, W1t is the output layer weight matrix at the current moment, and g 2 is the output layer bias term;
Step S324: calculating a loss function of the model, and updating model parameters by back-propagating the loss function;
Step S325: performing model training by repeatedly executing the steps S322, S323 and S324 for N times by adopting a gradient descent algorithm with momentum to obtain a patient weight prediction model;
Step S33: predicting the weight by adopting a patient weight prediction model to obtain the predicted weight of the patient;
in step S4, the post-operation readmission prediction is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, and includes the following steps:
Step S41: extracting features of clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, wherein the one-dimensional convolutional neural network comprises a one-dimensional convolutional layer, two maximum pooling layers, a flattening layer, a first full-connection layer, a second full-connection layer and a regularization layer;
Step S42: the limiting gradient lifting model is combined with the Bo Lu Da algorithm to obtain a Bo Lu Da algorithm combined with the limiting gradient lifting model;
step S43: the feature selection is carried out by adopting a Bo Lu Da algorithm combined with a limit gradient lifting model, and the method comprises the following steps:
Step S431: constructing an input feature matrix, namely randomly arranging the readmission feature matrix to obtain a shadow feature matrix, and splicing the readmission feature matrix and the shadow feature matrix to obtain the input feature matrix;
Step S432: training a limit gradient lifting model, specifically, training the model by minimizing a loss function and an objective function according to a gradient lifting principle, comprising the following steps:
Step S4321: calculating an objective function, which is used for determining an optimization target of the model, wherein a calculation formula is as follows:
Where K is the decision tree index, mf (K) is the objective function, j is the sample index, the samples are used to represent the samples in the input feature matrix, M is the number of samples, the number of samples are used to represent the number of samples in the input feature matrix, Is the loss function, x j is the actual value of the j-th sample,/>Is the predicted value of the jth sample, ψ (f K) is the regularization term for controlling the complexity of the model, f K is the predicted value of the kth decision tree;
step S4322: and calculating a loss function, wherein the loss function is used for measuring the difference between a model predicted value and an actual value, and the calculation formula is as follows:
step S4323: the performance of the K decision tree is evaluated by adopting a second-order Taylor expansion of the objective function, the performance is used for minimizing the objective function, and the calculation formula is as follows:
In the method, in the process of the invention, Is the second order Taylor expansion of the objective function, i is the leaf node index, lno is the number of leaf nodes, FD i is the first derivative of the loss function for the ith leaf node, SD i is the second derivative of the loss function for the ith leaf node, α is the regularization weight, β is the model hyper-parameter, for controlling the number of leaf nodes;
Step S433: and calculating a gain value of the leaf node for evaluating the importance of the feature, wherein the calculation formula is as follows:
Where leaf g is the gain value of the leaf node, FD L is the first derivative of the left node, SD L is the second derivative of the left node, FD R is the first derivative of the right node, SD R is the second derivative of the right node;
step S434: judging the importance of each feature in the readmission feature matrix according to the Z score, and removing the readmission feature from the readmission feature matrix if the Z score of the readmission feature is lower than the maximum Z score of the shadow feature, wherein the calculation formula of the Z score is as follows:
wherein Z (K) is the Z fraction, Is the average of the gain values of the leaf nodes,/>Is the standard deviation of the gain values of the leaf nodes;
Step S435: repeatedly executing the steps S431, S432, S433 and S434 for K max times to obtain a readmission critical feature matrix, wherein the readmission critical feature matrix is used for representing the readmission feature matrix after feature selection;
Step S44: according to the predicted weight of the patient and the readmission critical feature matrix, performing postoperative readmission prediction by adopting an artificial neural network model, dividing the patient into a readmission high risk group and a readmission non-high risk group, and obtaining postoperative readmission risk;
in step S5, the post-operation management policy is generated, and the post-operation management policy is generated according to the predicted weight of the patient and the risk of re-hospitalization after the operation, specifically, a health management plan is formulated based on the predicted weight of the patient, the health management plan includes diet, exercise and drug management plans, and the post-operation rehabilitation monitoring is enhanced and the frequency of re-diagnosis is increased for the patient with high risk of re-hospitalization based on the risk of re-hospitalization after the operation.
2. The artificial intelligence based weight loss surgical post-operation management method according to claim 1, wherein: in step S1, the patient clinical data includes patient basic information including sex, age, body mass index, specifically BMI index, a visit record describing a post-operation body weight and a reason for readmission and readmission within 3 months after the operation of the patient, a medical history, operation information, and a follow-up record, specifically body weight after 3 months after the operation.
3. An artificial intelligence based weight reduction postoperative management system for implementing the artificial intelligence based weight reduction postoperative management method according to any one of claims 1-2, wherein: the system comprises a patient data acquisition module, a data preprocessing module, a patient weight prediction module, a postoperative readmission prediction module and a postoperative management strategy generation module;
the patient data acquisition module is used for acquiring patient data, specifically acquiring related data after a weight-reduction operation to obtain patient clinical data, and sending the patient clinical data to the data preprocessing module;
The data preprocessing module is used for preprocessing data, specifically, clinical initial data is obtained by preprocessing clinical data of a patient, and the clinical initial data is sent to the patient weight prediction module and the postoperative hospital readmission prediction module;
The patient weight prediction module is used for predicting the weight of the patient after 3 months of the weight reduction operation, specifically, the clinical initial data is subjected to feature screening through a Pearson correlation coefficient to obtain weight prediction features, an Elman circulating neural network is adopted to construct a patient weight prediction model, a patient weight prediction model is obtained, the patient weight prediction model is adopted to conduct weight prediction to obtain the patient predicted weight, and the patient predicted weight is sent to the postoperative hospital-admission prediction module;
The post-operation readmission prediction module is used for performing post-operation readmission prediction according to the predicted weight of the patient and clinical initial data, specifically, extracting features of the clinical initial data through a one-dimensional convolutional neural network to obtain a readmission feature matrix, performing feature selection through a Bo Lu Da algorithm combined with a limit gradient lifting model to obtain a readmission key feature matrix, performing post-operation readmission prediction according to the predicted weight of the patient and the readmission key feature matrix through an artificial neural network model to obtain post-operation readmission risk, and sending the post-operation readmission risk to the post-operation management strategy generation module;
The post-operation management strategy generation module is used for generating a post-operation management strategy, and particularly for making a health management plan based on predicted weight of a patient, wherein the health management plan comprises diet, exercise and medicine management plans, and based on post-operation readmission risk, post-operation rehabilitation monitoring is enhanced and review frequency is increased for patients with high readmission risk.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410217422.8A CN117809811B (en) | 2024-02-28 | 2024-02-28 | Artificial intelligence-based weight-reduction operation postoperative management method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410217422.8A CN117809811B (en) | 2024-02-28 | 2024-02-28 | Artificial intelligence-based weight-reduction operation postoperative management method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117809811A CN117809811A (en) | 2024-04-02 |
| CN117809811B true CN117809811B (en) | 2024-05-10 |
Family
ID=90434794
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410217422.8A Active CN117809811B (en) | 2024-02-28 | 2024-02-28 | Artificial intelligence-based weight-reduction operation postoperative management method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117809811B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101001167B1 (en) * | 2010-04-26 | 2010-12-15 | 주식회사제론헬스케어 | Personal health record system connected with a specialized clinic for managing obesity, chronic diseases |
| EP2719291A1 (en) * | 2012-10-09 | 2014-04-16 | International Nutrition Research Company | Dietetic product for long-term administration to patients after bariatric surgery |
| CN107785062A (en) * | 2017-10-31 | 2018-03-09 | 南京鼓楼医院 | It is multidisciplinary to intervene fat health control method |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| CN112086195A (en) * | 2020-09-16 | 2020-12-15 | 电子科技大学 | Admission risk prediction method based on self-adaptive ensemble learning model |
| CN114822806A (en) * | 2022-04-22 | 2022-07-29 | 昆明市第一人民医院 | Postoperative tracking diagnosis and treatment system for weight-reducing metabolic surgery |
| WO2022226227A1 (en) * | 2021-04-21 | 2022-10-27 | Cedars-Sinai Medical Center | Deep neural networks for predicting post-operative outcomes |
| CN116344052A (en) * | 2023-03-28 | 2023-06-27 | 山西医科大学 | Readmission prediction model training method, readmission prediction device and readmission prediction equipment |
| CN117043881A (en) * | 2021-01-22 | 2023-11-10 | 西拉格国际有限公司 | Weight loss surgical postoperative monitoring |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230068453A1 (en) * | 2021-08-25 | 2023-03-02 | Koninklijke Philips N.V. | Methods and systems for determining and displaying dynamic patient readmission risk and intervention recommendation |
-
2024
- 2024-02-28 CN CN202410217422.8A patent/CN117809811B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101001167B1 (en) * | 2010-04-26 | 2010-12-15 | 주식회사제론헬스케어 | Personal health record system connected with a specialized clinic for managing obesity, chronic diseases |
| EP2719291A1 (en) * | 2012-10-09 | 2014-04-16 | International Nutrition Research Company | Dietetic product for long-term administration to patients after bariatric surgery |
| CN107785062A (en) * | 2017-10-31 | 2018-03-09 | 南京鼓楼医院 | It is multidisciplinary to intervene fat health control method |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| CN112086195A (en) * | 2020-09-16 | 2020-12-15 | 电子科技大学 | Admission risk prediction method based on self-adaptive ensemble learning model |
| CN117043881A (en) * | 2021-01-22 | 2023-11-10 | 西拉格国际有限公司 | Weight loss surgical postoperative monitoring |
| WO2022226227A1 (en) * | 2021-04-21 | 2022-10-27 | Cedars-Sinai Medical Center | Deep neural networks for predicting post-operative outcomes |
| CN114822806A (en) * | 2022-04-22 | 2022-07-29 | 昆明市第一人民医院 | Postoperative tracking diagnosis and treatment system for weight-reducing metabolic surgery |
| CN116344052A (en) * | 2023-03-28 | 2023-06-27 | 山西医科大学 | Readmission prediction model training method, readmission prediction device and readmission prediction equipment |
Non-Patent Citations (4)
| Title |
|---|
| Predicting hospital readmission in patients with mental or substance use disorders: A machine learning approach;Didier Morel et al;《International Journal of Medical Informatics》;20200731;全文 * |
| Risk for Hospital Readmission following Bariatric Surgery;Robert B. Dorman et al;《PLoS ONE》;20120331;全文 * |
| Using the Super Learner algorithm to predict risk of 30-day readmission after bariatric surgery in the United States;Matteo Torquati et al;《Surgery》;20210731;全文 * |
| 减重术后的管理策略;丁锐;姚琪远;;外科理论与实践;20171125(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117809811A (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ganji et al. | A fuzzy classification system based on Ant Colony Optimization for diabetes disease diagnosis | |
| Pradhan et al. | Predict the onset of diabetes disease using Artificial Neural Network (ANN) | |
| Verma et al. | Temporal deep learning architecture for prediction of COVID-19 cases in India | |
| CN112086195B (en) | Admission risk prediction method based on self-adaptive ensemble learning model | |
| CN113592809A (en) | Pneumonia image detection system and method based on channel attention residual error network | |
| CN116189847B (en) | Safety medicine recommendation method based on LSTM-CNN strategy of attention mechanism | |
| CN111387938B (en) | Patient heart failure death risk prediction system based on characteristic rearrangement one-dimensional convolutional neural network | |
| Kumar et al. | Predictive classification of covid-19: Assessing the impact of digital technologies | |
| CN115985513B (en) | Data processing method, device and equipment based on multiple groups of chemical cancer typing | |
| CN110349668A (en) | A kind of therapeutic scheme aid decision-making method and its system based on BP neural network | |
| Chadaga et al. | Predicting cervical cancer biopsy results using demographic and epidemiological parameters: a custom stacked ensemble machine learning approach | |
| CN111798324A (en) | Medical insurance fraud discovery method based on dynamic hospitalizing behavior alignment | |
| Azam et al. | Classification of COVID-19 symptoms using multilayer perceptron | |
| CN116072298B (en) | Disease prediction system based on hierarchical marker distribution learning | |
| CN117809811B (en) | Artificial intelligence-based weight-reduction operation postoperative management method and system | |
| CN117393153A (en) | Shock real-time risk early warning and monitoring method and system based on medical internet of things time sequence data and deep learning algorithm | |
| Mehrankia et al. | Prediction of heart attacks using biological signals based on recurrent GMDH neural network | |
| Sharma et al. | Comparing Data Mining Techniques Used For Heart Disease Prediction | |
| CN114613465A (en) | Cerebral apoplexy disease risk prediction and personalized treatment recommendation method and system | |
| Dey et al. | Deep Greedy Network: a tool for medical diagnosis on exiguous dataset of COVID-19 | |
| Saraswathi et al. | Designing the Prediction Protocol based on Sparrow Search Optimization Algorithm with Vibrational Auto Encoder in E-Healthcare | |
| Nguyen et al. | Enhancing disease prediction on imbalanced metagenomic dataset by cost-sensitive | |
| Saxena et al. | Adaptive Multi-Hop Deep Learning based Drug Recommendation System with Selective Coverage Mechanism | |
| Dutt et al. | A multi-layer feed forward neural network approach for diagnosing diabetes | |
| Kavitha et al. | Application of AI and ML Techniques for Revolutionized Health Care System |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |