Intelligent follow-up system and method after weight loss surgery

Abstract

The present invention relates to an intelligent follow-up system and method after weight loss surgery, which can obtain the weight change data and basic data of historical patients after weight loss surgery; create a standard weight change prediction model and train it; collect the basic data and current weight of the current postoperative patient, output the standard weight change data of the current postoperative patient through the model, and extract the postoperative diet of the historical patient with the highest matching degree as the current patient's diet planning scheme; collect the weight change value of the current patient at the first time node, and if the weight change value of the first time node is within the preset error range, return the information that the current patient reaches the standard at the first time node; collect the weight change value of the current patient at the second time node, judge that the weight change value of the second time node is within the preset error range, and return the information that the current patient reaches the standard at the second time node. Track the weight change of the patient after weight loss surgery to reduce the potential health risks of the patient after weight loss surgery.

Description

Technical Field

[0001] The present invention belongs to the technical field of postoperative management of weight loss surgery, and in particular relates to an intelligent follow-up system and method for postoperative weight loss surgery. Background Art

[0002] Bariatric surgery is an effective way to improve obesity and related metabolic diseases. It changes the patient's digestive tract anatomy and function through surgery, reduces stomach capacity, reduces intestinal absorption area, and changes gastrointestinal hormone regulation, thereby achieving the purpose of weight loss. Surgical methods include sleeve gastrectomy, gastric bypass, single anastomosis gastric bypass, biliopancreatic diversion, duodenal switch, etc.

[0003] After a patient undergoes weight loss surgery, the structure of the digestive tract changes, which poses a greater potential health risk to the patient. Therefore, how to track the patient's condition after weight loss surgery and conduct follow-up and management based on the tracking results to reduce the potential health risks of patients after weight loss surgery is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent follow-up system and method for weight loss surgery, which is used to track the patient's condition after weight loss surgery, and to follow up and manage according to the tracking results, so as to reduce the potential health risks of patients after weight loss surgery.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0006] In a first aspect, a method for intelligent follow-up after weight loss surgery comprises the following steps:

[0007] S1: Based on big data, the weight change data and basic data of historical patients after weight loss surgery are obtained, and the data of patients whose weight changes after weight loss surgery meet the preset requirements are extracted, and the historical patient data set is constructed after preprocessing. The basic data includes age, gender, and postoperative diet;

[0008] S2: creating a standard weight change prediction model, and inputting the historical patient data set into the standard weight change prediction model to train the model, and evaluating the prediction ability of the trained model to determine whether the evaluation result meets the preset requirements, if so, executing step S3, if not, adjusting the parameters of the standard weight change prediction model;

[0009] S3: collecting the basic data and current weight of the current postoperative patient and inputting them into the standard weight change prediction model, outputting the standard weight change data of the current postoperative patient through the model, extracting the postoperative diet of the historical patient with the highest matching degree with the basic data of the current patient in the historical patient data set as the current patient diet plan, and storing the standard weight change data and diet plan of the current patient in the database of the follow-up system;

[0010] S4: The weight value of the current patient at the first time node is collected and input into the weight recording module of the follow-up system, which transmits it to the data analysis module. The data analysis module compares the weight change value at the first time node with the weight change value at the first time node of the standard weight change data, and determines whether the weight change value at the first time node is within a preset error range. If so, information that the current patient has reached the standard at the first time node is returned, and step S5 is executed. If not, corresponding information that the standard is not reached is returned to the doctor, and the patient's diet is adjusted and the interval period between adjacent time nodes is shortened, and step S5 is executed;

[0011] S5: collecting the weight value of the current patient at the second time node and inputting it into the weight recording module of the follow-up system, which transmits it to the data analysis module, which compares the weight change value at the second time node with the weight change value at the second time node of the standard weight change data, and determines whether the weight change value at the second time node is within a preset error range. If so, returning information that the current patient has reached the standard at the second time node, and if not, adjusting the patient's diet plan;

[0012] S6: According to step S4, the weight change analysis at subsequent preset time nodes is performed in sequence and the corresponding analysis results, diagnosis and treatment recommendations and diet plans are returned to the current patient.

[0013] Preferably, the standard weight change prediction model in step S2 is a neural network model, and the standard weight change prediction model is configured with an output layer, a hidden layer and an output layer, the output layer is provided with multiple input dimensions for inputting age, gender, postoperative diet, and weight data, the hidden layer is provided with multiple neurons, and the output layer is provided with at least one output dimension for outputting standard weight change data.

[0014] Preferably, the hidden layer n The input of the layer neurons is represented as:

[0015] ;

[0016] The expression of the output result of the output layer is:

[0017] ;

[0018] in, wi Indicates that the lth layer is mapped to the n The weight of the layer, Indicates n The threshold of the layer neuron, function Represents a nonlinear mapping function for the calculation results. xi Indicates n Layer i Input data.

[0019] Preferably, the nonlinear mapping function is the Sigmoid activation function, and its specific expression is:

[0020] ;

[0021] in, x is the input of the Sigmoid activation function, σ is the output of the Sigmoid activation function. The Sigmoid activation function takes the input x Convert to a value between 0 and 1.

[0022] Preferably, the specific training process of the standard weight change prediction model is as follows:

[0023] S21: Initializing the weight and bias of the standard weight change prediction model, inputting the data in the historical patient data set into the input layer of the standard weight change prediction model in sequence, and inputting age, gender, postoperative diet, and weight data through different input dimensions respectively;

[0024] S22: The data of the input layer is passed to the hidden layer after weighted sum operation. The first layer of neurons in the hidden layer receives the signal from the input layer and passes through the nonlinear mapping function. After processing, it is passed to the next layer. The neurons in the next layer receive signals from the neurons in the previous layer and pass through the nonlinear mapping function. After processing, it is passed to the next layer until it is passed to the output layer, and the result is output through the output layer;

[0025] S23: Based on the back propagation algorithm, the weight and bias of each neuron in the hidden layer are iteratively adjusted to continuously reduce the output error value until the number of iterations is met or the error value is reduced to within a preset threshold.

[0026] Preferably, the specific process of the diet planning scheme in step S3 is as follows:

[0027] S31: collecting basic data and current weight of the current postoperative patient, and inputting them into the trained standard weight change prediction model, and the model outputs the standard weight change data of the current postoperative patient;

[0028] S32: matching the first batch of historical patient data whose age difference is within a preset range in the historical patient data set according to the age of the current patient, and extracting the second batch of historical patient data with the same gender as the current patient from the first batch of historical patient data;

[0029] S33: extracting the final historical patient data closest to the preoperative weight data of the current patient from the second batch of historical patient data;

[0030] S34: The postoperative diet of the final historical patient data is used as the current patient diet plan, and the current patient diet plan is stored in the database of the follow-up system, and the corresponding diet plan data is returned to the current patient at each preset time node.

[0031] Preferably, the specific process of adjusting the patient's diet in step S4 is as follows:

[0032] S41: calling the historical patient data set, obtaining the body mass index of each patient, and drawing a weight change curve of each patient after surgery;

[0033] S42: transmitting the body mass index and weight change curve of each patient to the data analysis module, and the data analysis module selects the patient whose body mass index is within a preset range and whose weight change curve has the largest change amplitude among the patients;

[0034] S43: Based on the selected patient, extract the dietary data thereof as the current patient's adjusted dietary planning plan.

[0035] In a second aspect, a postoperative intelligent follow-up system for weight loss surgery is provided, which is used to implement any one of the postoperative intelligent follow-up methods for weight loss surgery, including a data acquisition module, a model creation module, a standard weight change prediction model, a model training module, a weight recording module, a data analysis module, a diet planning module, and a data input module; the model creation module is connected to the standard weight change prediction model, the data acquisition module is connected to the standard weight change prediction model, the standard weight change prediction model is connected to the model training module, the weight recording module is connected to the data analysis module, and the data analysis module is connected to the diet planning module;

[0036] The data acquisition module is used to acquire the weight change data and basic data of historical patients after bariatric surgery based on big data, extract the data of patients whose weight changes after bariatric surgery meet the preset requirements, and construct the historical patient data set after preprocessing;

[0037] The model creation module is used to create a standard weight change prediction model;

[0038] The model training module is used to train the standard weight change prediction model;

[0039] The weight recording module is used to record the weight data of the current patient at each preset time node and transmit it to the data analysis module;

[0040] The diet planning module is used to plan the diet of the current patient;

[0041] The data analysis module is used to compare the weight change value at the specified time node with the weight change value at the corresponding time node of the standard weight change data, determine whether the weight change value at the specified time node is within a preset error range, and return the analysis result of the specified time node to the current patient;

[0042] The data input module is used to input the current patient's weight change data and basic data.

[0043] The beneficial effects of the present invention include:

[0044] The intelligent follow-up system and method for weight loss surgery provided by the present invention obtains the weight change data and basic data of historical patients after weight loss surgery; creates and trains a standard weight change prediction model; collects the basic data and current weight of the current postoperative patient, outputs the standard weight change data of the current postoperative patient through the model, and extracts the postoperative diet of the historical patient with the highest matching degree as the current patient's diet planning program; collects the weight change value of the current patient at the first time node, and the weight change value of the first time node is within the preset error range, and returns the information that the current patient reaches the standard at the first time node; collects the weight change value of the current patient at the second time node, judges that the weight change value of the second time node is within the preset error range, and returns the information that the current patient reaches the standard at the second time node. Tracking the weight change of patients after weight loss surgery reduces the potential health risks of patients after weight loss surgery.

[0045] First, a standard weight change prediction model is created, and the historical patient data set is input into the standard weight change prediction model to train the model. Standard weight change data is formulated based on the historical patient data and the standard weight change prediction model, which is equivalent to formulating with reference to the actual data of a large number of patients. This method is more adaptable than the prior art in which standard weight change data is directly formulated by doctors.

[0046] Secondly, by recording the corresponding weight values ​​according to the preset time nodes and analyzing the weight values ​​through the data analysis module, the weight change values ​​at the same time nodes are compared with the weight change values ​​of the standard weight change data to determine whether the weight change values ​​are within the preset error range, and then analyze whether the weight change in the corresponding time period is normal. If not, the corresponding non-compliance information is returned to the doctor, and the patient's diet is adjusted and the interval period between adjacent time nodes is shortened, thereby realizing the patient's weight tracking. When abnormalities occur, the doctor intervenes to provide diagnosis and treatment recommendations, and further formulates a follow-up plan in conjunction with the adjustment of diet and time nodes to reduce the patient's potential health risks after surgery.

[0047] Thirdly, the output layer, hidden layer and output layer are set through the standard weight change prediction model. The output layer has multiple input dimensions for inputting age, gender, postoperative diet and weight data. The hidden layer has multiple neurons. The output layer has at least one output dimension for outputting standard weight change data. n The input of the layer neurons is represented as: ; The expression of the output result of the output layer is: . It realizes the input of age, gender, postoperative diet and weight data, and outputs standard weight change data. It improves the prediction accuracy of standard weight change data.

[0048] Finally, the historical patient data set is called to obtain the body mass index of each patient, and a weight change curve of each patient after surgery is drawn; the body mass index and weight change curve of each patient are transmitted to the data analysis module, and the data analysis module selects the patient whose body mass index is within a preset range and whose weight change curve has the largest change; based on the selected patient, the diet data of the patient is extracted as the current patient's adjusted diet plan to adjust the patient's diet, so that the diet adjustment plan has higher reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The figure is a schematic diagram of the process of the intelligent follow-up method after weight loss surgery of the present invention.

[0050] Figure 2 This is a schematic diagram of the organizational structure of the intelligent follow-up system for weight loss surgery after the present invention.

[0051] Figure 3 The present invention is a flowchart of a diet planning program for the intelligent follow-up method after weight loss surgery. DETAILED DESCRIPTION

[0052] The following is combined with Figure 1~Figure 3 The present invention is further described in detail:

[0053] Example 1

[0054] See attached Figure 1 As shown, a smart follow-up method after weight loss surgery comprises the following steps:

[0055] S1: Based on big data, the weight change data and basic data of historical patients after weight loss surgery are obtained, and the data of patients whose weight changes after weight loss surgery meet the preset requirements are extracted, and the historical patient data set is constructed after preprocessing. The basic data includes age, gender, and postoperative diet. By obtaining the weight change data and basic data of historical patients after weight loss surgery, and extracting the data of patients whose weight changes after weight loss surgery meet the preset requirements as the data basis for subsequent model training, the subsequent prediction results have stronger reliability and accuracy.

[0056] S2: Create a standard weight change prediction model, and input the historical patient data set into the standard weight change prediction model to train the model, and evaluate the prediction ability of the trained model to determine whether the evaluation result meets the preset requirements. If so, execute step S3, if not, adjust the parameters of the standard weight change prediction model. By creating a standard weight change prediction model and inputting the historical patient data set into the standard weight change prediction model to train the model, standard weight change data is formulated based on historical patient data and the standard weight change prediction model, which is equivalent to formulating with reference to the actual data of a large number of patients. This method has stronger adaptability than the prior art that directly formulates by doctors.

[0057] S3: Collect the basic data and current weight of the current postoperative patient and input them into the standard weight change prediction model, output the standard weight change data of the current postoperative patient through the model, extract the postoperative diet of the historical patient with the highest matching degree with the basic data of the current patient in the historical patient data set as the current patient diet plan, and store the standard weight change data and diet plan of the current patient in the database of the follow-up system. The formulation of standard weight change data and diet plan based on the data of historical patients further improves the reliability and accuracy of the prediction results.

[0058] S4: Collect the weight value of the current patient at the first time node and input it into the weight recording module of the follow-up system. The weight recording module transmits it to the data analysis module. The data analysis module compares the weight change value at the first time node with the weight change value at the first time node of the standard weight change data to determine whether the weight change value at the first time node is within a preset error range. If so, return the information that the current patient meets the standard at the first time node and execute step S5. If not, return the corresponding non-standard information to the doctor, adjust the patient's diet, and shorten the interval period between adjacent time nodes, and execute step S5.

[0059] S5: Collect the weight value of the current patient at the second time node and input it into the weight recording module of the follow-up system. The weight recording module transmits it to the data analysis module. The data analysis module compares the weight change value at the second time node with the weight change value at the second time node of the standard weight change data to determine whether the weight change value at the second time node is within a preset error range. If so, return the information that the current patient has reached the standard at the second time node. If not, adjust the patient's diet plan.

[0060] By recording the corresponding weight values ​​according to the preset time nodes and analyzing the weight values ​​through the data analysis module, the weight change values ​​at the same time nodes are compared with the weight change values ​​of the standard weight change data to determine whether the weight change values ​​are within the preset error range, and then analyze whether the weight change in the corresponding time period is normal. If not, the corresponding non-compliance information is returned to the doctor, and the patient's diet is adjusted and the interval period between adjacent time nodes is shortened, thereby realizing the patient's weight tracking. When abnormalities occur, the doctor intervenes to provide diagnosis and treatment recommendations, and further formulates a follow-up plan in conjunction with the adjustment of diet and time nodes to reduce the patient's potential health risks after surgery.

[0061] S6: According to step S4, the weight change analysis at subsequent preset time nodes is performed in sequence and the corresponding analysis results, diagnosis and treatment recommendations and diet plans are returned to the current patient.

[0062] In another embodiment of the present invention, a standard weight change curve is drawn based on the standard weight change data. The patient can upload his weight data at any time point, and the data analysis module will automatically calculate the time period between the uploaded time point and the previous time point, thereby realizing the weight analysis and management of the patient at any time point after surgery.

[0063] Example 2

[0064] On the basis of Example 1, the standard weight change prediction model in step S2 is a neural network model, and the standard weight change prediction model is provided with an output layer, a hidden layer and an output layer, and the output layer is provided with multiple input dimensions for inputting age, gender, postoperative diet, and weight data, so that the standard weight change prediction model is based on age, gender, postoperative diet, and weight data for prediction, and the final standard weight change has stronger adaptability and has a higher matching degree with the current patient. The hidden layer is provided with multiple neurons, and the output layer is provided with at least one output dimension for outputting standard weight change data.

[0065] Among them, the hidden layer n The input of the layer neurons is represented as:

[0066] ;

[0067] The expression of the output result of the output layer is:

[0068] ;

[0069] in, wi Indicates that the lth layer is mapped to the n The weight of the layer, Indicates n The threshold of the layer neurons, function Represents a nonlinear mapping function for the calculation results. xi Indicates n Layer i The above formula can accurately quantify the standard weight change with high precision.

[0070] In this embodiment, the output layer, hidden layer and output layer are set through the standard weight change prediction model. The output layer has multiple input dimensions for inputting age, gender, postoperative diet, and weight data. The hidden layer has multiple neurons. The output layer has at least one output dimension for outputting standard weight change data. n The input of the layer neurons is represented as: ; The expression of the output result of the output layer is: . It realizes the input of age, gender, postoperative diet and weight data, and outputs standard weight change data. It improves the prediction accuracy of standard weight change data.

[0071] Nonlinear mapping function is the Sigmoid activation function, and its specific expression is:

[0072] ;

[0073] in, xis the input of the Sigmoid activation function, σ is the output of the Sigmoid activation function. The Sigmoid activation function takes the input x Convert to a value between 0 and 1.

[0074] The specific training process of the standard weight change prediction model is as follows:

[0075] S21: Initializing the weight and bias of the standard weight change prediction model, inputting the data in the historical patient data set into the input layer of the standard weight change prediction model in sequence, and inputting age, gender, postoperative diet, and weight data through different input dimensions respectively;

[0076] S22: The data of the input layer is passed to the hidden layer after weighted sum operation. The first layer of neurons in the hidden layer receives the signal from the input layer and passes through the nonlinear mapping function. After processing, it is passed to the next layer. The neurons in the next layer receive signals from the neurons in the previous layer and pass through the nonlinear mapping function. After processing, it is passed to the next layer until it is passed to the output layer, and the result is output through the output layer;

[0077] S23: Based on the back propagation algorithm, the weight and bias of each neuron in the hidden layer are iteratively adjusted to continuously reduce the output error value until the number of iterations is met or the error value is reduced to within a preset threshold.

[0078] Example 3

[0079] Based on Example 1 or Example 2, the specific process of the diet planning scheme in step S3 is as follows:

[0080] S31: collecting basic data and current weight of the current postoperative patient, and inputting them into the trained standard weight change prediction model, and the model outputs the standard weight change data of the current postoperative patient;

[0081] S32: matching the first batch of historical patient data whose age difference is within a preset range in the historical patient data set according to the age of the current patient, and extracting the second batch of historical patient data with the same gender as the current patient from the first batch of historical patient data;

[0082] S33: extracting the final historical patient data closest to the preoperative weight data of the current patient from the second batch of historical patient data;

[0083] S34: The postoperative diet of the final historical patient data is used as the current patient diet plan, and the current patient diet plan is stored in the database of the follow-up system, and the corresponding diet plan data is returned to the current patient at each preset time node.

[0084] In this embodiment, the specific process of adjusting the patient's diet in step S4 is as follows:

[0085] S41: calling the historical patient data set, obtaining the body mass index of each patient, and drawing a weight change curve of each patient after surgery;

[0086] S42: transmitting the body mass index and weight change curve of each patient to the data analysis module, and the data analysis module selects the patient whose body mass index is within a preset range and whose weight change curve has the largest change amplitude among the patients;

[0087] S43: Based on the selected patient, extract the dietary data thereof as the current patient's adjusted dietary planning plan.

[0088] A postoperative intelligent follow-up system for weight loss surgery is used to implement any one of the postoperative intelligent follow-up methods for weight loss surgery, including a data acquisition module, a model creation module, a standard weight change prediction model, a model training module, a weight recording module, a data analysis module, a diet planning module, and a data input module; the model creation module is connected to the standard weight change prediction model, the data acquisition module is connected to the standard weight change prediction model, the standard weight change prediction model is connected to the model training module, the weight recording module is connected to the data analysis module, and the data analysis module is connected to the diet planning module.

[0089] The data acquisition module is used to acquire the weight change data and basic data of historical patients after weight loss surgery based on big data, and extract the data of patients whose weight changes after weight loss surgery meet the preset requirements, and construct the historical patient data set after preprocessing; the model creation module is used to create a standard weight change prediction model; the model training module is used to train the standard weight change prediction model; the weight recording module is used to record the weight data of the current patient at each preset time node and transmit it to the data analysis module; the diet planning module is used to plan the diet of the current patient; the data analysis module is used to compare the weight change value at a specified time node with the weight change value at the corresponding time node of the standard weight change data, determine whether the weight change value at the specified time node is within a preset error range, and return the analysis result of the specified time node to the current patient; the data input module is used to input the weight change data and basic data of the current patient.

[0090] In summary, the intelligent follow-up system and method for weight loss surgery after surgery provided by the present invention creates a standard weight change prediction model, and inputs the historical patient data set into the standard weight change prediction model to train the model, and formulates standard weight change data based on historical patient data and the standard weight change prediction model, which is equivalent to formulating with reference to the actual data of a large number of patients. This method has stronger adaptability than the direct formulation by doctors in the prior art. By recording the corresponding weight value according to the preset time node, and analyzing the weight value through the data analysis module, the weight change value of the same time node is compared with the weight change value of the standard weight change data, and judging whether the weight change value is within the preset error range, and then analyzing whether the weight change in the corresponding time period is normal, if not, returning the corresponding non-standard information to the doctor, adjusting the patient's diet and shortening the interval period between adjacent time nodes, and then realizing the weight tracking of the patient, when abnormalities occur, the doctor intervenes to make diagnosis and treatment suggestions, and further formulates a follow-up plan in conjunction with the adjustment of diet and time nodes to reduce the patient's potential health risks after surgery.

[0091] The output layer, hidden layer and output layer are set by the standard weight change prediction model. The output layer is provided with multiple input dimensions for inputting age, gender, postoperative diet and weight data. The hidden layer is provided with multiple neurons. The output layer is provided with at least one output dimension to output standard weight change data. The standard weight change data is output based on the input of age, gender, postoperative diet and weight data. The prediction accuracy of standard weight change data is improved. The historical patient data set is called to obtain the body mass index of each patient, and the weight change curve of each patient after surgery is drawn; the body mass index and weight change curve of each patient are transmitted to the data analysis module, and the data analysis module selects the patient whose body mass index is within the preset range and whose weight change curve has the largest change; based on the selected patient, the diet data is extracted as the current patient's adjusted diet plan to adjust the patient's diet, so that the diet adjustment plan has higher reliability.

Claims

1. A method for intelligent follow-up after weight loss surgery, characterized in that: The following steps are involved: S1: Based on big data, the weight change data and basic data of historical patients after weight loss surgery are obtained, and the data of patients whose weight changes after weight loss surgery meet the preset requirements are extracted, and the historical patient data set is constructed after preprocessing. The basic data includes age, gender, and postoperative diet; S2: creating a standard weight change prediction model, and inputting the historical patient data set into the standard weight change prediction model to train the model, and evaluating the prediction ability of the trained model to determine whether the evaluation result meets the preset requirements, if so, executing step S3, if not, adjusting the parameters of the standard weight change prediction model; S3: collecting the basic data and current weight of the current postoperative patient and inputting them into the standard weight change prediction model, outputting the standard weight change data of the current postoperative patient through the model, extracting the postoperative diet of the historical patient with the highest matching degree with the basic data of the current patient in the historical patient data set as the current patient diet plan, and storing the standard weight change data and diet plan of the current patient in the database of the follow-up system; S4: The weight value of the current patient at the first time node is collected and input into the weight recording module of the follow-up system, which transmits it to the data analysis module. The data analysis module compares the weight change value at the first time node with the weight change value at the first time node of the standard weight change data, and determines whether the weight change value at the first time node is within a preset error range. If so, information that the current patient has reached the standard at the first time node is returned, and step S5 is executed. If not, corresponding information that the standard is not reached is returned to the doctor, and the patient's diet is adjusted and the interval period between adjacent time nodes is shortened, and step S5 is executed; S5: collecting the weight value of the current patient at the second time node and inputting it into the weight recording module of the follow-up system, which transmits it to the data analysis module, which compares the weight change value at the second time node with the weight change value at the second time node of the standard weight change data, and determines whether the weight change value at the second time node is within a preset error range. If so, returning information that the current patient has reached the standard at the second time node, and if not, adjusting the patient's diet plan; S6: Perform weight change analysis at subsequent preset time nodes in sequence according to step S4 and return corresponding analysis results, diagnosis and treatment suggestions and diet plans to the current patient; The specific process of the diet planning scheme in step S3 is as follows: S31: collecting basic data and current weight of the current postoperative patient, and inputting them into the trained standard weight change prediction model, and the model outputs the standard weight change data of the current postoperative patient; S32: matching the first batch of historical patient data whose age difference is within a preset range in the historical patient data set according to the age of the current patient, and extracting the second batch of historical patient data with the same gender as the current patient from the first batch of historical patient data; S33: extracting the final historical patient data closest to the preoperative weight data of the current patient from the second batch of historical patient data; S34: using the postoperative diet of the final historical patient data as the current patient diet plan, storing the current patient diet plan in a database of the follow-up system, and returning the corresponding diet plan data to the current patient at each preset time node; The specific process of adjusting the patient's diet in step S4 is as follows: S41: calling the historical patient data set, obtaining the body mass index of each patient, and drawing a weight change curve of each patient after surgery; S42: transmitting the body mass index and weight change curve of each patient to the data analysis module, and the data analysis module selects the patient whose body mass index is within a preset range and whose weight change curve has the largest change amplitude among the patients; S43: Based on the selected patient, extract the dietary data thereof as the current patient's adjusted dietary planning plan.

2. The intelligent follow-up method after bariatric surgery according to claim 1, characterized in that: The standard weight change prediction model in step S2 is a neural network model, which is provided with an output layer, a hidden layer and an output layer. The output layer is provided with multiple input dimensions for inputting age, gender, postoperative diet and weight data. The hidden layer is provided with multiple neurons, and the output layer is provided with at least one output dimension for outputting standard weight change data.

3. The intelligent follow-up method after bariatric surgery according to claim 2, characterized in that: The hidden layer n The input of the layer neurons is represented as: ； The expression of the output result of the output layer is: ； in, wi Indicates that the lth layer is mapped to the n The weight of the layer, Indicates n The threshold of the layer neuron, function Represents a nonlinear mapping function for the calculation results. xi Indicates n Layer i Input data.

4. The intelligent follow-up method after bariatric surgery according to claim 3, characterized in that: The nonlinear mapping function is the Sigmoid activation function, and its specific expression is: ； in, x is the input of the Sigmoid activation function, σ is the output of the Sigmoid activation function. The Sigmoid activation function takes the input x Convert to a value between 0 and 1.

5. The intelligent follow-up method after bariatric surgery according to claim 3, characterized in that: The specific training process of the standard weight change prediction model is as follows: S21: Initializing the weight and bias of the standard weight change prediction model, inputting the data in the historical patient data set into the input layer of the standard weight change prediction model in sequence, and inputting age, gender, postoperative diet, and weight data through different input dimensions respectively; S22: The data of the input layer is passed to the hidden layer after weighted sum operation. The first layer of neurons in the hidden layer receives the signal from the input layer and passes through the nonlinear mapping function. After processing, it is passed to the next layer. The neurons in the next layer receive signals from the neurons in the previous layer and pass through the nonlinear mapping function. After processing, it is passed to the next layer until it is passed to the output layer, and the result is output through the output layer; S23: Based on the back propagation algorithm, the weight and bias of each neuron in the hidden layer are iteratively adjusted to continuously reduce the output error value until the number of iterations is met or the error value is reduced to within a preset threshold.

6. An intelligent follow-up system after weight loss surgery, used to implement an intelligent follow-up method after weight loss surgery according to any one of claims 1 to 5, characterized in that: It includes a data acquisition module, a model creation module, a standard weight change prediction model, a model training module, a weight recording module, a data analysis module, a diet planning module, and a data input module; the model creation module is connected to the standard weight change prediction model, the data acquisition module is connected to the standard weight change prediction model, the standard weight change prediction model is connected to the model training module, the weight recording module is connected to the data analysis module, and the data analysis module is connected to the diet planning module; The data acquisition module is used to acquire the weight change data and basic data of historical patients after bariatric surgery based on big data, extract the data of patients whose weight changes after bariatric surgery meet the preset requirements, and construct the historical patient data set after preprocessing; The model creation module is used to create a standard weight change prediction model; The model training module is used to train the standard weight change prediction model; The weight recording module is used to record the weight data of the current patient at each preset time node and transmit it to the data analysis module; The diet planning module is used to plan the diet of the current patient; The data analysis module is used to compare the weight change value at the specified time node with the weight change value at the corresponding time node of the standard weight change data, determine whether the weight change value at the specified time node is within a preset error range, and return the analysis result of the specified time node to the current patient; The data input module is used to input the current patient's weight change data and basic data.

Images