Neural network-based postoperative diet intake nutrition evaluation system for weight loss surgery
By constructing a neural network-based post-bariatric surgery dietary intake nutritional assessment system, the problem of lack of data support in traditional nutritional assessment methods has been solved, enabling dynamic monitoring and optimization of patient dietary management and improving the accuracy and efficiency of nutritional assessment and management.
Patent Information
- Application Number
- CN202510943816.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-09
- Publication Date
- 2025-10-24
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional nutritional assessment methods lack objective data support, cannot track patients' dietary intake in real time, and are difficult to achieve dynamic monitoring and optimization of patients' dietary management.
A neural network-based postoperative dietary intake nutritional assessment system for bariatric surgery was constructed, including modules for data acquisition, preprocessing, neural network model, nutritional assessment, and dietary recommendation generation. The model was trained using the backpropagation algorithm, and personalized dietary recommendations were generated based on the individual needs of the patients.
It realizes dynamic monitoring and optimization of patient diet management, improves the accuracy and reliability of nutritional assessment, ensures that the diet plan meets the actual needs of patients, and enhances doctor-patient interaction and management efficiency.
Smart Images
Figure CN120833890A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of medical health management, more particularly to a weight loss surgery postoperative diet intake nutrition assessment system based on neural networks. BACKGROUND
[0002] Weight loss surgery is one of the effective means to treat severe obesity, but postoperative diet management of patients is crucial to the effectiveness of the surgery and physical health recovery. Reasonable diet intake not only ensures that patients obtain sufficient nutritional support and promote physical recovery, but also effectively prevents postoperative complications and maintains a healthy weight.
[0003] However, traditional nutrition assessment methods rely on the experience of medical staff and the subjective description of patients, lack objective and accurate data support, cannot track the diet intake of patients in real time, cannot discover diet problems in time and give targeted guidance and adjustment, and are difficult to achieve dynamic monitoring and optimization of patient diet management. In view of this, we propose a weight loss surgery postoperative diet intake nutrition assessment system based on neural networks. SUMMARY
[0004] The purpose of the present application is to provide a weight loss surgery postoperative diet intake nutrition assessment system based on neural networks, aiming to solve the problem that the prior art is difficult to achieve dynamic monitoring and optimization of patient diet management.
[0005] To solve the above technical problems, the present application provides the following technical scheme: a weight loss surgery postoperative diet intake nutrition assessment system based on neural networks, which comprises a data acquisition module, a data preprocessing module, a neural network model module, a nutrition assessment module, a diet suggestion generation module and a user interaction module; The data acquisition module is used to acquire basic information data, physiological index data and diet category and intake amount information data of patients; The data processing module cleans and removes outliers from the collected data, and then converts the diet category and intake amount information into nutrition intake amount data; The neural network model module is used to construct a neural network, and a training set is composed of historical postoperative data of patients, the data is labeled according to the evaluation results of professional doctors, and the neural network model is trained by using a back propagation algorithm to adjust the weights and biases, so that the model can more accurately assess the diet intake nutrition of patients; The nutrition assessment module inputs the preprocessed data of the current patient into the trained neural network, and outputs a nutrition assessment report; The diet suggestion generation module generates individualized diet suggestions and phased diet adjustment plans according to the results of the nutrition assessment module, combined with the diet preferences and taboos of patients; The user interaction module provides an interaction interface for patients and medical staff, supports patients to enter information, view report suggestions and feedback, and facilitates medical staff to track patient diet and adjust system parameters.
[0006] Preferably, the data acquisition module acquires basic information data of the patient through a hospital information system, and the basic information data includes age, gender, weight, height and operation type data of the patient.
[0007] Preferably, the data acquisition module acquires physiological index data of the patient through a hospital examination device, and the physiological index data includes blood glucose, blood lipid, blood pressure and body fat rate data of the patient.
[0008] Preferably, the data acquisition module manually inputs diet type and intake amount information data through a mobile application.
[0009] Preferably, the data preprocessing module removes repeated data, error format data and data obviously deviating from the normal range by a preset data cleaning rule when cleaning data.
[0010] Preferably, the cleaning rule includes: S1, for all data records collected, by comparing the patient's unique identifier, collection time and data content, if there are identical records, only one record is retained; S2, check whether the data conforms to the type and format of the data, and delete the data that does not conform; S3, set a reasonable threshold range for physiological index data and diet intake amount data, and regard the data exceeding the range and without special annotation as abnormal data and delete or mark for manual review.
[0011] Preferably, the neural network model module constructs a multilayer perceptron, the number of layers and the number of neurons in each layer are adjusted according to the complexity and feature quantity of historical postoperative data of patients, and the neural network construction formula is wherein, represents the output of the layer neuron, is an activation function, is the weight of the layer th neuron to the layer th neuron, is the output of the layer th neuron, is the bias of the layer th neuron, is the The number of layer neurons.
[0012] Preferably, the back propagation algorithm adopted by the neural network model module trains the training formula of the neural network model as , wherein, and are the update amount of weight and bias respectively, is the learning rate, is the loss function, which measures the model performance by calculating the error between the predicted value and the true value, and adjusts the weight and bias according to the formula, so that the model can more accurately evaluate the patient's dietary nutrient intake.
[0013] Preferably, the historical postoperative data of the patient is obtained by extracting the basic information, operation record, postoperative physiological index data of the patient who has completed the weight loss surgery from the electronic medical record system of the hospital, and obtaining the dietary data recorded by the patient after the operation through communication with the patient, collecting the examination report data generated by the patient after the operation in the hospital, and integrating these data.
[0014] Compared with the prior art, the beneficial effects of the present application are: 1、The data acquisition module of the present application continuously acquires the basic information, physiological indicators and dietary intake data of the patient, the data preprocessing module cleanses and converts the data, the neural network model module trains the model using historical data to accurately evaluate the current patient's nutrient intake, the nutrition evaluation module outputs a real-time evaluation report, and the dietary suggestion generation module generates a dynamically adjusted dietary plan combining the evaluation results and the personalized needs of the patient, thereby realizing dynamic monitoring and optimization of the patient's dietary management.
[0015] 2、The neural network model module in the present application constructs a multilayer perceptron, adjusts the number of layers and the number of neurons in each layer according to the complexity and the number of features of the historical postoperative data of the patient, and trains the model using the back propagation algorithm, so that the model can more accurately learn and fit the rules and patterns of the postoperative patient's dietary intake after weight loss surgery, thereby improving the accuracy and reliability of the nutrition evaluation.
[0016] 3、The data preprocessing module in the present application removes duplicate data, error format data and abnormal data from the collected data by predefining cleaning rules, converts the dietary category and intake amount information into nutrient intake data, improves the quality and usability of the data, and provides a reliable data basis for subsequent neural network model evaluation, thereby avoiding evaluation deviation caused by inaccurate data. BRIEF DESCRIPTION OF DRAWINGS
[0017] Figure 1 The schematic diagram of the architecture of the system in the present application. DETAILED DESCRIPTION
[0018] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. EMBODIMENT
[0019] A weight loss surgery postoperative diet intake nutrition assessment system based on a neural network, the system comprising a data acquisition module, a data preprocessing module, a neural network model module, a nutrition assessment module, a diet suggestion generation module, and a user interaction module; The data acquisition module is used to acquire basic information data, physiological index data, and diet category and intake amount information data of a patient, to ensure comprehensive patient information acquisition in multiple dimensions and provide a basis for accurate nutrition assessment; The data processing module cleans and removes outliers from the collected data, and then converts the diet category and intake amount information into nutrition intake amount data, to ensure data accuracy and reliability through data cleaning and avoid the influence of outliers on the assessment results; The neural network model module is used to construct a neural network, and a training set is composed of historical postoperative data of patients, the data is labeled according to professional doctor assessment results, and then a back propagation algorithm is used to train the neural network model to adjust the weight and bias, so that the model can more accurately assess the diet intake nutrition of the patient, the model is trained using historical postoperative data of patients to enable the model to learn the relationship between postoperative diet and nutrition intake of weight loss surgery patients, the back propagation algorithm optimizes the model parameters, and the accuracy of nutrition assessment is improved; The nutrition assessment module inputs the preprocessed data of the current patient into the trained neural network, and outputs a nutrition assessment report, to assess the current diet of the patient based on the trained model and generate a detailed nutrition assessment report, providing intuitive nutrition intake situation analysis for the patient and medical staff; The diet suggestion generation module generates personalized diet suggestions and phased diet adjustment plans according to the results of the nutrition assessment module, combined with the diet preferences and taboos of the patient, to ensure that the diet plan meets the actual situation of the patient, improve the patient's compliance, and the phased adjustment plan adapts to the nutritional needs in different recovery stages after surgery; The user interaction module provides an interactive interface for the patient and medical staff, supports the patient to input information, view report suggestions and feedback, and facilitates the medical staff to track the patient's diet and adjust the system parameters, so that the convenient interactive interface facilitates the patient to participate in diet management, the medical staff can track the patient's diet in real time and adjust the system parameters, and the medical staff can enhance the interaction and management efficiency.
[0020] Further, the data acquisition module acquires basic information data of the patient through the hospital information system, and the basic information data includes age, gender, weight, height and operation type data of the patient.
[0021] Further, the data acquisition module acquires physiological index data of the patient through the hospital examination equipment, and the physiological index data includes blood glucose, blood lipid, blood pressure and body fat rate data of the patient.
[0022] Further, the data acquisition module manually inputs diet type and intake amount information data through a mobile application.
[0023] Further, the data preprocessing module removes repeated data, error format data and data obviously deviating from the normal range through the preset data cleaning rule when cleaning the data, so as to effectively eliminate invalid data, improve the data quality, and provide reliable data input for subsequent model analysis.
[0024] Further, the cleaning rule includes: S1, for all data records collected, by comparing the patient's unique identifier, collection time and data content, if there are exactly the same records, only one is retained; S2, check whether the data conforms to the type and format of the data, and delete the data that does not conform; S3, set a reasonable threshold range for physiological index data and diet intake amount data, and regard the data exceeding the range and without special annotation as abnormal data and delete or mark for manual review and processing; Through the multi-dimensional cleaning rule, the data quality is guaranteed from the aspects of data repeatability, format correctness and numerical rationality, and the influence of data error on nutritional assessment is reduced.
[0025] Further, the neural network constructed by the neural network model module is a multi-layer perceptron, the number of layers and the number of neurons in each layer are adjusted according to the complexity and feature quantity of the historical patient postoperative data, and the neural network construction formula is wherein, represents the output of the layer neuron, is an activation function, is the weight of the layer th neuron to the layer th neuron, is the output of the layer th neuron, is the bias of the layer th neuron, is the first layer neural network, the number of layers of neurons, the multi-layer perceptron captures the complex nonlinear relationship between the data, adjusts the network structure according to the data characteristics, and improves the adaptability and accuracy of the model for postoperative dietary nutrition evaluation of weight loss surgery patients.
[0026] Further, the back propagation algorithm adopted by the neural network model module for training the neural network model is , wherein, and are the update amount of weight and bias respectively, is the learning rate, is the loss function, which measures the performance of the model by calculating the error between the predicted value and the true value, and adjusts the weight and bias so that the model can more accurately evaluate the patient's dietary nutrition, so as to minimize the error between the model predicted value and the true value by calculating the gradient of the loss function on the weight and bias, and iteratively optimize the model parameters, so as to improve the accuracy of the nutrition evaluation.
[0027] Further, the historical postoperative data of the patients are obtained by extracting the basic information, operation records, postoperative physiological index data of the patients who have completed weight loss surgery from the electronic medical record system of the hospital, and through communication with the patients, obtaining the dietary data recorded by the patients after the operation using the mobile application, and collecting the examination report data generated by the examination equipment in the hospital, and integrating these data to ensure the comprehensiveness and authenticity of the training set through multi-source data integration.
[0028] The embodiments disclosed by the present application are preferred embodiments, but are not limited thereto, and those skilled in the art can easily understand the spirit of the present application and make different inferences and changes according to the above embodiments, as long as they do not deviate from the spirit of the present application, they are within the protection scope of the present application.
Claims
1. A neural network-based post-bariatric surgery dietary intake nutrient assessment system, comprising: The system comprises a data acquisition module, a data preprocessing module, a neural network model module, a nutrition assessment module, a diet suggestion generation module, and a user interaction module; The data acquisition module is used to acquire basic information data, physiological index data, and diet category and intake information data of the patient; The data processing module cleans and removes outliers from the collected data, and then converts the diet category and intake information into nutrition intake data; The neural network model module is used to construct a neural network, and a training set is formed by collecting historical postoperative data of patients, the data is labeled according to the professional doctor's assessment results, and then a back propagation algorithm is used to train the neural network model to adjust the weight and bias, so that the model can more accurately assess the patient's diet intake nutrition; The nutrition assessment module inputs the preprocessed data of the current patient into the trained neural network, and outputs a nutrition assessment report; The diet suggestion generation module generates personalized diet suggestions and stage diet adjustment plans according to the results of the nutrition assessment module, combined with the patient's dietary preferences and taboos; The user interaction module provides an interactive interface for patients and medical staff, supports patients to enter information, view reports and suggestions, and feedback, and facilitates medical staff to track patient diet and adjust system parameters.
2. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The data acquisition module acquires basic information data of the patient through the hospital information system, and the basic information data includes age, gender, weight, height, and surgery type data of the patient.
3. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The data acquisition module acquires physiological index data of the patient through hospital examination equipment, and the physiological index data includes blood glucose, blood lipid, blood pressure, and body fat rate data of the patient.
4. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The data acquisition module manually inputs diet category and intake information data through a mobile application.
5. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, When cleaning data, the data preprocessing module removes duplicate data, incorrect format data, and data that deviates significantly from the normal range through preset data cleaning rules.
6. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 5, wherein, The cleaning rules include: S1, for all data records collected, by comparing the patient's unique identifier, collection time, and data content, if there are identical records, only one is retained; S2, check if the data conforms to the type and format of the data, and delete the data that does not conform; S3, set a reasonable threshold range for physiological index data and diet intake data, and regard the data that exceeds the range and has no special annotation as abnormal data and delete or mark it for manual review.
7. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The neural network model module constructs a multi-layer perceptron, the number of layers and the number of neurons in each layer are adjusted according to the complexity and the number of features of the historical postoperative data of the patients, and the neural network construction formula is wherein, represents the output of the layer neuron, is an activation function, is the weight of the layer the neuron to the layer the neuron, is the output of the layer the neuron, is the bias of the layer the neuron, is the number of neurons in the layer.
8. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The training formula of the neural network model trained by the back propagation algorithm adopted by the neural network model module is , , , , , , , , wherein, and are the update amount of weight and bias respectively, is the learning rate, is the loss function, the model performance is measured by calculating the error between the predicted value and the true value, and the weight and bias are adjusted according to the formula, so that the model can more accurately evaluate the nutritional status of the patient's dietary intake.
9. The neural network-based post-bariatric surgery dietary intake nutrient assessment system of claim 1, wherein, The historical postoperative data of the patient is obtained by extracting the basic information, surgery records, and postoperative physiological index data of the patient who has completed the weight loss surgery from the electronic medical record system of the hospital, and by communicating with the patient to obtain the postoperative diet data recorded by the patient through a mobile application. The diet data recorded by the mobile application is collected, and the examination report data generated by the examination equipment during the patient's postoperative hospitalization is also collected, and these data are integrated to obtain the historical postoperative data of the patient.