Health information management platform for patients with inflammatory bowel diseases

Abstract

The invention discloses a health information management platform for an inflammatory bowel disease patient. The health information management platform comprises the following steps: step 1, user registration and information input; 2, data preprocessing and feature extraction; 3, constructing and training a model; step 4, a real-time prediction and early warning mechanism; according to the scheme, by actively monitoring the health data of the patient and applying the real-time analysis capability of the machine learning model, potential health problems can be recognized in advance, timely early warning is achieved, the risk of acute symptom attack is reduced, and therefore the life quality of the patient is improved, multi-dimensional data analysis is utilized, and the accuracy of the patient is improved. Customized health management suggestions are provided for each patient, individual requirements of the patient are met, the control ability of the patient to the illness state of the patient is enhanced, the treatment effect is improved, the system can provide more basis treatment suggestions for doctors through continuous learning and optimization, clinical decisions are helped to be improved, and the accuracy and effectiveness of overall treatment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information management, and in particular to a health information management platform for patients with inflammatory bowel disease. Background Art

[0002] Inflammatory bowel disease (IBD), including Crohn's disease and ulcerative colitis, is a chronic intestinal disease. Patients often face multiple health challenges, including symptom fluctuations, malnutrition, emotional problems and decreased quality of life. With the development of digital health technology, health information management platforms provide IBD patients with a new way to manage their own health. An effective health information management platform can provide IBD patients with comprehensive self-management tools, promote communication between patients and medical teams, improve the efficiency of disease management, and ultimately improve patients' quality of life. The development and application of this platform not only helps patients' health management, but also provides clinicians with important data support to optimize treatment plans. This platform can integrate data from different sources, including hospital visit records, laboratory test results, diet records and patient self-reported symptom data. Through data analysis, doctors can better understand the changes in patients' conditions and develop personalized treatment plans.

[0003] However, the existing technology still has major deficiencies, such as:

[0004] Existing health management platforms usually rely on patients to actively report symptoms and lack real-time monitoring and prediction capabilities, which makes it difficult for patients to obtain timely warnings and intervention recommendations before serious symptoms occur. Traditional methods often cannot fully utilize patients' multi-dimensional health data, resulting in a lack of personalization in management plans and an inability to meet the unique needs and symptom changes of different patients. Summary of the invention

[0005] The purpose of the present invention is to provide a health information management platform for patients with inflammatory bowel disease to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A health information management platform for patients with inflammatory bowel disease, comprising the following steps:

[0008] Step 1: User registration and information entry;

[0009] Step 2: Data preprocessing and feature extraction;

[0010] Step 3: Model construction and training;

[0011] Step 4: Real-time prediction and early warning mechanism;

[0012] Step 5: Feedback and iteration.

[0013] Preferably, the step 1: user registration and information entry specifically includes:

[0014] User registration and information entry:

[0015] The patient downloads the app and creates a personal account through a registration process. During the registration process, the patient needs to fill in basic information, including medical history, allergy history, family medical history, and authorize the platform to access relevant health data;

[0016] Health Data Collection:

[0017] Through the app, patients enter a daily log of symptoms and record diet, emotional state, and medication use, integrating data from wearable devices and historical medical records;

[0018] Data storage and management:

[0019] The collected data is stored in a cloud database to ensure data security and accessibility.

[0020] Preferably, the step 2: data preprocessing and feature extraction specifically includes:

[0021] Data cleaning:

[0022] Clean the collected data, remove duplicates and missing values, and ensure the completeness and accuracy of the data;

[0023] Feature Engineering:

[0024] Extract key features, including symptom frequency, diet type, mood swings, and physiological parameters, and combine them with medical knowledge to build a feature set to ensure that the model is trained based on variables related to symptom occurrence.

[0025] Preferably, model building and training specifically include:

[0026] Select a machine learning algorithm:

[0027] The random forest algorithm is selected as the basic model, which has high accuracy and interpretability when dealing with classification problems;

[0028] Model training:

[0029] Use historical data sets to train the model, input feature sets and corresponding symptom records to build a predictive model;

[0030] Model optimization:

[0031] Through cross-validation and hyperparameter tuning, the model performance is optimized to ensure that the prediction accuracy reaches more than 80%.

[0032] Preferably, the real-time prediction and early warning mechanism specifically includes:

[0033] Real-time data input:

[0034] The system will receive patients’ health data in real time;

[0035] Symptoms prediction:

[0036] Use the trained machine learning model to analyze real-time data, predict potential symptom changes, and generate early warning information based on the prediction results.

[0037] Warning Notice:

[0038] When the model predicts that a patient may develop acute symptoms, a notification is automatically sent to remind the patient to take preventive measures.

[0039] Preferably, feedback and iteration specifically include:

[0040] User feedback collection:

[0041] After the early warning notification, collect patient feedback on the notification;

[0042] Model iterative update:

[0043] Based on user feedback, machine learning models are regularly updated and optimized to improve prediction accuracy and system adaptability.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. By actively monitoring patients’ health data and applying the real-time analysis capabilities of machine learning models, potential health problems can be identified in advance, timely warnings can be achieved, and the risk of acute symptom onset can be reduced, thereby improving patients’ quality of life;

[0046] 2. Use multi-dimensional data analysis to provide each patient with tailored health management suggestions to meet their individual needs, enhance their ability to control their own condition, and improve the effectiveness of treatment;

[0047] 3. Through continuous learning and optimization, the system can provide doctors with more reliable treatment recommendations, help improve clinical decision-making, and enhance the accuracy and effectiveness of overall treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] See also Figure 1 , the present invention provides a technical solution:

[0051] A health information management platform for patients with inflammatory bowel disease, comprising the following steps:

[0052] Step 1: User registration and information entry;

[0053] Step 2: Data preprocessing and feature extraction;

[0054] Step 3: Model construction and training;

[0055] Step 4: Real-time prediction and early warning mechanism;

[0056] Step 5: Feedback and iteration.

[0057] Step 1: User registration and information entry, including:

[0058] User registration and information entry:

[0059] The patient downloads the app and creates a personal account through a registration process. During the registration process, the patient needs to fill in basic information, including medical history, allergy history, family medical history, and authorize the platform to access relevant health data;

[0060] App download and user registration: Patients download the health management app designed specifically for inflammatory bowel disease (IBD) patients through their smartphone app store. The registration process is simple and intuitive, ensuring that users can quickly create a personal account.

[0061] Fill in basic information: During the registration process, patients need to fill in a series of basic information, including name, age, gender, contact information, etc. Important medical information includes medical history such as the type and course of IBD, allergy history (allergic reactions to food or drugs), family medical history (whether anyone in the family suffers from IBD or related diseases). Patients need to authorize the platform to access relevant health data so that the system can use this information in subsequent analysis;

[0062] Patients input their current symptoms, such as abdominal pain, diarrhea, fatigue, weight changes, etc., through the application every day. The system provides a convenient input interface that allows patients to quickly record the severity and duration of symptoms. Patients can also record their daily diet, including food types and intake, as well as emotional states such as anxiety and depression. This information helps to understand the relationship between symptoms and lifestyle. The application allows patients to record the name, dosage and frequency of medication used to better track the impact of medication on symptoms. The system integrates data from wearable devices, such as heart rate, activity (number of steps) and sleep quality, to establish a comprehensive health profile for patients. Patients can upload historical medical records, including laboratory test results and medical images, to provide the system with a more comprehensive health background. All collected data will be securely stored in a cloud database to ensure data security, integrity and accessibility. Encryption technology is used to protect patients' personal privacy information, and a regular data backup mechanism is established to ensure that data will not be lost under any circumstances. Patients can access and manage their health data at any time;

[0063] Through comprehensive data collection and integration, high-quality training data is provided for subsequent machine learning models, improving the model's prediction accuracy. The collected multi-dimensional data can realize personalized health management, making the management plan for each patient more targeted, thereby improving the patient's self-management ability and treatment compliance. By continuously collecting user feedback and health data, the system can be continuously optimized and updated to form a closed-loop management to ensure the effectiveness and continuous improvement of the early warning system.

[0064] Step 2: Data preprocessing and feature extraction, including:

[0065] Data cleaning:

[0066] Clean the collected data, remove duplicates and missing values, and ensure the completeness and accuracy of the data;

[0067] Feature Engineering:

[0068] Extract key features, including symptom frequency, diet type, mood swings, and physiological parameters, and combine them with medical knowledge to build a feature set to ensure that the model is trained based on variables related to symptom occurrence.

[0069] In the data set, the application automatically detects and deletes duplicate records to ensure that each symptom, diet or mood entry is retained only once to avoid bias. For missing data, appropriate methods are used to fill it, such as using the mean or median to fill it, or interpolating it based on other relevant data. At the same time, the proportion of missing data is recorded to ensure that its impact can be considered during data analysis. During the cleaning process, the integrity of the data is checked to ensure that each patient's record contains key information such as symptoms, diet, emotions, etc., and prompts are given as needed so that patients can supplement the missing information;

[0070] Extract a variety of key features from the cleaned data, including:

[0071] Symptom frequency: Calculate the frequency of a patient's symptoms over a period of time and identify the severity and changing trends of symptoms.

[0072] Diet typing: Categorize the patient's diet records to identify the effects of specific foods on symptoms.

[0073] Mood swings: By analyzing the patient's mood records, the amplitude and frequency of mood swings are extracted, and the relationship between mood and symptoms is explored.

[0074] Physiological parameters: Integrate physiological data collected by wearable devices such as heart rate, activity level, and sleep quality to analyze the relationship between these parameters and the occurrence of symptoms.

[0075] When constructing features, we combine medical expertise to ensure that the extracted features are closely related to the patient's health status and symptom occurrence. This can include analyzing the interactions between features to capture more complex symptom patterns;

[0076] Cleaning and feature extraction make the data input into the machine learning model cleaner and more relevant, thereby improving the model's training effect and prediction accuracy. This will directly affect the timeliness and accuracy of symptom prediction. Through in-depth analysis of features, the system can provide more personalized health management suggestions for each patient, achieve dynamic adjustments, and enhance patients' self-management awareness. Clear feature sets and high-quality data will provide strong support for subsequent data analysis and medical decision-making, helping doctors better understand patients' conditions and optimize treatment plans, thereby improving patients' overall health management level.

[0077] Step 3: Model construction and training, including:

[0078] Select a machine learning algorithm:

[0079] The random forest algorithm is selected as the basic model, which has high accuracy and interpretability when dealing with classification problems;

[0080] Model training:

[0081] Use historical data sets to train the model, input feature sets and corresponding symptom records to build a predictive model;

[0082] Model optimization:

[0083] Through cross-validation and hyperparameter tuning, the model performance is optimized to ensure that the prediction accuracy reaches more than 80%.

[0084] Select Random Forest as the base model. Random Forest is an ensemble learning algorithm that combines the prediction results of multiple decision trees. It can effectively reduce the risk of overfitting. When dealing with classification problems, Random Forest can usually provide higher accuracy, especially when the data dimension is high and the sample size is large: Random Forest provides feature importance evaluation, which can help us understand which features have the greatest impact on symptom prediction, helping doctors and patients better understand the output of the model;

[0085] The cleaned and feature-extracted dataset is divided into a training set and a test set. Usually 70% is used as a training set and 30% as a test set to ensure the effectiveness of model training and evaluation. The extracted feature set and corresponding symptom records are used to train the model. Random forests will form the final prediction result by building multiple decision trees and voting on their prediction results. During the training process, the model will continuously adjust its internal parameters to learn the relationship between features and symptoms.

[0086] Use k-fold cross validation, such as 5-fold or 10-fold, to evaluate the performance of the model on different data subsets to ensure the stability and generalization ability of the model, adjust the model's hyperparameters (such as the number of trees, maximum depth, etc.), and use grid search (GridSearch) or random search (Random Search) methods to find the best parameter combination to improve model performance. The goal is to ensure that the model's prediction accuracy reaches more than 80%. By evaluating the test set, observe the performance of the model on unseen data to ensure that it has good generalization ability;

[0087] Through precise model training and optimization processes, a model that can reliably predict symptoms is built. This is the core of the entire health management program, which directly affects the patient's health warning and personalized management. Accurate prediction results provide data support for doctors, helping them make more reasonable clinical decisions and improve the effectiveness of patient management. Once the model training is completed and put into use, the system can monitor the patient's health changes in real time and dynamically adjust the health management strategy according to the model prediction results to form a closed-loop management system.

[0088] Step 4: Real-time prediction and early warning mechanism, including:

[0089] Real-time data input:

[0090] The system will receive patients’ health data in real time;

[0091] Symptoms prediction:

[0092] Use the trained machine learning model to analyze real-time data, predict potential symptom changes, and generate early warning information based on the prediction results.

[0093] Warning Notice:

[0094] When the model predicts that a patient may develop acute symptoms, a notification is automatically sent to remind the patient to take preventive measures.

[0095] The system receives the patient's health data in real time through applications or wearable devices, including symptom records, diet, emotional state, physiological parameters (such as heart rate, activity level, and sleep quality), etc. The data will be automatically uploaded to the cloud in a scheduled or event-driven manner to ensure that the data is updated in real time and provide the latest health status. When inputting data, the system will implement data format standardization to ensure that data from different sources can be processed uniformly for subsequent analysis.

[0096] Perform necessary preprocessing on the data received in real time, including missing value processing and outlier detection, to ensure the quality and applicability of the data, and use the previously trained random forest model to analyze the real-time data. The model will input real-time features to generate predictions of potential symptom changes of patients. Based on the output of the model, the system will generate possible symptom changes (for example, the probability of symptom aggravation) and compare them with historical data to improve the accuracy of the prediction;

[0097] The system sets warning thresholds for different symptoms based on clinical experience and historical data. When the model predicts that the probability of a certain symptom exceeds the set threshold, the warning mechanism is automatically triggered. Once the model predicts that the patient may experience acute symptoms, the system will automatically generate and send a notification. Notifications can be sent in a variety of ways, such as push notifications within mobile apps, text messages, or emails to ensure that patients can receive information in a timely manner. In the notification, the system will provide specific preventive measures and suggestions to help patients take timely actions to avoid worsening symptoms (such as adjusting diet, increasing rest, etc.).

[0098] Through the real-time prediction and early warning mechanism, patients can receive timely warnings before symptoms worsen, so that they can take effective preventive measures, reduce the risk of acute attacks, and ensure patient safety. The system's intelligent notifications not only remind patients to pay attention to changes in symptoms, but also provide specific response strategies to enhance patients' participation in their own health management and improve their self-management capabilities. Timely warnings and interventions can reduce the need for emergency medical treatment caused by acute symptoms, thereby optimizing the utilization of medical resources, reducing the burden on the medical system, and improving overall medical efficiency.

[0099] Step 5: Feedback and iteration, including:

[0100] User feedback collection:

[0101] After the early warning notification, collect patient feedback on the notification;

[0102] Model iterative update:

[0103] Based on user feedback, machine learning models are regularly updated and optimized to improve prediction accuracy and system adaptability.

[0104] Systematically store feedback data to facilitate subsequent analysis and model updates. Ensure that feedback information includes basic patient information (such as age, gender, and medical history) and specific feedback content to promote in-depth understanding of different patient groups, regularly review collected user feedback, analyze its impact on model prediction results, identify common opinions and suggestions, and provide specific directions for model optimization;

[0105] Combine user feedback with historical data to analyze patients’ responses to warnings, identify the deficiencies of the model in specific situations, and iterate and update the model based on feedback data:

[0106] Feature reselection: Adjust input features, add or remove certain features to improve the accuracy of the model.

[0107] Algorithm adjustment: Based on the feedback results, try different algorithms or model combinations (such as XGBoost, deep learning, etc.) to find the optimal solution.

[0108] Hyperparameter re-tuning: Fine-tune the hyperparameters based on the new dataset to adapt to the new input data characteristics and changes.

[0109] After the update, the model is evaluated to ensure that the updated model has improved in accuracy, recall rate and other relevant indicators, and cross-validation is performed to ensure the generalization ability of the new model. By continuously optimizing the model and system, patients experience more accurate and timely services, which enhances their trust in the system, thereby increasing the frequency and effectiveness of use. The feedback and iteration mechanism ensures that the model can adapt to the individual differences and characteristics of different patients, improves the effectiveness of early warning, and reduces false alarms and missed reports, thereby achieving more accurate health management.

[0110] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A health information management platform for patients with inflammatory bowel disease, characterized by: The steps include: Step 1: User registration and information entry; Step 2: Data preprocessing and feature extraction; Step 3: Model construction and training; Step 4: Real-time prediction and early warning mechanism; Step 5: Feedback and iteration.

2. A health information management platform for patients with inflammatory bowel disease according to claim 1, characterized in that: The step 1: user registration and information entry, specifically includes: User registration and information entry: The patient downloads the app and creates a personal account through a registration process. During the registration process, the patient needs to fill in basic information, including medical history, allergy history, family medical history, and authorize the platform to access relevant health data; Health Data Collection: Through the app, patients enter a daily log of symptoms and record diet, emotional state, and medication use, integrating data from wearable devices and historical medical records; Data storage and management: The collected data is stored in a cloud database to ensure data security and accessibility.

3. A health information management platform for patients with inflammatory bowel disease according to claim 2, characterized in that: The step 2: data preprocessing and feature extraction, specifically includes: Data cleaning: Clean the collected data, remove duplicates and missing values, and ensure the completeness and accuracy of the data; Feature Engineering: Extract key features, including symptom frequency, diet type, mood swings, and physiological parameters, and combine them with medical knowledge to build a feature set to ensure that the model is trained based on variables related to symptom occurrence.

4. A health information management platform for patients with inflammatory bowel disease according to claim 3, characterized in that: The step three: model construction and training, specifically includes: Select a machine learning algorithm: The random forest algorithm is selected as the basic model, which has high accuracy and interpretability when dealing with classification problems; Model training: Use historical data sets to train the model, input feature sets and corresponding symptom records to build a predictive model; Model optimization: Through cross-validation and hyperparameter tuning, the model performance is optimized to ensure that the prediction accuracy reaches more than 80%.

5. A health information management platform for patients with inflammatory bowel disease according to claim 1, characterized in that: The fourth step: real-time prediction and early warning mechanism, specifically includes: Real-time data input: The system will receive patients’ health data in real time; Symptoms prediction: Use the trained machine learning model to analyze real-time data, predict potential symptom changes, and generate early warning information based on the prediction results. Warning Notice: When the model predicts that a patient may develop acute symptoms, a notification is automatically sent to remind the patient to take preventive measures.

6. A health information management platform for patients with inflammatory bowel disease according to claim 1, characterized in that: The step five: feedback and iteration, specifically includes: User feedback collection: After the early warning notification, collect patient feedback on the notification; Model iterative update: Based on user feedback, machine learning models are regularly updated and optimized to improve prediction accuracy and system adaptability.

Images