Early warning method and system for intestinal obstruction risk of colorectal cancer postoperative patient

Abstract

The invention relates to the technical field of medical treatment, in particular to a colorectal cancer postoperative patient intestinal obstruction risk early warning method and system, and the method comprises the steps: data collection and arrangement, data preprocessing, feature extraction, selection model construction and training, and real-time early warning. Compared with the prior art which mainly depends on manual monitoring and manual examination and treatment scheme making and has the defects of response lag, insufficient individuation, resource waste and the like, the scheme has the advantages that the recovery condition of the colorectal cancer postoperative patient is monitored in real time, the examination scheme is intelligently generated in combination with the risk score, and the registration process is automatically processed; meanwhile, a personalized treatment scheme is generated through intelligent matching based on an examination result and is automatically executed after being audited and confirmed by a doctor, so that the response speed and personalized accuracy are remarkably improved, medical resource utilization is effectively optimized, the auxiliary effect of the doctor in decision making is enhanced, and the efficiency and quality of medical services are integrally improved.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to an early warning method and system for intestinal obstruction risk in patients after colorectal cancer surgery. Background Art

[0002] In the recovery process of patients after colorectal cancer surgery, intestinal obstruction is a common and serious complication, and its early detection and timely intervention are crucial to the patient's recovery. However, existing technologies mainly rely on manual monitoring and manual preparation of examination and treatment plans, which has many shortcomings.

[0003] First, manual monitoring has the problem of delayed response. It is often difficult to capture subtle changes in the patient's postoperative recovery process in time, making it difficult to detect and deal with potential risks such as intestinal obstruction in a timely manner.

[0004] Secondly, the treatment plans under existing technologies lack sufficient personalization. Each patient's physical condition, surgical situation and postoperative recovery are different, but existing technologies often use standardized treatment plans, which are difficult to meet the personalized needs of patients, thus affecting the treatment effect.

[0005] In addition, manual monitoring and manual treatment planning also lead to a waste of medical resources. Due to the lack of efficient information technology, medical staff need to spend a lot of time and energy on data collection, organization and analysis, and patients also need to run between multiple departments, which not only increases the burden on patients, but also wastes precious medical resources.

[0006] In view of the above problems, the present invention proposes an early warning method and system for the risk of intestinal obstruction in patients after colorectal cancer surgery, aiming to improve the efficiency and quality of medical services by real-time monitoring of the patient's recovery and combining intelligent algorithms to generate personalized examination and treatment plans, while optimizing the utilization of medical resources. Summary of the invention

[0007] In order to overcome the problems raised in the above background technology, the present invention proposes an early warning method and system for the risk of intestinal obstruction in patients after colorectal cancer surgery.

[0008] The technical solution of the present invention is: a method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery, comprising the following steps:

[0009] S11: Data collection and organization, establishing medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and organizing and archiving the collected information;

[0010] S12: Data preprocessing: cleaning and organizing the collected data, removing invalid and redundant information, and standardizing the data;

[0011] S13: Feature extraction and selection: First, based on medical research and clinical experience, feature variables related to the risk of intestinal obstruction are extracted. Then, statistical methods and machine learning algorithms are used to screen out feature variables that have a significant impact on the prediction of intestinal obstruction risk.

[0012] S14: Model construction and training: select the appropriate machine learning algorithm according to the type and number of feature variables, train the model using historical data, and finally verify the model using the validation data set;

[0013] S15: Real-time warning: real-time monitoring of patients after colorectal cancer surgery, collection of data on their postoperative recovery, input of real-time monitoring data into the trained model, calculation of intestinal obstruction risk score, setting of warning threshold according to the risk score, and triggering of the warning mechanism when the score exceeds the threshold.

[0014] Preferably, when establishing the medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and arranging and archiving the collected information, the following steps are included:

[0015] S21: Establish a medical file. Establish an independent medical file for each patient who has undergone colorectal cancer surgery. The medical file is used to store and record relevant information of the patient who has undergone colorectal cancer surgery. The relevant information includes the patient's personal information, medical history, surgical records, and postoperative recovery status;

[0016] S22: Multi-channel information collection, collecting relevant information of patients after colorectal cancer surgery from multiple channels including the hospital electronic medical record system, clinical laboratory and imaging department;

[0017] S23: Information organization and archiving: Classify and organize the collected information, and archive the organized information into the medical records of patients after colorectal cancer surgery.

[0018] Preferably, when cleaning and arranging the collected data, removing invalid and redundant information and standardizing the data, the following steps are included:

[0019] S31: Data cleaning: first check whether there are missing values ​​in the data set and process the missing values, then process the outliers in the data set, and finally check and delete the duplicate values ​​in the data set;

[0020] S32: Data sorting: first, convert the data format to a unified format and convert text data to numerical data. Then, as needed, merge multiple related data sets into one data set and split the data with less correlation in one data set into multiple subsets.

[0021] S33: data standardization processing, including numerical standardization processing and category data coding processing;

[0022] S34: Perform quality check on the data by checking its distribution and statistical characteristics.

[0023] Preferably, when extracting characteristic variables related to the risk of intestinal obstruction based on medical research and clinical experience, and then using statistical methods and machine learning algorithms to screen out characteristic variables that have a significant impact on the prediction of the risk of intestinal obstruction, the following steps are included:

[0024] S41: Feature selection candidates, based on medical research and clinical experience, predetermine the feature variables that may be associated with the risk of intestinal obstruction, and use the determined feature variables as candidate features for feature extraction;

[0025] S42: feature extraction, extracting data of each item of candidate features from the original data by using feature extraction technology, wherein the feature extraction technology used includes correlation coefficient method, chi-square test method, principal component analysis method and linear discriminant analysis method;

[0026] S43: Feature selection. First, the extracted features are preliminarily screened using a machine learning algorithm to remove features with low correlation with the target variable and redundant features. Then, the feature importance evaluation method in the machine learning algorithm is used to rank the importance of the screened features. Finally, based on the feature importance evaluation results, the features that have a significant impact on the risk prediction of intestinal obstruction are selected as the final feature set. The feature importance evaluation methods used include the feature importance score in the random forest and the feature gain in the gradient boosting machine.

[0027] Preferably, when selecting a suitable machine learning algorithm according to the type and quantity of feature variables, training the model using historical data, and finally validating the model using a validation data set, the following steps are included:

[0028] S51: Algorithm selection, selecting a suitable machine learning algorithm according to the type and number of extracted feature variables, wherein the selectable machine learning algorithms include logistic regression algorithm, support vector machine algorithm, random forest algorithm and gradient boosting tree algorithm;

[0029] S52: Data preparation: collect and preprocess historical data, and then divide the preprocessed historical data into a training set, a validation set, and a test set;

[0030] S53: Model training: using the training set data to train the selected machine learning algorithm to obtain an initial model;

[0031] S54: Model validation, using validation set data to validate the trained model and evaluate the prediction performance of the model, where the selected evaluation indicators include accuracy, recall, F1 score, ROC curve and AUC value;

[0032] S55: Model testing: Use the test set data to test the model, evaluate the generalization ability and stability of the model, and make necessary adjustments and optimizations to the model based on the test results.

[0033] Preferably, real-time monitoring is performed on patients after colorectal cancer surgery to collect data on their postoperative recovery, the real-time monitoring data is input into a trained model, the intestinal obstruction risk score is calculated, and an early warning threshold is set according to the risk score. When the score exceeds the threshold, the early warning mechanism is triggered, including the following steps:

[0034] S61: Data collection: real-time monitoring of patients after colorectal cancer surgery and collection of data on their postoperative recovery;

[0035] S62: Data preprocessing, preprocessing the real-time monitoring data, wherein the adopted data preprocessing methods include data cleaning, data conversion and data standardization;

[0036] S63: Risk calculation, input the preprocessed data into the trained model, use the trained model to calculate the real-time monitoring data, and obtain the intestinal obstruction risk score;

[0037] S64: The early warning mechanism is triggered. When the intestinal obstruction risk score exceeds the pre-set early warning threshold, the early warning mechanism is automatically triggered;

[0038] S65: Doctor evaluation. After receiving the warning information triggered by the warning mechanism, the doctor on duty evaluates the condition of the patient after colorectal cancer surgery and processes the warning information triggered by the warning mechanism based on the evaluation results.

[0039] Preferably, real-time monitoring is performed on patients after colorectal cancer surgery to collect data on their postoperative recovery. The real-time monitoring data is input into a trained model to calculate the risk score of intestinal obstruction. An early warning threshold is set according to the risk score. When the score exceeds the threshold, the early warning mechanism is triggered, including the following steps:

[0040] S71: According to the specific conditions and risk scores of the postoperative colorectal cancer patient, a further examination plan is automatically generated and sent to the doctor for review, and then the generated examination plan is entered into the hospital's registration system for automatic registration. At the same time, the postoperative colorectal cancer patient and the person in charge of the hospital are notified. If the doctor does not review it within a limited time, the generated examination plan is directly entered into the hospital's registration system. The examination plan is obtained by matching the specific conditions and risk scores of the postoperative colorectal cancer patient with the data in the database;

[0041] S72: Processing and analyzing the examination results, identifying the specific physical health status of the patient after colorectal cancer surgery, and automatically generating a treatment plan based on the identification results, wherein the treatment plan includes a medication plan and a surgery plan, wherein the treatment plan is obtained by matching the examination results of the patient after colorectal cancer surgery with the data in the database;

[0042] S73: First, the generated treatment plan is sent to the doctor for review and confirmation. After the doctor adjusts and confirms the treatment plan, the treatment plan is automatically processed and executed.

[0043] A system for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery, comprising:

[0044] The data collection module is used to collect relevant information of patients after colorectal cancer surgery, wherein the relevant information collected from patients after colorectal cancer surgery includes medical history data, surgical records and postoperative recovery status;

[0045] A risk assessment module is used to assess the risk of intestinal obstruction in patients after colorectal cancer surgery using a risk assessment model based on the collected data;

[0046] The early warning prompt module is used to issue an early warning prompt when the risk assessment module identifies a high-risk patient;

[0047] The clinical intervention module is used to match the data in the database after the early warning prompt is issued, and automatically generate the inspection plan and clinical intervention plan;

[0048] The data tracking and feedback module is used to continuously track and monitor patients after colorectal cancer surgery.

[0049] Preferably, the data collection module includes the following data collection channels:

[0050] A11: Connect with the hospital's electronic medical record system to automatically collect and organize data in the electronic medical record system;

[0051] A12: Monitor the activity status and physiological parameters of patients after colorectal cancer surgery through wearable smart devices;

[0052] A13: Patients after colorectal cancer surgery regularly record their status through mini-programs and software. The recorded data includes text data and audio data.

[0053] A14: Connect with the hospital's monitoring system, record the video images inside the hospital, and identify the video images inside the hospital to identify the activity status and physical condition of patients after colorectal cancer surgery. The activity status of patients after colorectal cancer surgery includes the number of times and time the patients go to the toilet after colorectal cancer surgery, and the physical condition of patients after colorectal cancer surgery includes abdominal pain symptoms of patients after colorectal cancer surgery.

[0054] Preferably, the data collection module includes the following steps when working:

[0055] S81: data receiving, receiving data collected through channels A11-A14 and processing the data, wherein the adopted data processing method includes compressing audio data and video data;

[0056] S82: Data processing and identification, processing and identifying the collected data, wherein processing the collected data includes denoising and enhancing the video image data, cleaning the text data, parsing the audio data and converting it into text form, and identifying the collected data includes using image recognition technology to identify the activity status and physical condition of the postoperative colorectal cancer patient and parsing the text data input by the postoperative colorectal cancer patient and the text data converted from the audio data and extracting key information;

[0057] S83: Data integration: collect all the obtained data, clean the collected data, remove duplicate and invalid data, and form a data set.

[0058] Beneficial effects of the present invention:

[0059] 1. Compared with the existing technology that mainly relies on manual monitoring and manual formulation of examination and treatment plans, and has defects such as delayed response, insufficient personalization and waste of resources, this solution monitors the recovery of patients after colorectal cancer surgery in real time, intelligently generates examination plans based on risk scores and automatically handles the registration process. At the same time, it generates personalized treatment plans based on intelligent matching of examination results, and automatically executes them after review and confirmation by doctors. It not only significantly improves the response speed and personalization accuracy, but also effectively optimizes the utilization of medical resources, strengthens the auxiliary role of doctors in decision-making, and improves the efficiency and quality of medical services as a whole;

[0060] 2. Compared with the existing technology that mainly relies on a single hospital electronic medical record system to collect data, there are problems such as incomplete data, delayed updates and low patient participation. This solution integrates diversified data sources such as hospital electronic medical record systems, wearable smart devices, patient self-entry platforms and hospital monitoring systems, and combines efficient data reception, processing and recognition technologies, including data compression, denoising enhancement, text cleaning, audio analysis to text and image recognition, to form a comprehensive, real-time and high-quality data set, which not only improves patient participation, but also provides doctors with more accurate and personalized clinical decision support, significantly optimizing the management effect of intestinal obstruction risk in patients after colorectal cancer surgery;

[0061] 3. Through systematic collection, preprocessing, feature extraction and selection, model construction and training, and real-time warning, a method for accurately predicting the risk of intestinal obstruction is provided for patients after colorectal cancer surgery, which effectively improves the accuracy and timeliness of the warning, helps doctors take timely intervention measures, and reduces the incidence of intestinal obstruction and the risk of complications in patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 Shown is a schematic flow chart of a method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery of the present invention;

[0063] Figure 2 Shown is a schematic diagram of the structure of a system for the early warning method of the risk of intestinal obstruction in patients after colorectal cancer surgery of the present invention. DETAILED DESCRIPTION

[0064] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0065] See also Figure 1 The present invention provides an embodiment: a method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery, comprising the following steps:

[0066] S11: Data collection and organization, establishing medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and organizing and archiving the collected information;

[0067] S12: Data preprocessing: cleaning and organizing the collected data, removing invalid and redundant information, and standardizing the data;

[0068] S13: Feature extraction and selection: First, based on medical research and clinical experience, feature variables related to the risk of intestinal obstruction are extracted. Then, statistical methods and machine learning algorithms are used to screen out feature variables that have a significant impact on the prediction of intestinal obstruction risk.

[0069] S14: Model construction and training: select the appropriate machine learning algorithm according to the type and number of feature variables, train the model using historical data, and finally verify the model using the validation data set;

[0070] S15: Real-time warning: real-time monitoring of patients after colorectal cancer surgery, collection of data on their postoperative recovery, input of real-time monitoring data into the trained model, calculation of intestinal obstruction risk score, setting of warning threshold according to the risk score, and triggering of the warning mechanism when the score exceeds the threshold.

[0071] As described above, the present invention provides a method for accurately predicting the risk of intestinal obstruction for patients after colorectal cancer surgery through systematic collection, preprocessing, feature extraction and selection, model construction and training, and real-time warning, which effectively improves the accuracy and timeliness of the warning, helps doctors take intervention measures in time, and reduces the incidence of intestinal obstruction and the risk of complications in patients.

[0072] Preferably, when establishing the medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and arranging and archiving the collected information, the following steps are included:

[0073] S21: Establish a medical file. Establish an independent medical file for each patient who has undergone colorectal cancer surgery. The medical file is used to store and record relevant information of the patient who has undergone colorectal cancer surgery. The relevant information includes the patient's personal information, medical history, surgical records, and postoperative recovery status;

[0074] S22: Multi-channel information collection, collecting relevant information of patients after colorectal cancer surgery from multiple channels including the hospital electronic medical record system, clinical laboratory and imaging department;

[0075] S23: Information organization and archiving: Classify and organize the collected information, and archive the organized information into the medical records of patients after colorectal cancer surgery.

[0076] As described above, the present invention establishes independent medical records for patients after colorectal cancer surgery, and comprehensively collects relevant information from multiple channels such as the hospital's electronic medical record system, clinical laboratory, and imaging department, and then archives it after careful classification and organization, thereby effectively ensuring the integrity and accuracy of patient information, providing solid data support for the follow-up treatment, rehabilitation evaluation, and medical research of postoperative patients, and greatly improving the efficiency and quality of medical services.

[0077] Preferably, when cleaning and arranging the collected data, removing invalid and redundant information and standardizing the data, the following steps are included:

[0078] S31: Data cleaning: first check whether there are missing values ​​in the data set and process the missing values, then process the outliers in the data set, and finally check and delete the duplicate values ​​in the data set;

[0079] S32: Data sorting: first, convert the data format to a unified format and convert text data to numerical data. Then, as needed, merge multiple related data sets into one data set and split the data with less correlation in one data set into multiple subsets.

[0080] S33: data standardization processing, including numerical standardization processing and category data coding processing;

[0081] S34: Perform quality check on the data by checking its distribution and statistical characteristics.

[0082] As described above, the present invention effectively removes invalid and redundant information and ensures the accuracy, consistency and availability of data by implementing a series of steps such as data cleaning, sorting, standardization and quality inspection. The data cleaning step solves the problems of missing values, abnormal values ​​and duplicate values, and the data sorting step realizes the unification of data format and the numerical conversion of text data, while optimizing the structure of the data set. Data standardization improves the comparability of data, and data quality inspection further ensures the reliability and scientificity of data, laying a solid foundation for subsequent data analysis and application.

[0083] Preferably, when extracting characteristic variables related to the risk of intestinal obstruction based on medical research and clinical experience, and then using statistical methods and machine learning algorithms to screen out characteristic variables that have a significant impact on the prediction of the risk of intestinal obstruction, the following steps are included:

[0084] S41: Feature selection candidates, based on medical research and clinical experience, predetermine the feature variables that may be associated with the risk of intestinal obstruction, and use the determined feature variables as candidate features for feature extraction;

[0085] S42: feature extraction, extracting data of each item of candidate features from the original data by using feature extraction technology, wherein the feature extraction technology used includes correlation coefficient method, chi-square test method, principal component analysis method and linear discriminant analysis method;

[0086] S43: Feature selection. First, the extracted features are preliminarily screened using a machine learning algorithm to remove features with low correlation with the target variable and redundant features. Then, the feature importance evaluation method in the machine learning algorithm is used to rank the importance of the screened features. Finally, based on the feature importance evaluation results, the features that have a significant impact on the risk prediction of intestinal obstruction are selected as the final feature set. The feature importance evaluation methods used include the feature importance score in the random forest and the feature gain in the gradient boosting machine.

[0087] As described above, the present invention scientifically screens candidate features related to the risk of intestinal obstruction by integrating medical research and clinical experience, and uses a variety of feature extraction techniques such as correlation coefficient method, chi-square test method, principal component analysis method and linear discriminant analysis method to accurately extract feature information from the original data. Subsequently, with the help of machine learning algorithms, preliminary screening and feature importance evaluation were carried out to effectively eliminate low-correlation and redundant features, and the most influential feature set for intestinal obstruction risk prediction was selected based on feature importance sorting, which significantly improved the accuracy and reliability of the risk prediction model.

[0088] Preferably, when selecting a suitable machine learning algorithm according to the type and quantity of feature variables, training the model using historical data, and finally validating the model using a validation data set, the following steps are included:

[0089] S51: Algorithm selection, selecting a suitable machine learning algorithm according to the type and number of extracted feature variables, wherein the selectable machine learning algorithms include logistic regression algorithm, support vector machine algorithm, random forest algorithm and gradient boosting tree algorithm;

[0090] S52: Data preparation: collect and preprocess historical data, and then divide the preprocessed historical data into a training set, a validation set, and a test set;

[0091] S53: Model training: using the training set data to train the selected machine learning algorithm to obtain an initial model;

[0092] S54: Model validation, using validation set data to validate the trained model and evaluate the prediction performance of the model, where the selected evaluation indicators include accuracy, recall, F1 score, ROC curve and AUC value;

[0093] S55: Model testing: Use the test set data to test the model, evaluate the generalization ability and stability of the model, and make necessary adjustments and optimizations to the model based on the test results.

[0094] As described above, the present invention effectively ensures the adequacy of model training and the objectivity of verification by carefully selecting machine learning algorithms according to the characteristics of feature variables and scientifically dividing historical data into training sets, validation sets and test sets. After the initial model is trained with the training set data, the model performance is evaluated in multiple dimensions using the validation set data, including accuracy, recall rate, F1 score, ROC curve and AUC value, which fully guarantees the prediction accuracy of the model. Finally, the generalization ability and stability of the model are verified by the test set data, and targeted adjustments and optimizations are performed according to the test results, which significantly improves the practicality and reliability of the model.

[0095] Preferably, real-time monitoring is performed on patients after colorectal cancer surgery to collect data on their postoperative recovery, the real-time monitoring data is input into a trained model, the intestinal obstruction risk score is calculated, and an early warning threshold is set according to the risk score. When the score exceeds the threshold, the early warning mechanism is triggered, including the following steps:

[0096] S61: Data collection: real-time monitoring of patients after colorectal cancer surgery and collection of data on their postoperative recovery;

[0097] S62: Data preprocessing, preprocessing the real-time monitoring data, wherein the adopted data preprocessing methods include data cleaning, data conversion and data standardization;

[0098] S63: Risk calculation, input the preprocessed data into the trained model, use the trained model to calculate the real-time monitoring data, and obtain the intestinal obstruction risk score;

[0099] S64: The early warning mechanism is triggered. When the intestinal obstruction risk score exceeds the pre-set early warning threshold, the early warning mechanism is automatically triggered;

[0100] S65: Doctor evaluation. After receiving the warning information triggered by the warning mechanism, the doctor on duty evaluates the condition of the patient after colorectal cancer surgery and processes the warning information triggered by the warning mechanism based on the evaluation results.

[0101] As described above, the present invention monitors the recovery of patients after colorectal cancer surgery in real time, collects and pre-processes data, uses a trained model to accurately calculate the risk score of intestinal obstruction, and sets a warning threshold. Once the risk score exceeds the threshold, the warning mechanism is automatically triggered, and the doctor is notified in time for evaluation and treatment. This solution not only improves the timeliness and accuracy of intestinal obstruction risk monitoring, but also ensures that doctors can respond quickly and take effective measures, effectively reducing the risk of postoperative intestinal obstruction and improving the quality and safety of patient recovery.

[0102] Preferably, real-time monitoring is performed on patients after colorectal cancer surgery to collect data on their postoperative recovery. The real-time monitoring data is input into a trained model to calculate the risk score of intestinal obstruction. An early warning threshold is set according to the risk score. When the score exceeds the threshold, the early warning mechanism is triggered, including the following steps:

[0103] S71: According to the specific conditions and risk scores of the postoperative colorectal cancer patient, a further examination plan is automatically generated and sent to the doctor for review, and then the generated examination plan is entered into the hospital's registration system for automatic registration. At the same time, the postoperative colorectal cancer patient and the person in charge of the hospital are notified. If the doctor does not review it within a limited time, the generated examination plan is directly entered into the hospital's registration system. The examination plan is obtained by matching the specific conditions and risk scores of the postoperative colorectal cancer patient with the data in the database;

[0104] S72: Processing and analyzing the examination results, identifying the specific physical health status of the patient after colorectal cancer surgery, and automatically generating a treatment plan based on the identification results, wherein the treatment plan includes a medication plan and a surgery plan, wherein the treatment plan is obtained by matching the examination results of the patient after colorectal cancer surgery with the data in the database;

[0105] S73: First, the generated treatment plan is sent to the doctor for review and confirmation. After the doctor adjusts and confirms the treatment plan, the treatment plan is automatically processed and executed.

[0106] Compared with the existing technology that mainly relies on manual monitoring and manual formulation of examination and treatment plans, and has defects such as delayed response, insufficient personalization and waste of resources, this solution monitors the recovery of patients after colorectal cancer surgery in real time, intelligently generates examination plans based on risk scores and automatically handles the registration process. At the same time, it generates personalized treatment plans (including medication and surgery) based on intelligent matching of examination results, and automatically executes them after review and confirmation by the doctor. It not only significantly improves the response speed and personalization accuracy, but also effectively optimizes the utilization of medical resources, strengthens the auxiliary role of doctors in decision-making, and improves the efficiency and quality of medical services as a whole.

[0107] See also Figure 2 The present invention provides an embodiment: a system for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery, comprising:

[0108] The data collection module is used to collect relevant information of patients after colorectal cancer surgery, wherein the relevant information collected from patients after colorectal cancer surgery includes medical history data, surgical records and postoperative recovery status;

[0109] A risk assessment module is used to assess the risk of intestinal obstruction in patients after colorectal cancer surgery using a risk assessment model based on the collected data;

[0110] The early warning prompt module is used to issue an early warning prompt when the risk assessment module identifies a high-risk patient;

[0111] The clinical intervention module is used to match the data in the database after the early warning prompt is issued, and automatically generate the inspection plan and clinical intervention plan;

[0112] The data tracking and feedback module is used to continuously track and monitor patients after colorectal cancer surgery.

[0113] As described above, the present invention realizes comprehensive and intelligent management of the risk of intestinal obstruction in patients after colorectal cancer surgery by integrating modules such as data collection, risk assessment, early warning prompts, clinical intervention, and data tracking and feedback. It not only improves the accuracy and timeliness of risk assessment, but also can quickly generate personalized inspection and intervention plans when high-risk patients are identified, and ensure the treatment effect through continuous tracking and monitoring, thereby effectively reducing the risk of postoperative intestinal obstruction and improving the patient's rehabilitation quality and the overall efficiency of medical services.

[0114] Preferably, the data collection module includes the following data collection channels:

[0115] A11: Connect with the hospital's electronic medical record system to automatically collect and organize data in the electronic medical record system;

[0116] A12: Monitor the activity status and physiological parameters of patients after colorectal cancer surgery through wearable smart devices;

[0117] A13: Patients after colorectal cancer surgery regularly record their status through mini-programs and software. The recorded data includes text data and audio data.

[0118] A14: Connect with the hospital's monitoring system, record the video images inside the hospital, and identify the video images inside the hospital to identify the activity status and physical condition of patients after colorectal cancer surgery. The activity status of patients after colorectal cancer surgery includes the number of times and time the patients go to the toilet after colorectal cancer surgery, and the physical condition of patients after colorectal cancer surgery includes abdominal pain symptoms of patients after colorectal cancer surgery.

[0119] Preferably, the data collection module includes the following steps when working:

[0120] S81: data receiving, receiving data collected through channels A11-A14 and processing the data, wherein the adopted data processing method includes compressing audio data and video data;

[0121] S82: Data processing and identification, processing and identifying the collected data, wherein processing the collected data includes denoising and enhancing the video image data, cleaning the text data, parsing the audio data and converting it into text form, and identifying the collected data includes using image recognition technology to identify the activity status and physical condition of the postoperative colorectal cancer patient and parsing the text data input by the postoperative colorectal cancer patient and the text data converted from the audio data and extracting key information;

[0122] S83: Data integration: collect all the obtained data, clean the collected data, remove duplicate and invalid data, and form a data set.

[0123] Compared with the existing technology that mainly relies on a single hospital electronic medical record system to collect data, there are problems such as incomplete data, delayed updates and low patient participation. This solution integrates diversified data sources such as hospital electronic medical record systems, wearable smart devices, patient self-entry platforms and hospital monitoring systems, combined with efficient data reception, processing and recognition technologies, including data compression, denoising enhancement, text cleaning, audio analysis to text and image recognition, to form a comprehensive, real-time and high-quality data set, which not only improves patient participation, but also provides doctors with more accurate and personalized clinical decision support, significantly optimizing the management effect of intestinal obstruction risk in patients after colorectal cancer surgery.

[0124] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery; characterized by: The following steps are included: S11: Data collection and organization, establishing medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and organizing and archiving the collected information; S12: Data preprocessing: cleaning and organizing the collected data, removing invalid and redundant information, and standardizing the data; S13: Feature extraction and selection: First, based on medical research and clinical experience, feature variables related to the risk of intestinal obstruction are extracted. Then, statistical methods and machine learning algorithms are used to screen out feature variables that have a significant impact on the prediction of intestinal obstruction risk. S14: Model construction and training: select the appropriate machine learning algorithm according to the type and number of feature variables, train the model using historical data, and finally verify the model using the validation data set; S15: Real-time warning: real-time monitoring of patients after colorectal cancer surgery, collection of data on their postoperative recovery, input of real-time monitoring data into the trained model, calculation of intestinal obstruction risk score, setting of warning threshold according to the risk score, and triggering of the warning mechanism when the score exceeds the threshold.

2. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 1, characterized in that: The following steps are included in establishing medical records of patients after colorectal cancer surgery, collecting relevant information of patients after colorectal cancer surgery from multiple channels, and arranging and archiving the collected information: S21: Establish a medical file. Establish an independent medical file for each patient who has undergone colorectal cancer surgery. The medical file is used to store and record relevant information of the patient who has undergone colorectal cancer surgery. The relevant information includes the patient's personal information, medical history, surgical records, and postoperative recovery status; S22: Multi-channel information collection, collecting relevant information of patients after colorectal cancer surgery from multiple channels including the hospital electronic medical record system, clinical laboratory and imaging department; S23: Information organization and archiving: Classify and organize the collected information, and archive the organized information into the medical records of patients after colorectal cancer surgery.

3. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 2, characterized in that: When cleaning and organizing the collected data, removing invalid and redundant information and standardizing the data, the following steps are included: S31: Data cleaning: first check whether there are missing values ​​in the data set and process the missing values, then process the outliers in the data set, and finally check and delete the duplicate values ​​in the data set; S32: Data sorting: first, convert the data format to a unified format and convert text data to numerical data. Then, as needed, merge multiple related data sets into one data set and split the data with less correlation in one data set into multiple subsets. S33: data standardization processing, including numerical standardization processing and category data coding processing; S34: Perform quality check on the data by checking its distribution and statistical characteristics.

4. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 3, characterized in that: Based on medical research and clinical experience, characteristic variables related to the risk of intestinal obstruction are extracted, and then statistical methods and machine learning algorithms are used to screen out characteristic variables that have a significant impact on the prediction of intestinal obstruction risk, including the following steps: S41: Feature selection candidates, based on medical research and clinical experience, predetermine the feature variables that may be associated with the risk of intestinal obstruction, and use the determined feature variables as candidate features for feature extraction; S42: feature extraction, extracting data of each item of candidate features from the original data by using feature extraction technology, wherein the feature extraction technology used includes correlation coefficient method, chi-square test method, principal component analysis method and linear discriminant analysis method; S43: Feature selection. First, the extracted features are preliminarily screened using a machine learning algorithm to remove features with low correlation with the target variable and redundant features. Then, the feature importance evaluation method in the machine learning algorithm is used to rank the importance of the screened features. Finally, based on the feature importance evaluation results, the features that have a significant impact on the risk prediction of intestinal obstruction are selected as the final feature set. The feature importance evaluation methods used include the feature importance score in the random forest and the feature gain in the gradient boosting machine.

5. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 4, characterized in that: The following steps are involved in selecting a suitable machine learning algorithm based on the type and number of feature variables, training the model using historical data, and finally validating the model using a validation dataset: S51: Algorithm selection, selecting a suitable machine learning algorithm according to the type and number of extracted feature variables, wherein the selectable machine learning algorithms include logistic regression algorithm, support vector machine algorithm, random forest algorithm and gradient boosting tree algorithm; S52: Data preparation: collect and preprocess historical data, and then divide the preprocessed historical data into a training set, a validation set, and a test set; S53: Model training, using the training set data to train the selected machine learning algorithm to obtain an initial model; S54: Model validation, using validation set data to validate the trained model and evaluate the prediction performance of the model, where the selected evaluation indicators include accuracy, recall, F1 score, ROC curve and AUC value; S55: Model testing: Use the test set data to test the model, evaluate the generalization ability and stability of the model, and make necessary adjustments and optimizations to the model based on the test results.

6. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 5, characterized in that: Real-time monitoring of patients after colorectal cancer surgery is performed to collect data on their postoperative recovery. The real-time monitoring data is input into the trained model to calculate the intestinal obstruction risk score. The warning threshold is set according to the risk score. When the score exceeds the threshold, the warning mechanism is triggered, including the following steps: S61: Data collection: real-time monitoring of patients after colorectal cancer surgery and collection of data on their postoperative recovery; S62: Data preprocessing, preprocessing the real-time monitoring data, wherein the adopted data preprocessing methods include data cleaning, data conversion and data standardization; S63: Risk calculation, input the preprocessed data into the trained model, use the trained model to calculate the real-time monitoring data, and obtain the intestinal obstruction risk score; S64: The early warning mechanism is triggered. When the intestinal obstruction risk score exceeds the pre-set early warning threshold, the early warning mechanism is automatically triggered; S65: Doctor evaluation. After receiving the warning information triggered by the warning mechanism, the doctor on duty evaluates the condition of the patient after colorectal cancer surgery and processes the warning information triggered by the warning mechanism based on the evaluation results.

7. The method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 6, characterized in that: Real-time monitoring of patients after colorectal cancer surgery is performed to collect data on their postoperative recovery. The real-time monitoring data is input into the trained model to calculate the intestinal obstruction risk score. The warning threshold is set according to the risk score. When the score exceeds the threshold, the warning mechanism is triggered, including the following steps: S71: According to the specific conditions and risk scores of the postoperative colorectal cancer patient, a further examination plan is automatically generated and sent to the doctor for review, and then the generated examination plan is entered into the hospital's registration system for automatic registration. At the same time, the postoperative colorectal cancer patient and the person in charge of the hospital are notified. If the doctor does not review it within a limited time, the generated examination plan is directly entered into the hospital's registration system. The examination plan is obtained by matching the specific conditions and risk scores of the postoperative colorectal cancer patient with the data in the database; S72: Processing and analyzing the examination results, identifying the specific physical health status of the patient after colorectal cancer surgery, and automatically generating a treatment plan based on the identification results, wherein the treatment plan includes a medication plan and a surgery plan, wherein the treatment plan is obtained by matching the examination results of the patient after colorectal cancer surgery with the data in the database; S73: First, the generated treatment plan is sent to the doctor for review and confirmation. After the doctor adjusts and confirms the treatment plan, the treatment plan is automatically processed and executed.

8. The system of the method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 7, characterized in that: Included are: The data collection module is used to collect relevant information of patients after colorectal cancer surgery, wherein the relevant information collected from patients after colorectal cancer surgery includes medical history data, surgical records and postoperative recovery status; A risk assessment module is used to assess the risk of intestinal obstruction in patients after colorectal cancer surgery using a risk assessment model based on the collected data; The early warning prompt module is used to issue an early warning prompt when the risk assessment module identifies a high-risk patient; The clinical intervention module is used to match the data in the database after the early warning prompt is issued, and automatically generate the inspection plan and clinical intervention plan; The data tracking and feedback module is used to continuously track and monitor patients after colorectal cancer surgery.

9. The system of the method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 8, characterized in that: The data collection module includes the following data collection channels: A11: Connect with the hospital's electronic medical record system to automatically collect and organize data in the electronic medical record system; A12: Monitor the activity status and physiological parameters of patients after colorectal cancer surgery through wearable smart devices; A13: Patients after colorectal cancer surgery regularly record their status through mini-programs and software. The recorded data includes text data and audio data. A14: Connect with the hospital's monitoring system, record the video images inside the hospital, and identify the video images inside the hospital to identify the activity status and physical condition of patients after colorectal cancer surgery. The activity status of patients after colorectal cancer surgery includes the number of times and time the patients go to the toilet after colorectal cancer surgery, and the physical condition of patients after colorectal cancer surgery includes abdominal pain symptoms of patients after colorectal cancer surgery.

10. The system of the method for early warning of the risk of intestinal obstruction in patients after colorectal cancer surgery according to claim 9, characterized in that: The data collection module includes the following steps when working: S81: data receiving, receiving data collected through channels A11-A14 and processing the data, wherein the adopted data processing method includes compressing audio data and video data; S82: Data processing and identification, processing and identifying the collected data, wherein processing the collected data includes denoising and enhancing the video image data, cleaning the text data, parsing the audio data and converting it into text form, and identifying the collected data includes using image recognition technology to identify the activity status and physical condition of the postoperative colorectal cancer patient and parsing the text data input by the postoperative colorectal cancer patient and the text data converted from the audio data and extracting key information; S83: Data integration: collect all the obtained data, clean the collected data, remove duplicate and invalid data, and form a data set.

Images