Prediction method for intestinal preparation of patient before colonoscopy based on machine learning

Abstract

The invention discloses a method for predicting intestinal preparation of a patient before colonoscopy based on machine learning, and the method comprises the following steps: S1, collecting clinical data of the patient, and carrying out the preprocessing of the clinical data; s2, performing feature selection on the collected clinical data by using a Boruta algorithm, and identifying important features related to insufficient intestinal preparation before colonoscopy; s3, constructing a prediction model by utilizing a machine learning algorithm, evaluating the prediction model by adopting a cross validation method, and selecting an optimal prediction model; s4, explaining the optimal prediction model by adopting an SHAP explaining tree model, and identifying a feature which has the greatest influence on a prediction result; s5, clinical data of a to-be-detected patient are transmitted to the optimal prediction model, and intestinal preparation of the patient before colonoscopy is predicted; according to the method, the risk of insufficient intestinal preparation can be accurately predicted, personalized guidance can be provided, the clinical working efficiency is remarkably improved, and the method has important clinical significance.

Description

Technical Field

[0001] The present invention relates to the field of computer processing technology, and in particular to a method for predicting intestinal preparation of patients before colonoscopy based on machine learning. Background Art

[0002] Colonoscopy is an important means of diagnosing and treating colon diseases, and good bowel preparation is an important prerequisite for ensuring the success of colonoscopy. Traditional bowel preparation assessment methods rely on clinical experience and paper-based guidance, and have problems with accuracy and individualization. Existing studies have shown that machine learning technology can provide more accurate predictions by analyzing large amounts of data, and has broad application prospects in the medical field.

[0003] Patients, especially elderly patients, are at higher risk of inadequate bowel preparation due to factors such as age, medical history, and medication use. Existing assessment methods rely on clinical experience and lack scientific prediction models, leading to the following problems: 1) Insufficient prediction accuracy, unable to accurately identify high-risk patients; 2) Low degree of individualization, unable to provide personalized advice based on the patient's specific circumstances; 3) Heavy clinical workload, difficult to effectively manage bowel preparation for a large number of patients.

[0004] Therefore, the existing methods for evaluating bowel preparation before colonoscopy lack effective analysis and utilization of large amounts of data, making it difficult to achieve accurate predictions. In particular, the existing technologies do not fully consider the special needs of elderly patients, are not sufficiently individualized, and lack identification and quantitative analysis of key influencing factors, resulting in insufficient robustness and reliability of the prediction model. Summary of the invention

[0005] To solve the above problems, the present invention proposes a method for predicting intestinal preparation of patients before colonoscopy based on machine learning. The method deploys the optimal prediction model in a web application, and clinicians can obtain prediction results and guidance suggestions by inputting patient information. The constructed optimal prediction model based on machine learning can not only accurately predict the risk of insufficient intestinal preparation, but also provide personalized guidance, which significantly improves clinical work efficiency and has important clinical significance.

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

[0007] A method for predicting bowel preparation of patients before colonoscopy based on machine learning, comprising the following steps:

[0008] S1. Collect the patient's clinical data and pre-process the clinical data;

[0009] S2. Use the Boruta algorithm to perform feature selection on the collected clinical data to identify important features associated with inadequate bowel preparation before colonoscopy;

[0010] S3. Use machine learning algorithms to build prediction models, and use cross-validation methods to evaluate the prediction models and select the best prediction model;

[0011] S4. Use the SHAP interpretation tree model to explain the best prediction model and identify the features that have the greatest impact on the prediction results;

[0012] S5. The clinical data of the patient to be tested is transmitted to the optimal prediction model to predict the patient's bowel preparation before colonoscopy.

[0013] Preferably, the clinical data in step S1 include age, gender, medical history, medication use, daily bowel movements, education and activity.

[0014] Preferably, the preprocessing of clinical data in step S1 includes data cleaning and normalization.

[0015] Preferably, the machine learning algorithm in step S3 includes support vector machine, gradient boosting machine, decision tree, random forest and extreme gradient boosting.

[0016] Preferably, the specific process of step S5 is:

[0017] S51, model deployment: deploy the trained best prediction model on the server;

[0018] S52. Front-end design: Build the front-end user interface for clinicians to input patient information and view prediction results; the front-end is designed using modern Web technology;

[0019] S53, back-end design: The back-end uses Python programming language and implements the API interface of the best prediction model through Flask or Django framework, which interacts data between the server, back-end and front-end, and is used to receive patient information input by the front-end, call the best prediction model and return the results;

[0020] S54. Database management: Use a relational database to store patient data and prediction results;

[0021] S55. Result prediction: Input the clinical data of the patient to be tested in the user interface, analyze the important features and the features that have the greatest impact on the predicted results, and generate predicted results and medical recommendations.

[0022] Preferably, step S5 also includes implementing data encryption and access control measures to protect the security and privacy of patient data.

[0023] After adopting the above technical solution, the present invention has the following beneficial effects:

[0024] 1. The present invention deploys the optimal prediction model in a web application. Clinicians can obtain prediction results and guidance suggestions by inputting patient information. The optimal prediction model constructed based on machine learning can not only accurately predict the risk of insufficient intestinal preparation, but also provide personalized guidance, which significantly improves clinical work efficiency and has important clinical significance.

[0025] 2. The present invention adopts a multi-level, multi-model interpretable machine learning framework to improve the reliability and user interactivity of the system. High prediction accuracy: A large amount of clinical data is analyzed using machine learning algorithms to improve the accuracy of predictions. Individualized guidance: The optimal prediction model takes into account the individual characteristics of elderly patients and provides accurate bowel preparation guidance. Convenient operation: Through the web application, clinical medical staff can easily enter patient information and obtain prediction results, which improves work efficiency. Data security: Strict data encryption and access control measures are implemented to ensure the security and privacy protection of patient data. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0028] like Figure 1 As shown, a method for predicting bowel preparation before colonoscopy based on machine learning includes the following steps:

[0029] S1. Collect the patient's clinical data and pre-process the clinical data;

[0030] The clinical data in step S1 include age, gender, medical history, medication use, daily bowel movements, education, and activity;

[0031] The preprocessing of clinical data in step S1 includes data cleaning and normalization;

[0032] S2. Use the Boruta algorithm to perform feature selection on the collected clinical data to identify important features associated with inadequate bowel preparation before colonoscopy;

[0033] S3. Use machine learning algorithms to build prediction models, and use cross-validation methods to evaluate the prediction models and select the best prediction model;

[0034] The machine learning algorithms described in step S3 include support vector machine, gradient boosting machine, decision tree, random forest and extreme gradient boosting;

[0035] S4. Use the SHAP interpretation tree model to explain the best prediction model and identify the features that have the greatest impact on the prediction results;

[0036] S5. Transmitting the clinical data of the patient to be tested to the optimal prediction model to predict the patient's bowel preparation before colonoscopy;

[0037] The specific process of step S5 is:

[0038] S51, model deployment: deploy the trained best prediction model on the server;

[0039] S52. Front-end design: Build the front-end user interface for clinicians to input patient information and view prediction results; the front-end is designed using modern Web technology;

[0040] S53, back-end design: The back-end uses Python programming language and implements the API interface of the best prediction model through Flask or Django framework, which interacts data between the server, back-end and front-end, and is used to receive patient information input by the front-end, call the best prediction model and return the results;

[0041] S54. Database management: Use a relational database to store patient data and prediction results;

[0042] S55, result prediction: input clinical data of the patient to be tested in the user interface, analyze the important features and the features that have the greatest impact on the prediction results, and generate prediction results and medical advice;

[0043] Step S5 also includes implementing data encryption and access control measures to protect the security and privacy of patient data.

[0044] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for predicting intestinal preparation of patients before colonoscopy based on machine learning, characterized in that: The following steps are involved: S1. Collect the patient's clinical data and pre-process the clinical data; S2. Use the Boruta algorithm to perform feature selection on the collected clinical data to identify important features associated with inadequate bowel preparation before colonoscopy; S3. Use machine learning algorithms to build prediction models, and use cross-validation methods to evaluate the prediction models and select the best prediction model; S4. Use the SHAP interpretation tree model to explain the best prediction model and identify the features that have the greatest impact on the prediction results; S5. The clinical data of the patient to be tested is transmitted to the optimal prediction model to predict the patient's bowel preparation before colonoscopy.

2. The method for predicting the intestinal preparation of a patient before colonoscopy based on machine learning according to claim 1, characterized in that: The clinical data in step S1 include age, gender, past medical history, medication use, daily bowel movements, education, and activity.

3. The method for predicting intestinal preparation of a patient before colonoscopy based on machine learning as claimed in claim 1, characterized in that: The preprocessing of clinical data in step S1 includes data cleaning and normalization.

4. The method for predicting the intestinal preparation of a patient before colonoscopy based on machine learning as claimed in claim 1, characterized in that: The machine learning algorithms described in step S3 include support vector machine, gradient boosting machine, decision tree, random forest and extreme gradient boosting.

5. The method for predicting the intestinal preparation of a patient before colonoscopy based on machine learning as claimed in claim 1, characterized in that: The specific process of step S5 is: S51, model deployment: deploy the trained best prediction model on the server; S52. Front-end design: Build the front-end user interface for clinicians to input patient information and view prediction results. The front-end is designed using modern Web technology. S53, back-end design: The back-end uses Python programming language and implements the API interface of the best prediction model through Flask or Django framework, which interacts data between the server, back-end and front-end, and is used to receive patient information input by the front-end, call the best prediction model and return the results; S54. Database management: Use a relational database to store patient data and prediction results; S55. Result prediction: Input the clinical data of the patient to be tested in the user interface, analyze the important features and the features that have the greatest impact on the predicted results, and generate predicted results and medical recommendations.

6. The method for predicting the intestinal preparation of a patient before colonoscopy based on machine learning as claimed in claim 1, characterized in that: Step S5 also includes implementing data encryption and access control measures to protect the security and privacy of patient data.

Images