Construction of a prediction model for postoperative survival of pancreatic cancer based on SMOTE-ENN combined with XGBoost algorithm
10.3969/j.issn.1673-9701.2025.28.006
- VernacularTitle:基于SMOTE-ENN结合XGBoost算法构建胰腺癌术后生存预测模型
- Author:
Jiaqi WANG
1
;
Yanhong LUO
;
Yarong GUO
Author Information
1. 山西医科大学医学科学院,山西太原 030001
- Publication Type:Journal Article
- Keywords:
Pancreatic cancer;
Imbalanced data;
XGBoost;
Outcome prediction
- From:
China Modern Doctor
2025;63(28):23-28,34
- CountryChina
- Language:Chinese
-
Abstract:
Objective A survival outcome prediction model for postoperative pancreatic cancer patients was constructed by applying large-scale data based on the new version of American Joint Committee on Cancer(AJCC)staging using different machine learning algorithms.Methods Based on the Surveillance,Epidemiology,and End Results(SEER)database,synthetic minority over-sampling technique(SMOTE)and synthetic minority over-sampling technique and edited nearest neighbors(SMOTE-ENN)algorithms were used to process unbalanced data,random forest,support vector machine,decision tree,and extreme gradient boosting(XGBoost)algorithms were used to build and compare prognostic models,and Shapley additive explanation(SHAP)was introduced to interpret the models.Results The SMOTE-ENN combined with XGBoost model had the best performance(accuracy rate was 86.2％,precision rate was 95.2％,recall rate was 71.2％,F1 value was 0.762,area under the curve was 0.884,Brier score was 0.108).The calibration curve and decision curve respectively showed that this model had good calibration effect and high clinical application value.In addition,SHAP analysis showed that the most important impact on prognostic outcomes was N stage.Conclusion The XGBoost model has the best performance and can be used as a new high-performance postoperative prognosis prediction model under AJCC staging that conforms to the current clinical staging system,providing theoretical support for predicting postoperative patient survival outcomes and formulating personalized treatment plans.
