Machine learning-based real-time prediction of duodenal stump leakage from gastrectomy in gastric cancer patients

Abstract

Purpose: This study aimed to develop a machine learning (ML) model for real-time prediction of duodenal stump leakage (DSL) following gastrectomy in patients with gastric cancer (GC) using a comprehensive set of clinical variables to improve postoperative outcomes and monitoring efficiency.

Methods: A retrospective analysis was conducted on 1,107 patients with GC who underwent gastrectomy at Pusan National University Yangsan Hospital between 2019 and 2022. One hundred eighty-nine features were extracted from each patient record, including demographic data, preoperative comorbidities, and blood test outcomes from the subsequent seven postoperative days (POD). Six ML algorithms were evaluated: Logistic Regression (LR), K-nearest neighbors (KNN), Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGB), and Neural Network (NN). The models predicted DSL occurrence preoperatively and on POD 1, 2, 3, 5, and 7. Performance was assessed using the Area Under the Receiver Operating Characteristic Curve (AUROC) and Recall@K.

Results: Among the 1,107 patients, 29 developed DSL. XGB demonstrated the highest AUROC score (0.880), followed by RF (0.858), LR (0.823), SVM (0.819), NN (0.753), and KNN (0.726). The RF achieved the best Recall@K score of 0.643. Including additional POD features improved the predictive performance, with the AUROC value increasing to 0.879 on POD 7. The confidence scores of the model indicated that the DSL predictions became more reliable over time.

Conclusion: The study concluded that ML models, notably the XGB algorithm, can effectively predict DSL in real-time using comprehensive clinical data, enhancing the clinical decision-making process for GC patients.

Keywords: duodenal stump leakage; gastrectomy; gastric cancer; machine learning; predictive modeling.

Figure 1

Consort diagram.

Figure 2

Distribution of 189 features across treatment phases.

Figure 3

3-fold cross-validation of performance.

Figure 4

Comprehensive performance of all models across all prediction time (A) AUROC of machine learning models. (B) Recall @ 10%. LR, logistic regression; KNN, K-nearest neighbors; SVM, support vector machine; RF, random forest; XGB, extreme gradient boosting; NN, neural network.

Figure 5

Improvement of AUROC scores with additional features. AUROC, area under receiver operating characteristic curve; SD, Standard deviation.

Figure 6

Increase of predictive performance as more POD information is provided. AUROC, area under receiver operating characteristic curve; POD, postoperative day.

Figure 7

Evolution of confidence score by time. DSL, duodenal stump leakage.

Figure 8

Top 30 important features. CRPDn, CRP on postoperative day n; CRP Δn, rate of decrease in CRP levels on postoperative day n; HbDn, Hemoglobin on postoperative day n; ASTDn, AST on postoperative day n; ALTDn, ALT on postoperative day n; JP_{L or R} Amylase or Lipase Δn, rate of decrease in amylase or lipase levels on postoperative day n; WBCDn, WBC on postoperative day n; WBC Δn, rate of decrease in WBC levels on postoperative day n; SD, Soft diet.