To Establish an Early Prediction Model for Acute Respiratory Distress Syndrome in Severe Acute Pancreatitis Using Machine Learning Algorithm

Abstract

Objective: To develop binary and quaternary classification prediction models in patients with severe acute pancreatitis (SAP) using machine learning methods, so that doctors can evaluate the risk of patients with acute respiratory distress syndrome (ARDS) and severe ARDS at an early stage.

Methods: A retrospective study was conducted on SAP patients hospitalized in our hospital from August 2017 to August 2022. Logical Regression (LR), Random Forest (RF), Support Vector Machine (SVM), Decision Tree (DT), and eXtreme Gradient Boosting (XGB) were used to build the binary classification prediction model of ARDS. Shapley Additive explanations (SHAP) values were used to interpret the machine learning model, and the model was optimized according to the interpretability results of SHAP values. Combined with the optimized characteristic variables, four-class classification models, including RF, SVM, DT, XGB, and Artificial Neural Network (ANN), were constructed to predict mild, moderate, and severe ARDS, and the prediction effects of each model were compared.

Results: The XGB model showed the best effect (AUC = 0.84) in the prediction of binary classification (ARDS or non-ARDS). According to SHAP values, the prediction model of ARDS severity was constructed with four characteristic variables (PaO₂/FiO₂, APACHE II, SOFA, AMY). Among them, the overall prediction accuracy of ANN is 86%, which is the best.

Conclusions: Machine learning has a good effect in predicting the occurrence and severity of ARDS in SAP patients. It can also provide a valuable tool for doctors to make clinical decisions.

Keywords: acute respiratory distress syndrome; acute severe pancreatitis; artificial neural network; machine learning.

Figure 1

ROC curves for the five different models in the test set.

Figure 2

Ranking of feature importance in the Random Forest (RF) and eXtreme Gradient Boosting (XGB) models, combined with shap value. (A): Ranking of feature importance in the RF model; (B): Ranking of feature importance in the XGB model.

Figure 3

Receiver Operator Characteristic (ROC) curves of the five models after optimization.

Figure 4

SHAP summary plot to determine the importance ranking of the seven features that have the greatest impact on the RF and XGB models. (A): SHAP summary plot of RF model; (B): SHAP summary plot of XGB model.