Table 4 Performance of scSE combined with different classification models for the diagnosis of H. pylori infection by single endoscopic image from gastric body.

Method	Accuracy	Sensitivity	Specificity	PPV	NPV	AUC
scSE	0.88	0.93	0.80	0.89	0.86	0.86
scSE-KNN	0.89	0.93	0.81	0.90	0.87	0.87
scSE-SVM	0.83	0.90	0.71	0.85	0.80	0.81
scSE-RF	0.88	0.94	0.76	0.88	0.88	0.85
scSE-GBDT	0.84	0.91	0.72	0.86	0.81	0.82
scSE-AdaBoost	0.87	0.90	0.81	0.90	0.82	0.86
scSE-XGBoost	0.89	0.93	0.80	0.90	0.86	0.87
scSE-LGBoost	0.90	0.93	0.83	0.91	0.87	0.88
scSE-CatBoost	0.88	0.93	0.80	0.89	0.86	0.86

scSE, KNN, SVM, RF, GBDT, AdaBoost, XGBoost, LGBoost, CatBoost, PPV, NPV and AUC are short for Spatial Squeeze and Channel Excitation Block, K-Nearest Neighbor, Support Vector Machine, Random Forest, Gradient Boosting Decision Tree, Adaptive Boosting, eXtreme Gradient Boosting, Light Gradient Boosting, Categorical Boosting, Positive Predictive Value, Negative Predictive Value and Area Under the ROC curve. No differences in the performances of accuracy, sensitivity, specificity, PPV, NPV and AUC between groups.

Quick links

Search