Skip to main content
Fig. 3 | BMC Medical Imaging

Fig. 3

From: Machine learning for predicting the risk stratification of 1–5 cm gastric gastrointestinal stromal tumors based on CT

Fig. 3

Receiver operating characteristic (ROC) curves of three models to predict the risk stratification of gastric GISTs. (A). ROC curve of three models in the training cohort. The largest area under the curve (AUC) was GBDT (0.981), followed by DT (0.883) and LR (0.815). (B). ROC curve of three models in the internal validation cohort. GBDT and LR had the equal AUC (0.815), and DT gained the smallest AUC (0.803). (C). ROC curve of three models in the external test cohort. LR gained the best AUC (0.910), followed by GBDT (0.819) and DT (0.700)

Back to article page