Skip to main content
Fig. 5 | BMC Cancer

Fig. 5

From: Ultrasound-based radiomics machine learning models for diagnosing cervical lymph node metastasis in patients with non-small cell lung cancer: a multicentre study

Fig. 5

The ROC curves of all models for (A) five-fold cross-validation and (B) hold-out test. The average AUCs for the five-fold cross-validation represent the mean AUC values calculated across the five individual folds. ROC, receiver operating characteristic; LR, logistic regression; RF, random forest; Combined, semantic-radiomics combined; AUC, the area under the curve

Back to article page