Skip to main content
Fig. 2 | Cancer Imaging

Fig. 2

From: Using neighborhood gray tone difference matrix texture features on dual time point PET/CT images to differentiate malignant from benign FDG-avid solitary pulmonary nodules

Fig. 2

ROC curves of physicians’ visual interpretations with and without texture features. AUC showed that with the help of texture features, physicians performed better in differentiating malignant from benign lesions. By employing the best performance threshold, visual interpretation with reference to the texture features obtained 76.71, 90.63 and 80.91% sensitivity, specificity, and accuracy, respectively. By employing the best performance threshold, visual interpretation without reference to the texture features obtained 76.71, 75.00 and 76.20% sensitivity, specificity, and accuracy, respectively

Back to article page