Fig. 2: Performance of the NCLS and radiologists.
From: Deep learning approach for screening neonatal cerebral lesions on ultrasound in China
a ROC curve of the NCLS and the AUC of each radiologist. b The diagnostic performance of NCLS and radiologists on specific conditions. For cases with normal, ependymal cyst, ventricular dilation, and grade I or II IVH, predicting the outcome as non-severe is considered a correct prediction. For other conditions, the reverse applies. The left column corresponds to the internal test set, while the right column corresponds to the external test set. In the confusion matrix, the specific numbers represent the count of the diagnostic outcomes of the model or radiologists for a specific condition. Source data are provided as a Source data file. AUC area of curve, NCLS neonatal cerebral lesions screening, PVL periventricular leukomalacia.
