Skip to main content
. 2022 Jan 27;2022:1473977. doi: 10.1155/2022/1473977

Table 4.

Performance of human experts and deep learning during classification of orthodontic images in the external dataset.

Operator Method Accuracy Macro-AUC Time taken to archive (min)
Deep learning Automatic 0.994 1.00 0.08
Expert 1 Manual 0.988 0.992 19.19
Deep learning assistance 0.998 0.997 8.27
Expert 2 Manual 0.987 0.985 18.97
Deep learning assistance 0.997 0.996 7.92
Expert 3 Manual 0.983 0.983 18.63
Deep learning assistance 0.996 0.996 8.10

AUC: area under the curve.