Table 3 Accuracies of the classification model for internal and external test sets.

From: Realistic high-resolution lateral cephalometric radiography generated by progressive growing generative adversarial network and quality evaluations

 

Real images

 + Generated images

Internal test set

Overall

0.8344

0.8493

 Each class

  Class I

0.8429

0.8556

  Class II

0.8917

0.9108

  Class III

0.9342

0.9321

External test set

Overall

0.8287

0.8619

 Each class

  Class I

0.8287

0.8619

  Class II

0.8785

0.9061

  Class III

0.9503

0.9558

  1. Training using only the real dataset is indicated by Real images and training using the real dataset with generated images is indicated by + Generated images.
Back to article page