Table 4 Classification performance for five suture lines. In each entry, n(m) indicates n metrics from the three-view approach (including Towne view) and m represents metrics from a two-view approach.

From: Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays

Suture Line

F1

Sen

Spec

Pre

Acc

Sagittal

0.870(0.833)

0.909(0.909)

0.913(0.870)

0.833(0.769)

0.912(0.882)

Lt coronal

0.737(0.696)

0.700(0.800)

0.917(0.792)

0.778(0.615)

0.853(0.853)

Rt coronal

0.838(0.800)

0.928(0.714)

0.800(0.950)

0.765(0.909)

0.853(0.862)

Lt lambdoid

0.667(0.600)

0.750(0.750)

0.933(0.900)

0.600(0.500)

0.912(0.882)

Rt lambdoid

0.545(0.500)

0.750(0.500)

0.867(0.933)

0.428(0.500)

0.853(0.882)

Back to article page