Skip to main content

Table 3 Performance evaluation of different MRI modalities on Inception-v3

From: Convolutional neural network-based magnetic resonance image differentiation of filum terminale ependymomas from schwannomas

Input

Sensitivity

Specificity

Accuracy

AUC

Kappa

T1-WI

73.5%

73.5%

73.5%

0.813

0.471

T2-WI

76.5%

85.3%

80.9%

0.912

0.618

CE-T1

88.2%

76.5%

82.4%

0.888

0.647

ALL

94.1%

79.4%

86.8%

0.928

0.735