Table 3 Performance of the C, CT, MBR, CMBR and CTMBR models.

From: MRI radiomics independent of clinical baseline characteristics and neoadjuvant treatment modalities predicts response to neoadjuvant therapy in rectal cancer

Evaluation index

C model

CT model

MBR model

CMBR model

CTMBR model

Training set

Validation set

Training set

Validation set

Training set

Validation set

Training set

Validation set

Training set

Validation set

AUC

0.7053

0.6103

0.7512

0.6294

0.9841

0.7835

0.9891

0.7871

0.9907

0.7916

Accuracy

81.80%

66.42%

83.65%

70.11%

94.96%

92.59%

96.15%

93.19%

96.44%

93.33%

PPV

92.59%

86.02%

93.57%

88.48%

95.34%

93.45%

95.76%

93.65%

96.13%

94.02%

NPV

60.43%

40.23%

63.43%

44.37%

88.26%

89.74%

91.57%

91.03%

91.67%

91.08%

  1. AUC area under the curve, PPV positive predictive value, NPV negative predictive value.