Skip to main content

Table 3 Predictive performances of significant variables and the final model

From: Histogram-based analysis of diffusion-weighted imaging for predicting aggressiveness in papillary thyroid carcinoma

Variable

AUC(95% CI)

Accuracy

Sensitivity

Specificity

PPV

NPV

ADC_firstorder_TotalEnergy

0.77 (0.66–0.88)

0.67

0.63

0.81

0.91

0.40

ADC_firstorder_InterquartileRange

0.68 (0.55–0.82)

0.80

0.91

0.43

0.84

0.60

ADC_firstorder_10Percentile

0.68 (0.54–0.81)

0.64

0.60

0.76

0.89

0.37

ADC_firstorder_Maximum

0.66 (0.52–0.80)

0.70

0.72

0.62

0.86

0.41

ADC_firstorder_RobustMeanAbsoluteDeviation

0.65 (0.51–0.80)

0.79

0.91

0.38

0.83

0.57

Predictive model

0.88 (0.81–0.95)

0.75

0.69

0.95

0.98

0.49