Skip to main content

Table 3 ROC curve analysis of three single functional models in differentiation between NPC and NPH in volunteers

From: The utility of texture analysis based on quantitative synthetic magnetic resonance imaging in nasopharyngeal carcinoma: a preliminary study

Parameters

NPC

NPH

P

Model fit

AUC (95% CI)

Sensitivity%

Specificity%

Accuracy%

T1 map derived model

   

0.870

0.972 (0.948, 0.996)

88.1

95.8

88.8

T1_Median

1405.10 (1366.50, 1519.00)

1617.13 (1464.85, 1761.93)

0.000

     

T1_Range

2718.00 (2170.00, 3256.40)

2357.80 (1910.30, 3016.78)

0.041

     

T1_Total Energy (× 106)

8800.08 (5866.79, 13,639.98)

1682.82 (1083.85, 2798.88)

0.000

     

T2 map derived model

   

1.000

0.986 (0.970, 1.000)

96.6

93.7

94.4

T2_Minimum

53.70 (47.10, 59.20)

65.20 (62.93, 68.08)

0.000

     

T2_Total Energy (× 106)

27.36 (20.71, 40.30)

5.18 (2.88, 7.80)

0.000

     

PD map derived model

   

0.890

0.975 (0.952, 0.998)

89.8

95.8

90.7

PD_90th percentile

100.10 (97.7, 100.7)

101.59 (100.86, 102.12)

0.000

     

PD_Total Energy (× 106)

29.01 (21.29, 49.53)

5.12 (3.16, 7.75)

0.000

     
  1. Data was expressed as median (interquartile range)
  2. ROC, receiver operating characteristic; NPC, nasopharyngeal carcinoma; AUC, area under the curve; 95% CI, 95% confidence interval