Skip to main content

Table 5 Prediction of unfavourable outcome using the optimal cutoffs of the ADC and the percent ADC

From: Diffusion-weighted magnetic resonance imaging for predicting the clinical outcome of comatose survivors after cardiac arrest: a cohort study

Grey matter structures

Optimal cutoff

Sensitivity with 95% CI

Specificity with 95% CI

PPV with 95% CI

NPV with 95% CI

 

ADC

Percent ADC**

    

Frontal cortex

0.685

79

73% (45 to 85%)

100% (56 to 100%)

100% (68 to 100%)

64% (32 to 88%)

Parietal cortex*

0.674

77

87% (58 to 98%)

100% (56 to 100%)

100% (72 to 100%)

78% (40 to 96%)

Temporal cortex

0.640

69

67% (39 to 87%)

100% (56 to 100%)

100% (66 to 100%)

58% (29 to 84%)

Occipital cortex*

0.740

82

93% (66 to 100%)

100% (56 to 100%)

100% (73 to 100%)

88% (47 to 99%)

Precentral cortex*

0.606

84

87% (58 to 98%)

100% (56 to 100%)

100% (72 to 100%)

78% (40 to 96%)

Postcentral cortex

0.625

86

73% (45 to 91%)

100% (56 to 100%)

100% (68 to 100%)

64% (32 to 88%)

Caudate nucleus

0.621

76

67% (39 to 87%)

100% (56 to 100%)

100% (66 to 100%)

58% (29 to 84%)

Putamen*

0.590

75

93% (66 to 100%)

100% (56 to 100%)

100% (73 to 100%)

88% (47 to 99%)

Thalamus*

0.647

85

87% (58 to 98%)

100% (56 to 100%)

100% (72 to 100%)

78% (40 to 96%)

  1. ADC unit × 10-3 mm2/sec; PPV, positive predictive value; NPV, negative predictive value
  2. Optimal cutoff values predicting unfavourable outcome were determined by ROC curve analysis in patients and controls.
  3. *Area under the ROC curve was greater than 0.9 with a P value less than 0.001.
  4. **Percent ADC expressed the (absolute ADC value/mean normal control ADC value) × 100.