Skip to main content

Table 5 Diagnostic performance and validation of prediction models

From: Prediction of osteoporosis using radiomics analysis derived from single source dual energy CT

 

Radiomics signature

Clinical-radiomics model

 
 

Training

Validation

Training

Validation

P value*

AUCa

0.914(0.892,0.931)

0.902(0.880,0.928)

0.938(0.903,0.952)

0.988(0.967,0.998)

0.643

Cutoff

1.05

0.09

-1.46

0.15

 

Accuracy (%)

86.8

86.0

88.6

94.0

 

Specificity (%)

75.0

80.0

89.1

97.2

 

Sensitivity (%)

90.7

87.5

86.4

85.7

 

NPV (%)

72.4

61.5

96.5

94.6

 

PPV (%)

91.8

94.6

65.5

92.3

 
  1. AUC Area under the receiver operating characteristics curve, NPV Negative predictive value, PPV Positive predictive value
  2. aData in parentheses are 95% confidence interval
  3. *Difference between training and validation cohorts
Back to article page

BMC Musculoskeletal Disorders

ISSN: 1471-2474

Contact us