Skip to main content

Table 3 Cross validation results from the models built with different

From: Utilizing heart rate variability to predict ICU patient outcome in traumatic brain injury

Set #

Features

Training

Testing

AUC

[95% CI]

Sensitivity

[95% CI]

Specificity

[95% CI]

AUC

[95% CI]

Sensitivity

[95% CI]

Specificity

[95% CI]

1

APACHE II

0.73

[0.7, 0.76]

0.23

[0.15, 0.31]

1

[1, 1]

0.74

[0.63, 0.84]

0.16

[0.01, 0.31]

1

[1, 1]

2

APACHE III

0.6

[0.56, 0.64]

0.07

[0.01, 0.13]

1

[0.99, 1]

0.59

[0.43, 0.75]

0

[0, 0]

0.98

[0.95, 1.01]

3

SAPS

0.73

[0.7, 0.75]

0.08

[0.02, 0.14]

0.96

[0.95, 0.98]

0.74

[0.65, 0.84]

0.04

[0, 0.12]

0.95

[0.91, 0.99]

4

APACHE II

APACHE III

0.77

[0.74, 0.79]

0.2

[0.12, 0.28]

0.96

[0.95, 0.98]

0.65

[0.55, 0.75]

0

[0, 0]

0.96

[0.92, 1]

5

APACHE II APACHE III SAPS

0.78

[0.76, 0.81]

0.23

[0.15, 0.31]

0.96

[0.95, 0.98]

0.64

[0.53, 0.75]

0

[0, 0]

0.9

[0.84, 0.96]

6

LF_Hz

HF_Hz

LF_perc

LFHF

0.8

[0.77, 0.83]

0.35

[0.26, 0.44]

0.97

[0.97, 0.98]

0.75

[0.66,0.84]

0.22

[0.08,0.35]

0.93

[0.89,0.97]

7

DistHR

DistHF_Hz

DistHF_perc

DistLF/HF

0.87

[0.85, 0.9]

0.6

[0.54, 0.66]

0.96

[0.95, 0.96]

0.76

[0.64, 0.88]

0.52

[0.34, 0.7]

0.93

[0.9, 0.97]

8

HR

DistHR

DistHF_Hz

DistVLF_perc

DistHF_perc

DistLF/HF

0.9

[0.87, 0.92]

0.68

[0.62, 0.73]

0.96

[0.96, 0.97]

0.77

[0.66, 0.89]

0.65

[0.48, 0.82]

0.92

[0.88, 0.96]

9

HR

DistHR DistHF_Hz DistLF/HF

0.86

[0.83, 0.88]

0.62

[0.56, 0.69]

0.95

[0.94, 0.96]

0.76

[0.64, 0.88]

0.5

[0.31, 0.68]

0.93

[0.89, 0.97]

  1. *Sensitivity and specificity were calculated with a probability of 0.5 from LR as a cut-off, which means when the output from LR is over than 0.5, it was classified as positive, otherwise negative