Skip to main content

Table 4 Odds ratio from fitting the LogitBoost logistic regression on the AMD diagnosis outcome (dataset imputing missing values with population’s mean/mode)

From: Combining macula clinical signs and patient characteristics for age-related macular degeneration diagnosis: a machine learning approach

Variable (mode)

Odds ratio

Lower 95% CI

Upper 95% CI

P-value

Age (per year older)

1.09

1.07

1.11

<0.0001

Gender (M vs F)

1.05

0.71

1.57

0.7985

Soft drusen (pos vs neg)

19.30

7.82

47.65

<0.0001

Macular scar (pos vs neg)

1.75

0.57

5.41

0.329

RPE defect/pigment mottling (pos vs neg)

2.20

1.20

4.04

0.0109

Depigmentation area (pos vs neg)

1.35

0.73

2.51

0.3349

Subretinal fluid (pos vs neg)

3.21

1.70

6.08

0.0003

Macular tickness (per unit increase)

1.00

0.99

1.00

0.139

Subretinal fibrosis (pos vs neg)

4.60

1.39

15.24

0.01245

Subretinal hemorrhage (pos vs neg)

5.91

1.49

23.42

0.01138

  1. Statistically significant p-values are reported in bold.