Table 1 Internal and external validation of the DLS in image models, clinical models and mixed models for 5-year high myopia prediction.

From: Deep learning system to predict the 5-year risk of high myopia using fundus imaging in children

Model

Input Variables

Internal Validation (School 2 and 3)

External Validation (School 1)

Threshold

AUC (C.I.)

Accuracy (C.I)

Sensitivity (C.I.)

Specificity (C.I.)

Threshold

AUC (C.I.)

Accuracy (C.I)

Sensitivity (C.I.)

Specificity (C.I.)

Image model

Fundus photo

1.94

0.93 (0.89–0.98)

0.84 (0.84–0.84)

0.92 (0.92–0.92)

0.84 (0.84–0.84)

1.94

0.93 (0.88–0.97)

0.86 (0.86–0.86)

0.88 (0.87–0.88)

0.86 (0.85–0.86)

Fundus photo + Age + Race + Gender

1.84

0.95 (0.92–0.99)

0.86 (0.86–0.86)

0.85 (0.85–0.85)

0.86 (0.86–0.86)

1.84

0.91 (0.85–0.96)

0.78 (0.78–0.78)

0.88 (0.87–0.88)

0.77 (0.77–0.78)

Clinical model

Baseline SEa

1.74

0.90 (0.84–0.96)

0.80 (0.80–0.80)

0.77 (0.77–0.77)

0.80 (0.80–0.80)

1.74

0.93 (0.87–0.98)

0.65 (0.64–0.65)

1.00 (1.00–1.00)

0.61 (0.61–0.62)

Baseline SEa + Age + Race + Gender

1.99

0.97 (0.93–0.99)

0.87 (0.87–0.87)

0.85 (0.85–0.85)

0.87 (0.87–0.87)

1.99

0.94 (0.90–0.97)

0.74 (0.74–0.74)

1.00 (1.00–1.00)

0.72 (0.71–0.72)

Baseline SEa + Age + Race + Gender + 1-year SE* progression

2.06

0.98 (0.96–0.99)

0.92 (0.92–0.92)

0.92 (0.92–0.92)

0.92 (0.92–0.92)

2.06

0.97 (0.94–0.99)

0.84 (0.83–0.84)

1.00 (1.00–1.00)

0.82 (0.82–0.82)

Mixed model

Fundus photo + Baseline SEa

1.98

0.97 (0.94–0.99)

0.93 (0.93–0.93)

0.92 (0.92–0.92)

0.93 (0.93–0.93)

1.98

0.97 (0.94–0.99)

0.82 (0.82–0.82)

1.00 (1.00–1.00)

0.80 (0.80–0.81)

Fundus photo + Baseline SEa + Age + Race + Gender

1.98

0.97 (0.95–0.99)

0.93 (0.93–0.93)

0.92 (0.92–0.92)

0.93 (0.93–0.93)

1.98

0.98 (0.96–1.00)

0.82 (0.82–0.82)

1.00 (1.00–1.00)

0.80 (0.80–0.81)

Fundus photo + Baseline SEa + Age + Race + Gender + 1-year SEa progression

2

0.99 (0.97–1.00)

0.94 (0.94–0.94)

0.92 (0.92–0.92)

0.94 (0.94–0.94)

2

0.95 (0.92–0.98)

0.88 (0.88–0.88)

0.94 (0.94–0.94)

0.88 (0.88–0.88)

  1. AUC area-under-curves, C.I. confidence interval.
  2. aSpherical equivalent (SE) was performed using cycloplegic autorefraction.