Table 1 Baseline characteristics of the datasets

From: LensAge index as a deep learning-based biological age for self-monitoring the risks of age-related diseases and mortality

 

Traditional slit-lamp images

Smartphone images

 

Reference dataset

Analysis dataset

Reference dataset

Analysis dataset

No. of participants

1990

3433

50

102

Nationality, n (%)

Chinese

1952 (98.1%)

3370 (98.2%)

50 (100%)

99 (97.1%)

Non-Chinese

38 (1.9%)

63 (1.8%)

0

3 (2.9%)

Age in years (mean ± s.d.)

55.3 ± 18.0

66.0 ± 11.5

64.6 ± 11.4

62.0 ± 10.8

Distribution of chronological age, n (%)

≥ 20 and < 30

245 (12.3%)

8 (0.2%)

0

0

≥ 30 and < 40

231 (11.6%)

55 (1.6%)

3 (6.0%)

4 (3.9%)

≥ 40 and < 50

246 (12.4%)

224 (6.5%)

0

6 (5.9%)

≥ 50 and < 60

292 (14.7%)

624 (18.2%)

15 (30.0%)

37 (36.3%)

≥ 60 and < 70

507 (25.5%)

1139 (33.2%)

12 (24.0%)

28 (27.5%)

≥ 70 and < 80

334 (16.8%)

976 (28.4%)

18 (36.0%)

23 (22.5%)

≥ 80

134 (6.7%)

407 (11.9%)

2 (4.0%)

4 (3.9%)

Sex, n (%)

Male

732 (36.8%)

1482 (43.2%)

20 (40.0%)

45 (44.1%)

Female

1258 (63.2%)

1951 (56.8%)

30 (60.0%)

57 (55.9%)

Images, n

Diffuse-light images

4542

5641

N/A

N/A

Slit-lamp images

3713

5663

157

389