Table 1 Characteristics of the developmental, internal and external validation datasets for the detection of SBI

From: A deep learning system for detecting silent brain infarction and predicting stroke risk

 

Developmental

Internal

External-1

External-2

External-3

External-4

External-5

SDPP

WTHM

PRECISE

MeLODY

UKB

I-OPTA

Fundus images

1,080

484

876

5,576

76

312

254

Participants

540

242

438

2788

38

156

127

Race, n (%)

       

Chinese

540 (100.0)

242 (100.0)

438 (100.0)

2,788 (100.0)

15 (39.5)

–

–

Indian

–

–

–

–

8 (21.0)

–

–

Malay

–

–

–

–

15 (39.5)

–

–

Asian

–

–

–

–

–

4 (2.6)

–

White

–

–

–

–

–

137 (87.8)

127 (100.0)

Black

–

–

–

–

–

1 (0.6)

–

Other

–

–

–

–

–

14 (9.0)

–

Age (years)

50.1 ± 8.7

51.1 ± 7.9

61.4 ± 12.6

61.0 ± 6.6

65.6 ± 8.9

58.2 ± 9.2

71.1 ± 8.1

Male, n (%)

377 (69.8)

200 (82.6)

323 (73.7)

1,461 (52.4)

21 (55.3)

76 (48.7)

79 (62.2)

Smoking, n (%)

137 (25.4)

81 (33.5)

109 (24.9)

950 (34.1)

–

71 (45.5)

53 (41.7)

BMI (kg m−2)

24.5 ± 3.0

25.2 ± 2.9

25.3 ± 2.5

23.8 ± 3.0

26.4 ± 1.9

26.6 ± 4.1

27.0 ± 4.6

Systolic BP (mmHg)

120.9 ± 15.2

123.1 ± 14.8

130.2 ± 17.3

129.2 ± 3.0

145.8 ± 19.2

137.2 ± 19.9

145.1 ± 6.2

Diastolic BP (mmHg)

74.2 ± 10.9

76.2 ± 10.2

78.9 ± 11.0

75.2 ± 9.0

76.4 ± 8.1

81.2 ± 10.6

84.8 ± 13.8

FPG (mmol l−1)

5.3 ± 0.9

5.4 ± 1.0

6.8 ± 2.7

5.9 ± 1.6

10.5 ± 7.9

5.1 ± 1.4

–

HbA1c (%)

5.7 ± 0.6

5.8 ± 0.7

6.9 ± 1.6

5.9 ± 0.9

8.0 ± 2.2

5.4 ± 0.7

6.5 ± 1.5

TC (mmol l−1)

4.9 ± 1.1

5.0 ± 0.9

4.2 ± 1.0

5.3 ± 1.0

4.5 ± 0.7

5.8 ± 1.2

4.8 ± 0.8

TG (mmol l−1)

1.6 ± 1.2

1.8 ± 1.6

1.9 ± 1.8

1.8 ± 1.2

1.6 ± 0.9

1.7 ± 1.0

1.4 ± 0.7

HDL-C (mmol l−1)

1.3 ± 0.3

1.3 ± 0.3

1.1 ± 0.3

2.8 ± 0.8

1.2 ± 0.3

1.5 ± 0.4

1.6 ± 0.3

LDL-C (mmol l−1)

3.0 ± 0.7

3.2 ± 0.8

2.5 ± 0.9

1.4 ± 0.3

2.5 ± 0.7

3.6 ± 0.9

2.5 ± 0.9

Hypertension, n (%)

182 (33.7)

104 (43.0)

127 (29.0)

1,196 (42.9)

11 (28.9)

77 (49.4)

39 (30.7)

Antihypertensives, n (%)

109 (20.2)

68 (28.1)

87 (19.9)

2,043 (73.3)

12 (31.6)

27 (17.3)

49 (38.6)

Diabetes, n (%)

50 (9.3)

30 (12.4)

39 (8.9)

269 (9.6)

38 (100.0)

6 (3.8)

9 (7.1)

SBI, n (%)

11 (2.0)

5 (2.1)

40 (9.1)

295 (10.6)

4 (10.5)

18 (11.5)

1 (0.8)

  1. Data are presented as mean ± s.d. or n (%) as appropriate. The en dash means not available or not applicable. PRECISE, Polyvascular Evaluation for Cognitive Impairment and Vascular Events; MeLODY, the Multiethnic Lifestyle, Obesity, and Diabetes Registry in Malaysia Diabetes Registry in Malaysia cohort; UKB, UK Biobank; I-OPTA, Identification of patient-reported barriers to treatment with anti-VEGF for neovascular AMD; FPG, fasting plasma glucose; BP, blood pressure; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol.
Back to article page