Table 2 Cohort demographics and statistics for non-gated model.

From: Automated coronary calcium scoring using deep learning with multicenter external validation

Characteristic

Retrospective non-gated coronary CT Cohort

External non-gated validation cohorts

Stanford

MESA

Training

Validation

Test

Total

Training

Validation

Test

Total

Site 1

Site 2

Site 3

Site 4

Patient sex

 M

87

11

24

122

67

11

21

99

22

53

46

61

 F

64

11

18

93

59

11

24

94

0

22

25

74

 N/Aa

0

0

0

0

37

1

1

39

0

0

0

0

No. of exams

151

22

42

215

163

23

46

232

22

75

71

135

Mean age (range [± standard deviation])

60.9 (18–80[±9.8])

60.6 (31–82[±12.1])

60.6 (47–73[±7.5])

60.8 (18–82[±9.6])

70.0 (54–111[±9.0])

65.0 (55–80[±7.3])

65.9 (55–91[±9.0])

67.8 (54–111[±9.0])

64.6 (38–76[±10.1]

62.4 (31–88[±13.3])

66.5 (43–83[±9.0])

N/A

Median time intervalb

8.0

8.2

3.3

6.2

0.0

0.0

0.0

0.0

2.7

1.0

5.6

2.4

Scanner

 GE

55

9

20

84

40

40

0

0

22

39

0

131

 Siemens

90

13

20

123

123

123

23

46

0

11

71

3

 Other

6

0

2

8

0

0

0

0

0

25

0

1

CAC Score Bucket

 I

59

12

18

89

48

6

7

61

4

22

28

52

 II

12

5

3

20

12

3

4

19

1

4

4

11

 III

35

3

5

43

38

3

14

55

3

17

13

22

 IV

24

1

10

35

34

7

13

54

4

16

14

32

 V

21

1

6

28

30

4

8

42

11

16

12

18

  1. aInformation was not available due to site specific patient privacy rules.
  2. bMedian time interval (months) between non-gated routine chest CT and gated coronary CT for each test set.