Table 1 Characteristics of patients from our datasets.

From: Development and evaluation of an artificial intelligence system for COVID-19 diagnosis

a

# Subject (# Scan)

Non-pneumonia

CAP

Influenza-A/B

COVID-19

In total

Training cohort

1230 (1233)

666 (666)

41 (70)

751 (1294)

2688 (3263)

Test cohort

1229 (1234)

666 (668)

42 (62)

751 (1235)

2688 (3199)

CC-CCII test cohort

829 (1079)

964 (1528)

0

726 (1257)

2539 (3784)

MosMedData test cohort

254 (254)

0

0

856 (856)

1110 (1110)

In total

3562 (3818)

2296 (2862)

83 (132)

3084 (4542)

9025 (11356)

b

# Subject

Non-pneumonia

CAP

Influenza-A/B

COVID-19

 

Training cohort

Test cohort

Training cohort

Test cohort

Training cohort

Test cohort

Training cohort

Test cohort

Gender

 Male

25

34

404

424

27

27

348

369

 Female

104

95

262

244

14

15

403

382

Age

 ≤20

0

1

62

69

2

0

1

3

 21–40

122

118

116

98

5

7

139

116

 41–60

6

10

193

198

7

8

268

261

 61–80

1

0

241

243

19

20

314

335

 >80

0

0

45

50

8

7

29

36

 No record

0

0

9

10

0

0

0

0

# Stage

 1

129

129

666

664

28

27

323

341

 2

0

0

0

4

7

11

300

301

 3

0

0

0

0

1

1

103

87

 >3

0

0

0

0

5

3

25

22

  1. a Number of subjects and scans of each class. b Detailed characteristics of subjects of training and test cohort from Wuhan. Note: as clinical characteristics of publicly available databases are not available due to anonymization, Table b lists only the data from Wuhan.