Skip to main content

Table 1 The number of radiographs (hips) and classification from each contributing hospital

From: Application of deep learning for automated diagnosis and classification of hip dysplasia on plain radiographs

   

Non-dysplastic

 

Dysplastic

Total

 

Normal

Borderline

 

C 1

C 2

C 3

C 4

HA 1

HA 2

HA 3

Aalborg, Denmark

95

 

85

8

 

2

0

0

0

2

0

0

Boston, USA

47

 

29

10

 

7

1

0

0

7

1

0

Emmen, Netherlands

12

 

8

0

 

4

0

0

0

4

0

0

Gothenburg, Sweden

28

 

15

5

 

8

0

0

0

8

0

0

Grapewine, USA

12

 

3

4

 

5

0

0

0

5

0

0

Hartford, USA

35

 

24

8

 

3

0

0

0

3

0

0

Hvidovre, Denmark

67

 

55

9

 

3

0

0

0

3

0

0

Kanazawa, Japan

393

 

74

32

 

122

77

71

17

187

80

20

London, UK

16

 

10

5

 

1

0

0

0

1

0

0

Madrid, Spain

20

 

16

2

 

2

0

0

0

2

0

0

New York, USA

36

 

27

6

 

3

0

0

0

3

0

0

Odense, Denmark

56

 

44

8

 

4

0

0

0

4

0

0

Oslo, Norway

50

 

28

9

 

12

0

1

0

12

1

0

Plano, USA

48

 

33

10

 

5

0

0

0

5

0

0

Stanford, USA

60

 

31

19

 

10

0

0

0

10

0

0

Uddevalla, Sweden

47

 

31

11

 

5

0

0

0

5

0

0

Total

1,022

 

513

146

 

196

78

72

17

261

82

20

  1. C = Crowe, HA = Hartofilakidis
  2. All hospitals are from data source 1 except Kanazawa (data source 2 & 3)