Skip to main content

Table 2 Summary of quantitative measures for structures for the HR case benchmarking study patient

From: Automatic segmentation of male pelvic anatomy on computed tomography images: a comparison with multiple observers in the context of a multicentre clinical trial

CTV

     

Observer

ViPlan (cm3)

Δ V iPlan Obs (%)

VObsiPlan (cm3)

DSC Obs iPlan

I Obs ObsiPlan

iPlan

54.0

-

-

-

-

H

50.2

−6.9

40.3

0.77

0.80

A

75.2

39.4

51.0

0.79

0.68

I

60.0

11.2

48.0

0.84

0.80

B

84.3

56.3

50.8

0.74

0.60

C

41.2

−23.6

35.8

0.75

0.87

D

87.0

61.2

52.1

0.74

0.60

E

78.6

45.6

46.2

0.70

0.59

F

70.8

31.2

47.6

0.76

0.67

J

158.8

194.3

54.0

0.51

0.34

K

82.5

53.0

48.8

0.72

0.59

L

62.8

16.4

47.1

0.81

0.75

Mean A-F, H-L (SD)

77.4 (30.7)

43.5 (56.8)

47.4 (5.3)

0.74 (0.09)

0.66 (0.15)

BLADDER

     

Observer

ViPlan (cm3)

Δ V Obs iPan (%)

VObsiPlan (cm3)

DSC Obs iPlan

I iPlan Obs

iPlan

160.3

-

-

-

-

H

-a

-a

-a

-a

-a

A

172.9

7.9

153.4

0.92

0.89

I

152.6

−4.8

147.6

0.94

0.97

B

165.7

3.4

151.5

0.93

0.91

C

157.8

−1.6

146.9

0.92

0.93

D

158.2

−1.3

145.8

0.92

0.92

E

164.4

2.6

152.3

0.94

0.93

F

149.5

−6.7

140.0

0.90

0.94

J

184.9

15.3

153.1

0.89

0.83

K

169.1

5.5

152.1

0.92

0.90

L

172.7

7.7

139.2

0.84

0.81

Mean A-F, H-L (SD)

164.8 (10.7)

2.8 (6.7)

148.2 (5.3)

0.91 (0.03)

0.90 (0.05)

RECTUM

     

Observer

ViPlan (cm3)

Δ V Obs iPan (%)

VObsiPlan (cm3)

DSC Obs iPlan

I Obs ObsiPlan

iPlan

62.3

-

-

-

-

H

49.6

−20.5

40.3

0.72

0.81

A

63.9

2.5

47.2

0.75

0.74

I

53.9

−13.5

45.0

0.77

0.84

B

60.4

−3.1

46.8

0.76

0.77

C

39.8

−36.2

31.1

0.61

0.78

D

58.0

−7.0

45.5

0.76

0.78

E

52.2

−16.3

46.1

0.80

0.88

F

51.2

−17.9

44.3

0.78

0.87

J

65.1

4.4

48.1

0.76

0.74

K

44.7

−28.3

35.1

0.66

0.79

L

67.6

8.4

52.2

0.80

0.77

Mean A-F, H-L (SD)

55.1 (8.8)

−11.6 (14.1)

43.8 (6.1)

0.74 (0.06)

0.80 (0.05)

  1. a Outline missing.