Skip to main content

Table 3 The percentages (%) of cases for each muscle auto segmented by DLAS and ABAS which were worse than that achieved by physicians (mean DSC was used to compare the results)

From: Deep learning vs. atlas-based models for fast auto-segmentation of the masticatory muscles on head and neck CT images

 

M-R

M-L

T-R

T-L

LP-R

LP-L

MP-R

MP-L

DLAS

62.1%

51.7%

20.7%

24.1%

65.5%

65.5%

44.8%

37.9%

(18/29)

(15/29)

(6/29)

(7/29)

(19/29)

(19/29)

(13/29)

(11/29)

ABAS

96.6%

89.7%

48.3%

41.4%

96.6%

82.8%

79.3%

69.0%

(28/29)

(26/29)

(14/29)

(12/29)

(28/29)

(24/29)

(23/29)

(20/29)

P value*

0.02

0.03

0.05

0.26

0.01

0.23

0.01

0.03

  1. * represents Chi-Square test was performed between DLAS and ABAS