Skip to main content

Table 3 Synovial sheath segmentation results of Chuang’s dataset (mean ± standard deviation)

From: Segmentation of finger tendon and synovial sheath in ultrasound image using deep convolutional neural network

Methods

Clear

Fuzzy

DSC

MAD

HD

Yasnoff

CHD

DSC

MAD

HD

Yasnoff

CHD

ATASM [16]

0.87 ± 0.04

5.12 ± 1.67

–

–

–

0.88 ± 0.04

4.54 ± 1.37

–

–

–

U-Net

0.87 ± 0.07

8.10 ± 6.25

26.69 ± 34.07

0.51 ± 0.57

10.27 ± 14.45

0.88 ± 0.05

8.62 ± 4.34

19.01 ± 19.46

0.64 ± 0.39

8.77 ± 8.97

FC-DenseNet

0.89 ± 0.05

5.30 ± 4.78

18.04 ± 21.51

0.24 ± 0.39

6.17 ± 9.81

0.89 ± 0.05

4.87 ± 3.64

11.39 ± 5.10

0.23 ± 0.30

3.75 ± 3.77

DFC-DN

0.90 ± 0.04

4.00 ± 1.92

12.34 ± 4.96

0.11 ± 0.06

2.78 ± 2.91

0.90 ± 0.04

3.64 ± 1.58

10.49 ± 5.62

0.10 ± 0.05

2.51 ± 2.76

D2FC-DN

0.90 ± 0.05

3.92 ± 1.74

12.38 ± 4.90

0.12 ± 0.06

2.60 ± 3.61

0.91 ± 0.04

3.29 ± 1.43

9.53 ± 3.68

0.09 ± 0.04

1.66 ± 0.87

Back to article page

BioMedical Engineering OnLine

ISSN: 1475-925X

Contact us