Skip to main content

Table 2 Comparison of ACNM with existing brain extraction approaches on data set 2: BSE in Brainsuite software, BET in MRIcro software, WAT, HWA, GCUT and ROBEX using IBSR data set 2 (20 × 3.1 mm scans)

From: Brain extraction from cerebral MRI volume using a hybrid level set based active contour neighborhood model

Method

DS mean (SD)[range]

JS mean (SD)[range]

FP_Rate(%) mean (SD)[range]

FN_Rate(%) mean (SD)[range]

HD mean (SD)[range]

BSE

0.93(0.05)

0.88(0.08)

6.8(2.7)

6.3(9.5)

23.5(7.4)

[0.75-0.96]

[0.60-0.93]

[4.2-12.5]

[1.1-36.5]

[10-30]

BET

0.85(0.08)

0.75(0.11)

22.9(8.0)

9.0(11.3)

25.4(7.0)

[0.67-0.93]

[0.50-0.87]

[11.8-39.8]

[0.2-29.6]

[7-30]

WAT

0.76 (0.14)

0.64 (0.18)

18.4 (14.1)

24.5 (22.7)

_

[0.47–0.92]

[0.31–0.86]

[5.2–61.2]

[0.1–62.7]

HWA

0.78 (0.21)

0.68 (0.21)

131.2 (308.2)

1.9 (6.5)

_

[0.16–0.88]

[0.09–0.78]

[19.4–1060.2]

[0.0–28.9]

GCUT

0.85 (0.09)

0.75 (0.10)

38.3 (40.1)

0.01 (0.02)

_

[0.49–0.90]

[0.33–0.81]

[23.1–207.5]

[0.0-0.06]

ROBEX

0.96(0.08)

_

_

_

13.3(2.6)

ACNM One

0.94(0.02)

0.89(0.03)

9.14(4.6)

2.46(1.94)

12.6(5.6)

[0.91-0.96]

[0.84-0.93]

[3.6-17.5]

[0.26-8.51]

[6-25]