Skip to main content

Table 3 Results of nodule classification at wavelet first decomposition level

From: Automated system for lung nodules classification based on wavelet feature descriptor and support vector machine

Base

Angle

Sub-band

Features

Specificity

Sensitivity

Preciseness

Db1

LH

Autc-Clsh

82.60%

63.63%

73.33%

Db2

LH

Clsh-IMC2

86.95%

45.45%

66.66%

Db4

HH

Autc-Diss

60.86%

95.45%

77.77%

Db1

45°

LH

Autc-Clsh

86.95%

59.09%

73.33%

Db2

45°

LH

Autc-IMC2

69.56%

68.18%

68.88%

Db4

45°

LH

Autc-IMC2

69.56%

90.90%

80%

Db1

90°

LH

Clpr-Clsh

73.91%

90.90%

82.22%

Db2

90°

LH

Clpr-Svar

82.60%

59.09%

71.11%

Db4

90°

HH

Autc-Diffv

52.17%

100%

75.55%

Db1

135°

LH

Clpr-Sent

86.95%

72.72%

80%

Db2

135°

LH

Clsh-Diffe

86.95%

72.72%

80%

Db4

135°

HH

Autc-Cont

65.21%

90.90%

77.77%

  1. The bold data represent the best value obtained.