Skip to main content

Table 3 Epilepsy detection performance achieved using multiscale feature extraction methods on the CHB-MIT dataset

From: Technological Vanguard: the outstanding performance of the LTY-CNN model for the early prediction of epileptic seizures

Patient id

Acc (%)

Sen (%)

Spe (%)

AUC

Pt 01

98.6

98.9

91.8

0.929

Pt 02

99.5

99.6

97.2

0.943

Pt 03

99.3

99.3

98.2

0.945

Pt 04

99.9

99.9

100.0

0.950

Pt 05

99.1

99.2

80.0

0.900

Pt 06

98.4

99.0

90.0

0.937

Pt 07

99.5

99.5

86.7

0.906

Pt 08

97.8

98.0

92.7

0.924

Pt 09

99.7

99.8

94.6

0.953

Pt 10

98.1

98.4

90.5

0.923

Pt 11

99.7

99.7

100.0

0.900

Pt 13

99.0

99.1

96.7

0.943

Pt 14

97.5

97.5

99.5

0.914

Pt 15

99.9

99.9

100.0

0.950

Pt 16

97.9

98.3

95.6

0.958

Pt 17

98.7

98.7

100.0

0.918

Pt 18

99.1

99.4

89.1

0.932

Pt 19

99.6

99.7

95.6

0.963

Pt 21

98.8

98.9

80.0

0.887

Pt 22

99.0

99.3

83.3

0.907

Pt 23

99.1

99.5

96.6

0.980

Average

99.0

99.1

93.2

0.932