Skip to main content

Table 2 Comparison of performance of various methods to distinguish MCI-C from MCI-NC and MCI from CN

From: A combination of 3-D discrete wavelet transform and 3-D local binary pattern for classification of mild cognitive impairment

Dataset

Model

Original Number of Features

Accuracy

Specificity

Sensitivity

Average Number of Features

MCI-C vs. MCI-NC

3D DWT

1,359,872

0.8574 ± 0.0073

0.8564 ± 0.0068

0.8555 ± 0.0099

271.8 ± 79.35

LBP-3D

262,144

0.7644 ± 0.0395

0.7497 ± 0.0291

0.8100 ± 0.0237

232.6 ± 29.28

LBP-309

309

0.8361 ± 0.0209

0.8453 ± 0.0264

0.8363 ± 0.0238

194.5 ± 59.68

LBP-20

20

0.8432 ± 0.0131

0.8416 ± 0.0093

0.8411 ± 0.0132

16 ± 2.49

3D DWT + LBP-20

140

0.8877 ± 0.0167

0.8916 ± 0.0216

0.9016 ± 0.0054

15.3 ± 2.40

MCI vs. CN

3D DWT

1,359,872

0.8834 ± 0.0072

0.8846 ± 0.0064

0.8802 ± 0.0072

235.6 ± 95.01

LBP-3D

262,144

0.7763 ± 0.0267

0.7805 ± 0.0331

0.7900 ± 0.0342

230.4 ± 3

8.29

LBP-309

309

0.8791 ± 0.0134

0.8780 ± 0.0167

0.8631 ± 0.0158

188 ± 67.42

LBP-20

20

0.8847 ± 0.0086

0.8699 ± 0.0137

0.8575 ± 0.0115

15.1 ± 2.33

3D DWT + LBP-20

140

0.9031 ± 0.0137

0.9015 ± 0.0168

0.9022 ± 0.0134

16.2 ± 1.39