Table 2 The statistical significance of the extracted features.

From: A new strategy for the early detection of alzheimer disease stages using multifractal geometry analysis based on K-Nearest Neighbor algorithm

Image

Kaggle dataset

ADNI dataset

P value

No dementia

Very mild dementia

Mild dementia

Moderate dementia

CN

MCI

AD

Kaggle

ADNI

D1

1.69 ± 0.016

1.62 ± 0.021

1.58 ± 0.032

1.453 ± 0.018

1.82 ± 0.012

1.76 ± 0.021

1.65 ± 0.03

1e-4

2.3e-4

D2

1.596 ± 0.025

1.533 ± 0.012

1.463 ± 0.032

1.402 ± 0.028

1.74 ± 0.025

1.65 ± 0.011

1.54 ± 0.024

1e-4

6.2e-4

αmax

2.318 ± 0.055

2.613 ± 0.027

2.635 ± 0.063

2.776 ± 0.052

2.42 ± 0.032

2.62 ± 0.044

2.83 ± 0.036

7.2e-5

3e-4

f(αmax)

0.294 ± 0.078

0.19697 ± 0.018

0.224 ± 0.042

0.122 ± 0.014

0.634 ± 0.011

0.435 ± 0.063

0.235 ± 0.034

1e-4

5.2e-4

αmin

1.65 ± 0.028

1.64 ± 0.03

1.634 ± 0.034

1.601 ± 0.037

1.53 ± 0.022

1.568 ± 0.042

1.635 ± 0.027

3.2e-5

1e-4

f(αmin)

1.542 ± 0.072

1.481 ± 0.027

1.383 ± 0.097

1.37 ± 0.082

1.21 ± 0.021

1.335 ± 0.066

1.46 ± 0.036

4e-6

1e-4

Spectrum width

0.67 ± 0.062

0.796 ± 0.042

0.974 ± 0.079

1.175 ± 0.069

0.935 ± 0.071

1.124 ± 0.068

1.436 ± 0.056

5.2e-5

1e-4

α0

1.69 ± 0.016

1.702 ± 0.022

1.724 ± 0.019

1.75 ± 0.049

1.623 ± 0.032

1.733 ± 0.023

1.832 ± 0.019

2.3e-4

1e-4

Symmetrical shift

0.289 ± 0.027

0.341 ± 0.032

0.401 ± 0.035

0.442 ± 0.042

0.211 ± 0.041

0.334 ± 0.025

0.452 ± 0.015

1e-4

3.3e-4

A(mm2)

17.71 ± 0.158

15.46 ± 0.151

13.47 ± 0.148

11.31 ± 0.207

18.63 ± 0.132

16.32 ± 0.169

14.58 ± 0.201

3.6e-6

6.3e-4

Back to article page