Skip to main content

Table 2 Statistical test results in comparing our image-based method with voxel-based method

From: Automated identification of cell-type-specific genes in the mouse brain by image computing of expression patterns

Folds

Sections

A vs. Neg.

N vs. Neg.

O vs. Neg.

O vs. A

A vs. N

N vs. O

1.5

Coronal

0.0822

1.7E-06

4.7E-06

0.0036

1.7E-06

1.7E-06

 

Sagittal

1.7E-06

8.5E-06

0.1306

2.9E-6

1.7E-06

2.1E-06

 

Cor.+Sag.

3.5E-06

1.2E-05

0.0017

1.7E-06

2.6E-06

1.7E-06

10

Coronal

6.6E-04

1.7E-06

0.9263

9.3E-06

8.7E-05

0.7343

 

Sagittal

0.0558

1.7E-06

5.5E-4

0.0916

0.0180

0.0052

 

Cor.+Sag.

0.0387

1.1E-05

0.5038

0.1086

0.3389

0.4908

20

Coronal

0.0612

1.9E-06

0.9590

0.0001

0.7188

0.0100

 

Sagittal

0.0387

0.0026

0.0157

2.7E-5

5.7E-6

0.0614

 

Cor.+Sag.

0.6435

9.7E-05

0.0114

4.0E-4

0.3359

0.0349

  1. We employed two-sided Wilcoxon signed rank tests on the AUC values produced by 30 random trials, and the p-values were reported. We also performed the one-sided statistical test to compare the mean of image-based multiple trials with that of voxel-based method. The bold values indicate tasks on which image-based method outperforms voxel-based method significantly.