Table 5 Performances of the using single layers and the using different layer concatenations.

From: An improved deep learning approach for detection of thyroid papillary cancer in ultrasound images

 

TPR

FPR

TNR

FNR

1

0.416

0.612

0.388

0.584

2

0.479

0.596

0.404

0.521

3

0.514

0.504

0.496

0.486

4

0.659

0.365

0.635

0.341

5

0.897

0.276

0.724

0.103

3 & 5

0.935

0.185

0.815

0.065

4 & 5

0.910

0.205

0.795

0.090

2 & 5

0.901

0.265

0.735

0.099

1 & 5

0.872

0.277

0.723

0.128

3 & 4

0.691

0.329

0.671

0.309

2 & 4

0.641

0.410

0.590

0.359

2 & 3

0.512

0.509

0.491

0.488

1 & 3

0.462

0.558

0.442

0.538

1 & 2

0.407

0.619

0.381

0.593

  1. The numbers in first column represent the convolutional layers. ‘&’ represents concatenation.
Back to article page