Table 1 Confusion matrix of the trained SegNet.

From: Deep learning-based quantitative estimation of lymphedema-induced fibrosis using three-dimensional computed tomography images

Actual

Predicted

Air

Muscle/Water

Fat

Skin

Fibrosis

Air

19,362,257

0

0

9

0

Muscle/Water

601

2,848,499

17,059

977

69,293

Fat

0

33,531

1,190,847

3,354

136,657

Skin

13

6

15,746

447,883

4348

Fibrosis

0

39,396

62,347

42,085

497,700

Back to article page