Skip to main content
EURASIP Journal on Advances in Signal Processing

Table 3 Radiologist-like segmentation (RLS) results comparison of different methods on COVID-19 radiologist dataset

From: A deep adversarial model for segmentation-assisted COVID-19 diagnosis using CT images

Method

Class

PA

IOU

Mask-RCNN [52]

1

77.08%

31.42%

RLS with L1loss

1

66.43%

46.93%

RLS with MSEloss

1

76.02%

48.15%

Mask-RCNN [52]

2

46.74%

24.66%

RLS with L1loss

2

42.42%

26.82%

RLS with MSEloss

2

62.82%

32.10%

Mask-RCNN [52]

3

26.35%

13.62%

RLS with L1loss

3

2.33%

0.27%

RLS with MSEloss

3

30.48%

18.12%

Back to article page