Table 1 Model performances for targeted and broad-spectrum CPE detection.

From: A web-based artificial intelligence system for label-free virus classification and detection of cytopathic effects

ID

Virus

Cell

Accuracy (%)

F1 score

Precision

Recall

Loss/test

12

BoAHV-1

MDBK

97.94

0.9797

0.9807

0.9788

0.0586

16

BoAHV-1

Vero

90.76

0.9020

0.9119

0.8924

0.2105

17

BoGHV-4 (MOVAR 33/63)

MDBK

95.16

0.9491

0.9496

0.9486

0.1326

18

BoGHV-4 (MOVAR 33/63)

Vero

87.99

0.8728

0.9220

0.8285

0.2759

19

BoGHV-4 (DN-599)

Vero

92.99

0.9296

0.9479

0.9119

0.2507

20

BoGHV-4 (DN-599)

MDBK

100.00

1.0000

1.0000

1.0000

0.0048

21

BPIV3

Vero

98.74

0.9820

0.9920

0.9722

0.0342

22

BPIV3

MDBK

97.85

0.9752

0.9849

0.9656

0.0418

23

BAdV-1

Vero

95.79

0.9582

0.9618

0.9547

0.1425

24

BAdV-1

MDBK

97.77

0.9782

0.9787

0.9777

0.0757

25

SARS-CoV-2

Vero

96.18

0.9554

0.9502

0.9606

0.1675

Average

95.56

0.9531

0.9618

0.9446

0.1268

26

Broad-spectrum

 

95.28

0.9495

0.9509

0.9482

0.1598

Back to article page