Table 2 Clinicopathological information of 350 HCC cases in TCGA dataset classified by the tumor ploidy determined using the HIPT2_unfrozen2 model

From: Selective identification of polyploid hepatocellular carcinomas with poor prognosis by artificial intelligence-based pathological image recognition

Variables

Diploid

Polyploid

p value

Number of cases

123

227

-

 

Age

62 ± 12.5

60 ± 13.4

0.35

1

Sex (Proportion of male)

72.4

65.4

0.18

2

Tumor stage

T1

75

98

0.002

4

T2 / T2a / T2b

23

65

T3 / T3a / T3b

22

52

T4

1

11

N.D.

2

1

-

 

LN metastasis

N0/N1/N.D.

89/0/34

154/3/70

0.48

2

Distant metastasis

M0/M1/N.D.

93/1/29

162/2/63

1

2

Pathological stage

Stage I

72

92

0.005

4

Stage II

23

60

Stage III / IIIA / IIIB / IIIC

22

56

Stage IV / IVA / IVB

1

4

Tumor grade

G1/G2/G3/G4/N.D.

21/69/29/3/1

25/99/89/10/4

0.001

4

Child-Pugh grade

A/B/N.D.

97/8/18

112/13/102

0.47

2

Vascular invasion

Macro/Micro/None/N.D.

6/25/85/7

9/63/111/44

0.07

4

Fibrosis

0 - No Fibrosis

42

31

0.26

4

1,2 - Portal Fibrosis

9

19

3,4 - Fibrous Speta

14

14

5 - Nodular Formation and Incomplete Cirrhosis

3

5

6 - Established Cirrhosis

30

35

N.D.

25

123

-

 

Serum prothrombin time (sec)

1.1 ± 4.89

1.1 ± 5.05

0.294

3

Serum alpha fetoprotein (ng/mL)

9 ± 14255

23 ± 162348

0.040

3

Serum albumin (g/dL)

4.05 ± 5.5

4 ± 1.03

0.502

3

Platelet (Ă—103/μL)

212.5 ± 61667

211 ± 86405.6

0.675

3

Serum total bilirubin (mg/dL)

0.7 ± 1.49

0.7 ± 1.63

0.549

3

  1. HE images of 350 HCC cases in TCGA dataset were analyzed using the HIPT2_unfrozen2 model (cutoff of polyploidy score = 0.363).
  2. 1 Student t test, 2 Pearson’s χ2 test, 3 Wilcoxon-Mann-Whitney U test, 4 Cochran-Armitage trend test. N.D. not determined.
Back to article page