Deep learning-based radiomics based on contrast-enhanced ultrasound predicts early recurrence and survival outcome in hepatocellular carcinoma

Table 2 Univariate and multivariable analyses of early recurrence of hepatocellular carcinoma patients

	Univariate cox regression				Multivariate logistic regression
	HR	[0.025	0.975]	P	HR	[0.025	0.975]	P
Age, yr	0.991	0.971	1.011	0.383
Gender	1.005	0.463	2.181	0.990
HBsAg-positive	1.138	0.472	2.740	0.773
AFP	0.622	0.361	1.071	0.087	0.709	0.406	1.239	0.227
CEA	0.944	0.352	2.535	0.910
CA125	1.811	0.424	7.752	0.423
CA19-9	1.636	0.727	3.684	0.234
ALT, U/L	1.248	0.697	2.321	0.267
AST, U/L	1.566	0.397	2.108	0.675
FIB-4 score	1.212	0.431	1.986	0.742
Liver cirrhosis	1.142	0.506	2.121	0.657
Tumor size x	1.000	0.882	1.143	0.951
Tumor size y	0.988	0.856	1.142	0.873
Gray-scale echogenicity	0.731	0.493	1.087	0.121
Arterial phase enhancement	0.924	0.351	2.438	0.874
Portal phase enhancement	1.321	0.832	2.100	0.238
Portal phase enhancement	0.982	0.512	1.885	0.957
Enhancing capsules	0.930	0.413	2.096	0.862
Satellite nodules	4.843	1.917	12.244	0.001	4.194	1.368	12.871	0.012
Unsmooth margins	0.839	0.462	1.522	0.563
Constant					0.990	0.707	1.387	0.953

Fibrosis-4 score: Age (years) × aspartate aminotransferase (U/L) / [platelet count (10⁹/L) × alanine aminotransferase (U/L)]^1/2. HR: Hazard ratio; AFP: Alpha-fetoprotein; CEA: Carcinoembryonic antigen; CA125: Carbohydrate antigen 125; CA19-9: Carbohydrate antigen 19-9; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; FIB-4 score: Fibrosis-4 score; HBsAg: Hepatitis B surface antigen.