Scoring model of MRI features for predicting proliferative hepatocellular carcinoma
10.13929/j.issn.1003-3289.2024.06.016
- VernacularTitle:MRI特征评分模型预测增殖型肝细胞癌
- Author:
Mengtian LU
1
,
2
;
Xueqin ZHANG
;
Tao ZHANG
;
Qi QU
;
Zuyi YAN
;
Chunyan GU
;
Lei XU
;
Jifeng JIANG
Author Information
1. 南通大学医学院,江苏 南通 226006
2. 南通市第三人民医院放射科,江苏 南通 226006
- Keywords:
carcinoma,hepatocellular;
magnetic resonance imaging;
liver imaging reporting and data system
- From:
Chinese Journal of Medical Imaging Technology
2024;40(6):874-879
- CountryChina
- Language:Chinese
-
Abstract:
Objective To observe the value of the scoring model of MRI features for predicting proliferative hepatocellular carcinoma(HCC).Methods Data of 241 patients with pathologically confirmed HCC,including 90 cases of proliferative HCC and 151 cases of non-proliferative HCC were analyzed retrospectively.Univariate and multivariate logistic regression were used to compare the clinical and MRI findings evaluated according to liver imaging reporting and data system version 2018 between groups.The independent predictive factors of proliferative HCC were screened,and scores were assigned according to the weight,then a scoring model was constructed.Receiver operating characteristic(ROC)curve was drawn,and the area under the curves(AUC)were calculated to assess the predictive efficacy of this model.The patients were divided into high and low proliferation risk subgroups based on the optimal score thresholds.The recurrence free survival(RFS)rates and early RFS rates were compared between groups and subgroups.Results MRI showed tumor corona enhancement,arterial phase annular hyper-enhancement,intratumoral vessels,much focus parenchymal low enhancement and irregular tumor margins were all independent predictive factors for proliferative HCC(OR=3.287,2.362,4.542,2.997,2.379,all P＜0.05),which were then were scored with 7,5,9,7 and 5,respectively,with a total score of 0-33.AUC of the obtained scoring model for predicting proliferative HCC was 0.818.Taken 9 points as the optimal score thresholds,97 cases were assigned into high proliferation subgroup and 144 into low proliferation risk subgroups).Significant differences of RFS rates and early RFS rates were found between groups and subgroups(all P＜0.05).Conclusion MRI features scoring model could effectively predict proliferative HCC.
