关键词: 肝细胞癌, 细胞分化, 增强CT扫描, 纹理分析, 预测

Abstract: Objective The aim of this study was to explore feasibility of enhanced CT texture parameters in predicting pathological tumor cell differentiation in patients with single-centered hepatocellular carcinoma (HCC). Methods 86 patients with single-centered HCC were encountered in our hospital between January 2023 and March 2025, and all patients received enhanced CT scan before surgery to record average CT value, kurtosis, skewness, energy and entropy in regions of interest. Tumor cell differentiation was determined by pathological exam. Univariate and multivariate Logistic regression analyses were applied to identify the influencing factors of poorly differentiated HCC, and the receiver operating characteristic (ROC) curves were drawn to evaluate the predictive efficacy of enhanced CT texture parameters for the degree of pathological differentiation of tumor cells. Results Of the 86 patients with solitary HCC, histo-pathological examination diagnosed poorly differentiated tumors in 27 cases (31.4%) and moderately to well-differentiated tumors in 59 cases (68.6%); the average CT value, kurtosis, skewness and entropy in the poor differentiation group were(88.7±10.3), (5.1±0.6), (1.4±0.5) and (2.4±0.4), all much higher than [(75.1±8.2), (4.6±0.7), (0.9±0.3) and (1.9±0.3), respectively, P<0.05], while energy was(4.1±0.3)×10⁶, much lower than [(4.9±0.6)×10⁶,P<0.05] in moderately to well-differentiation group; multivariate Logistic regression analysis showed that kurtosis(OR=1.844, 95%CI:1.020-3.333), skewness(OR=1.684, 95%CI:1.029-2.754) and entropy(OR=1.824, 95%CI:1.091-3.048) were all the independent impacting factors for poor differentiation(P<0.05); ROC analysis demonstrated that the AUC was 0.919, with sensitivity of 88.9% and specificity of 94.9%, when CT kurtosis was combined with skewness and entropy in predicting poor differentiation of tumor cells, much superior to any single parameter doing (P<0.05). Conclusion We tentatively recommend enhanced CT texture parameters for prediction of tumor cell differentiation in patients with single-centered HCC, which needs further clinical verification.

Key words: Hepatocellular carcinoma, Cell differentiation, Enhanced CT scan, CT texture parameters, Prediction