目的 建立一种基于外周血mRNA的原发性肝细胞癌多参数基因诊断模型。方法GeXP法检测9个基因（GPC3、HGF、ANXA1、FOS、SPAG9、HSPA1B、CXCR4、PFN1和CALR）的含量。103例早期原发性肝癌患者和54例健康对照人群用于建立诊断模型。52例早期原发性肝癌患者和34例健康对照人群用于验证模型。使用二元Logistic回归分析、判别分析、分类树和人工神经网络建立多参数基因诊断模型用于评价其诊断价值。受试者工作曲线、曲线下面积、敏感性和特异性用于诊断价值的评价。结果 9个基因的人工神经网络模型具有最好的诊断价值，其曲线下面积、敏感性和特异性分别为0.943，98%和85%。验证后，其敏感性和特异性分别为96%和86%。结论 多参数基因诊断模型的诊断价值优于单一基因诊断价值，可以作为潜在的早期原发性肝癌辅助诊断方法。

Objective We aimed to build a multi-parameter genes diagnostic model of hepatocellular carcinoma based on peripheral blood mRNA. is pressing. Methods GeXP method was used to detect the expression of 9 genes (GPC3, HGF, ANXA1, FOS, SPAG9, HSPA1B, CXCR4, PFN1 and CALR). 103 early stage of HCC patients and 54 age matched health normal control were used to build the diagnosis model. 52 early stage of HCC patients and 34 healthy people were used for validation. Logistic regression analysis, discriminant analysis, classification tree analysis and artificial neural network were used for the multi-parameter genes expression analysis method. The ROC curve, area under the curve, sensitivity and specificity were used as the diagnosis indicators. Results: Artificial neural network of the total 9 genes had the best diagnosis value, the AUC, sensitivity and specificity were separately 0.943, 98% and 85%. At last, 52 HCC patients and 34 Healthy normal controls were used for validation. The sensitivity and specificity were separately 96% and 86%. Conclusions: Multi-parameter analysis methods may increase the diagnosis value compared to the single factor, and it may be potential assistant diagnostic method.