论文部分内容阅读
目的 :通过 4项肿瘤标志物联合检测 ,运用人工神经网络技术 ,提高小细胞肺癌 (SCLC)与非小细胞肺癌(NSCLC)正确判别率。方法 :用放射免疫法测定了 5 1例肺癌患者血清癌胚抗原 (CEA)、糖类抗原 12 5 (CA12 5 )、促胃液素、神经元特异性烯醇化酶 (NSE)水平 ,采用人工神经网络技术 ,探讨了 4项肿瘤标志物在肺癌组织分型中的应用价值。结果 :SCLC患者促胃液素、NSE水平明显高于NSCLC患者 ,而CEA、CA12 5水平却低于非小细胞肺癌患者。人工神经网络技术在判别SCLC与NSCLC类型中 ,总的符合率为 87.5 %。结论 :该 4项肿瘤标志物联合检测在肺癌组织分型方面可为临床提供有价值的参考资料 ,同时表明人工神经网络技术在肺癌组织分型中具有一定的实用价值。
Objective : To improve the correct discrimination rate of small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC) by using four joint detection of tumor markers and using artificial neural network technology. METHODS: Serum CEA, carbohydrate 12 5 (CA 12 5 ), gastrin, and neuron-specific enolase (NSE) levels in 51 patients with lung cancer were measured using radioimmunoassay. Artificial nerves were used. Network technology explores the value of four tumor markers in lung cancer tissue typing. Results: Gastrin and NSE levels in SCLC patients were significantly higher than those in NSCLC patients, while CEA and CA12 5 levels were lower than those in NSCLC patients. Artificial neural network technology in the identification of SCLC and NSCLC types, the overall compliance rate was 87.5%. Conclusion : The combined detection of these four tumor markers can provide valuable clinical reference for lung cancer tissue classification. It also shows that artificial neural network technology has certain practical value in lung cancer tissue typing.