论文部分内容阅读
目的探讨福建省肝癌发病的主要危险因素;建立肝癌危险因素神经网络预测模型。方法采用病例-对照研究,用问卷进行调查。用logistic回归分析筛选出主要危险因素,建立肝癌危险因素神经网络模型。结果感染乙型肝炎、经常食用腌制食品、霉变食品处理后食用等6项,是肝癌的危险因素;吃饭时间≥10min为保护因素;神经网络预测结果显示:对照组符合率为90.4%,病例组符合率为92.1%,总符合率为91.3%。结论神经网络建立的模型对肝癌具有良好的预测作用。
Objective To explore the main risk factors of liver cancer in Fujian Province and to establish a neural network prediction model of liver cancer risk factors. Methods A case-control study was conducted with a questionnaire. Logistic regression analysis was used to screen out the main risk factors and establish a neural network model of risk factors for liver cancer. Results Hepatitis B infection, regular consumption of preserved foods and consumption of mildewed foods were the risk factors for liver cancer. The time of eating ≥10 minutes was the protective factor. The neural network prediction showed that the coincidence rate of control group was 90.4% The coincidence rate of case group was 92.1% and the total coincidence rate was 91.3%. Conclusion The model established by neural network has a good predictive value for liver cancer.