论文部分内容阅读
目的利用人工计算模型建立起对脂肪肝的辅助诊断的新方法。方法基于对甘肃省某电厂196名职工的血糖、总胆固醇、低密度脂蛋白胆固醇、高密度脂蛋白胆固醇、甘油三酯的检测指标,采用基因表达式编程(GEP)方法对是否是脂肪肝的人群进行分类判别。结果 GEP方法获得的特异度和敏感度分别为99%和67%,准确率为95.92%。阳性预测值(PPV)为99.33%,阴性预测值(NPV)为96.13%。结论 GEP方法建立的诊断模型可用于脂肪肝的辅助诊断。
Objective To establish a new method of auxiliary diagnosis of fatty liver by artificial calculation model. Methods Based on the detection of blood glucose, total cholesterol, low density lipoprotein cholesterol, high density lipoprotein cholesterol and triglyceride in 196 workers in a power plant in Gansu Province, gene expression programming (GEP) Crowd classification discrimination. Results The specificity and sensitivity of the GEP method were 99% and 67%, respectively, with an accuracy of 95.92%. The positive predictive value (PPV) was 99.33% and the negative predictive value (NPV) was 96.13%. Conclusion The diagnostic model established by GEP method can be used to diagnose fatty liver.