引用本文:[点击复制]
[点击复制]
【打印本页】 【在线阅读全文】【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览 2931次   下载 71 本文二维码信息
码上扫一扫!
肝功能检验建议性报告软件临床分析管理的优势
0
()
摘要:
介绍一种方便于临床分析、鉴别诊断和分析管理的临床检验专家系统。利用人工神经网络(Artificial Neural. Networks,ANN)挖掘患者基本信息和实验室数据,进行综合统计分析,对已确诊的患者(肝炎、肝癌、肝硬化和胆囊结石)的生化检验项目进行梳理,得出相应疾病生化指标的临床阳性预测值,然后以各项生化指标为多因素变量、诊断为输出变量建立ANN预测模型。另外抽取肝功异常并已明确诊断的患者60例进行生化指标测定,利用ANN系统,综合分析预测临床符合率,同时构建对初诊病人的树状筛查程序和直观判读报告软件的开发。利用专家诊断系统对肝脏疾病预测的准确率分别为:肝炎80.0%、肝硬化86.7%、肝癌66.7%和胆囊结石73.3%,优化后的报告格式直观反映患者的病情变化。借助于建议性报告软件可快速简易地作出对肝胆系统疾病鉴别诊断,达到医疗资源优化利用的目的。
关键词:  专家诊断系统  肝功能检验  报告软件  临床阳性率
DOI:
基金项目:
The Advantages of Clinical Analysis Management Liver Laboratory Test Proposal Report
()
Abstract:
Basic information and laboratory data from patients for comprehensive statistical analysis are introduced.Based on Artificial Neural Networks(ANN),biochemical tests of definite diagnosis deseases are analyzed,such as hepatitis, hepatic carcinoma cirrhosis, cholecystolithiasis, and positive predictive value of biochemical indexes from the corresponding diseases is obtained. Finally, ANN forecasting model is established relied on multivariate variables taken from biochemical indexes and output variables taken from diagnosis. Furthermore, the predictive clinial coincidence rate of 60 patients, who are definitely diagnosed as abnormal liver function is comprehensively analyzed,and the devolopment of tree-like screening program and visual interpretation report software is constructed. Through expert diagnosis system, the forecasting accurate rate of hepatic deseases is used as follows: hepatitis 80.0%, hepatic carcinoma 66.7%, cirrhosis 86.7%, and cholecystolithiasis 73.3%. The optimal report form directly reflects disease evolution of patients. By means of proposal report software, differential diagnosis of hepatobiliary system disease could be rapidly and simplely made,and optimal utilization of medical resources is consequently achieved .
Key words:  expert diagnosis system, clinical biochemical tests, report software, clinical positive incidence

用微信扫一扫

用微信扫一扫