摘要
应用最大似然法和逐步判别分析法对238份出血热、急性肾炎、流脑、上感患者的资料进行了分析,建立了计量诊断表和判别式。其特点是把临床化验数值及部分临床特点结合起来,成为科学诊断依据,同时,也为电子计算机辅助诊断提供一种方法。通过组内四代观察和另外163份病例组外回代验证,出血热的诊断符合率达98.4~100%,四种疾病总诊断符合率达96.9~99.6%,对流行性出血热早期鉴别诊断提供了准确的方法。
Analysis of 238 cases of EHF, acute gIomerulonephritis, epidemic cerebrospinal meningitis and upper respiratory tract infections was made with stepwise discriminatory analysis and maximum likelihood method, a quantitative diagnostic table and discriminatory formulas were established.Laboratory tests were combined with part of clinical features, thus making diagnosis more scientific,and the application of computers available. Through verifying the 238 cases and the other 163 cases, the rates of coincidence of EHF reached 98.4-100%, the total rates 96.9-99.6%. It has prored to be an accurate diagnostic method for EHF at its early stage.
出处
《数理医药学杂志》
1992年第3期27-32,共6页
Journal of Mathematical Medicine
关键词
计量诊断
逐步判别分析
最大似然法
流行出血热/诊断
Quantitative diaqnosis
Stepwise discriminatory analysis
Maximum likelihood method
Epidemic hemorrhagic/Diagnosis