论文部分内容阅读
目的：探讨判别分析法在弥漫性肺疾病鉴别诊断方面的应用。方法：收集五种常见经临床和病理证实的弥漫性肺疾病１９３例，以ＨＲＣＴ观察的３５项指标为依据，采用概率型判别分析方法——最大似然判别法进行判别分析。结果：经判别分析，１９３ 例样本病例的诊断符合率为９２％ ，所追踪观察的５９例实验病例的诊断符合率为８５％ 。结论：判别分析方法能够综合多个观察指标对某类疾病诊断的贡献率，提高诊断符合率。
Objective: To explore the application of discriminant analysis in the differential diagnosis of diffuse pulmonary diseases. Methods: Ninety-five cases of diffuse pulmonary disease confirmed clinically and pathologically were collected. Based on 35 indicators observed by HRCT, the discriminant analysis was performed by the method of probability discriminant analysis - maximum likelihood. Results: The discriminant analysis showed that the diagnostic coincidence rate of the 193 cases was 92%, and the follow-up of 59 cases was 85%. Conclusion: Discriminant analysis can synthesize the contribution rate of multiple observation indicators to the diagnosis of certain diseases and improve the diagnostic coincidence rate.