A Research On The Characteristic Symptoms Of The Fei Disease Based On A Variety Of Data Analysis Methods

Objective: Recognizing the characteristic symptoms of lung diseases with the help of the statistics on the first-hand clinical cases. Methods: Visual FoxPro 6.0, as the development environment, had been used to develop a Chinese interface and related database structure. The cases of the past dynasties were inputted and a digit-structured data base was established. To obtain the characteristic symptoms of the Fei diseases, some programs were compiled in SAS 6.12. Frequency test,φcorrelation test and non-conditional logistic multivariate regression was employed to select the characteristic symptoms of the Fei. Clustering analysis of the symptoms was employed for automatically clustering. To determine the meaning of characteristic symptoms of the Fei, mamy descriptions on the clinical manifestation in the cases of the past dynasties were summarized. According to the stipulation of denomination of a scientific term, the name and definition of characteristic symptoms of the Fei were summarized. The characteristic symptoms of the Fei diseases were summaried in the paper with the help of the statistics and the knowledge of TCM. Results: 64 symptoms were appointed through frequency test, 53 symptoms which screening on the step were obtained throughφcorrelation test, 30 symptoms were obtained through non-conditional logistic multivariate regression, 11 clusters of symptoms were selected through clustering analysis finally. Conclusions: 7 groups of characteristic symptoms that repsesent 30 characteristic symptoms of the Fei were summaried:①Symptoms of respiratory in disorder.②Symptoms of surface skin.③Symptoms of nose and pharyngeal.④Symptoms of deficiency of Yin.⑤Symptoms of exuberance of Yang.⑥Symptoms of deficiency of Fluid.⑦Symptoms of water metabolism in disorder.