Yue Su1, Lei Wang, Hua Zhang. 1. Research Center for Traditional Chinese Medicine Complexity System, Shanghai University of Traditional Chinese Medicine, Shanghai. shibingsu@yahoo.com
Abstract
OBJECTIVE: To explore the similarity between traditional Chinese medicine (TCM) syndromes and clinical symptoms and biological parameters in patients with post-hepatitis cirrhosis (PHC). METHODS: After the variants had been normalized, 4 methods for similarity analysis, i.e. method of cosine distance, correlation coefficient, D and spectral similarity, were used to analyze the similarity of clinical characteristic information (symptoms), biological parameters and TCM syndromes obtained from 279 patients with PHC. RESULTS: The corresponding similarity matrixes were used to reflect the similarity between TCM syndromes and symptoms and biological parameters respectively, results obtained by the 4 methods were basically identical. As compared with the traditional correlation coefficient analysis, the other three methods showed a higher level of matching, sensitivity, rationality in quantitative accepting-rejecting and reliability, and were more accordant with clinical practice. CONCLUSION: The resemblance between TCM syndromes and clinical information on disease obtained from similarity analysis is basically in accordance with the clinical practice, so, the technique could be taken as a method for finding the characteristics of TCM syndrome with simplified clinical parameters.
OBJECTIVE: To explore the similarity between traditional Chinese medicine (TCM) syndromes and clinical symptoms and biological parameters in patients with post-hepatitis cirrhosis (PHC). METHODS: After the variants had been normalized, 4 methods for similarity analysis, i.e. method of cosine distance, correlation coefficient, D and spectral similarity, were used to analyze the similarity of clinical characteristic information (symptoms), biological parameters and TCM syndromes obtained from 279 patients with PHC. RESULTS: The corresponding similarity matrixes were used to reflect the similarity between TCM syndromes and symptoms and biological parameters respectively, results obtained by the 4 methods were basically identical. As compared with the traditional correlation coefficient analysis, the other three methods showed a higher level of matching, sensitivity, rationality in quantitative accepting-rejecting and reliability, and were more accordant with clinical practice. CONCLUSION: The resemblance between TCM syndromes and clinical information on disease obtained from similarity analysis is basically in accordance with the clinical practice, so, the technique could be taken as a method for finding the characteristics of TCM syndrome with simplified clinical parameters.