摘要
目的应用相似度分析方法,对亚健康状态的人群的临床症状进行特征分析,探讨各个症状间的相似程度。方法使用4种不同计算方法分析亚健康的流行病学调查数据;同时通过相似度分析提炼亚健康状态的重要临床特征。结果采用4种方法对亚健康症状的分析结果符合临床实际,精简了亚健康的症状数量。结论使用夹角余弦、相似度D、谱图相似度3种相似度分析方法比传统的相关分析准确度高,与临床实际基本相符。该方法可简化临床症状指标数量,可形成亚健康状态的某些特征要素,为亚健康状态的评价研究提供方法学参考。
Objective To analyze the symptoms of people with sub-health in order to explore the similarity among symptoms with the method for similarity analysis.Methods 4 methods were used to analyze the epidemiological data in sub-healthy people and summarize the clinical characteristics.Results The analisis results with 4 methods were accordant with clinical symptoms,simplified the quantity of sub-health symptoms.Conclusion As compared with the traditional correlation coefficient analysis,method of cosine distance,correlation coefficient,D and special similarity,showed a higher level of accuracy.The results obtained were basically matched with clinical practice.So,the technique could be taken as a method for simplifying clinical parameters and finding the typical characteristic elements.
出处
《中国卫生统计》
CSCD
北大核心
2011年第5期543-545,共3页
Chinese Journal of Health Statistics
基金
上海市教委资助项目(No.2010JW29)
国家科技支撑计划"亚健康范畴及测量标准研究"项目(No.2006BAI13B01)
上海市卫生局科研基金项目(No.2009164)
关键词
相似度分析
亚健康
症状
特征要素
Similarity analysis
Sub-health
Symptom
Characteristic element
