摘要
目的提出一种基于疗效时间序列分析的方法,探讨它在绝经综合征中医药临床疗效评价研究的应用。方法采用聚类和主成分分析方法,对考察指标进行归类并确定其权重;并基于病例、考察指标和时间等构建疗效矩阵,采用矩阵Frobenius范数作为矩阵的相似度,分别将疗效矩阵转化为疗效时间序列;采用欧氏距离对疗效时间序列进行分析。结果这一方法与临床判定结果较为吻合,并发现中医药疗效分析结果与生存质量分析结果基本一致。结论以上方法有助于为绝经综合征中医药治疗效果提供有力的科学证据,并通过指标的调整,可辅助临床专家建立较完整的中医药临床疗效评价指标体系。
Objective To propose a method based on weight number therapeutic effect-time series analysis, and to investigate its applicable in assessment of Traditional Chinese Medicine（TCM） clinical therapeutic effect on menopausal syndrome. Methods Using clustering and principal component analysis, to classify and weight the observed indexes, to establish a therapeutic effect matrix model based on cases, observed indexes and time. Changes the therapeutic effect matrix into therapeutic effect-time series with the Frobenius matrix norm as its similarity, to make an analysis of the therapeutic effect-time series with Euclidean distance. Results We found that this method was quite in accordance with clinical judged results, and that TCM therqpeutic effect and quality of life had concordance on efficacy. Conclusion The above is helpful to offer strong and scientific evidence for TCM therapeutic effect on menopausal syndrome. To establis a more complete index system of TCM clinical therapeutic effect assessment, it assists clinical experts to examine and adjudge the observed indexes.
出处
《中国卫生统计》
CSCD
北大核心
2008年第4期387-390,共4页
Chinese Journal of Health Statistics
基金
国家科技部二项“十五”攻关课题(项目编号分别为:2001BA701A19,2004BA716B02)
广东省2007年科技攻关项目(项目编号:2007B031402002)
关键词
绝经综合征
数据挖掘
疗效评价
Menopauseal syndrome
Data mining
Therapeutic effect assessment
