基于甲亢临床指标的多维时间序列关联度分析

Multi-dimensional Time Series Correlation Analysis Based on Hyperthyroidism Clinical Parameters

近年来,对临床医疗数据的挖掘分析越来越热,医疗数据中包含很多有价值的信息等待被挖掘。对基于时间序列的临床指标数据流进行关联度分析,从中发现临床指标相互之间变化趋势的相关性,对于开展精准医疗具有非常重要的价值。本文将高斯混合模型运用于临床医疗数据流的关联度分析中,提出关联支持度的方法来衡量指标之间的关联关系的强弱程度。最后通过分析100多位甲亢患者临床指标数据流,计算出各个指标对的关联支持度,得出各指标相互之间关联度的强弱关系。

In recent years,clinical data mining analysis is getting more and more popular,and medical data contain a lot of valuable information that is waiting to be tapped. With clinical test indicators data time series correlation analysis,we found that relevance of the trend among clinical test indicators,which has a very important value for the conduct of precision medical. This article makes correlation analysis among clinical data which Gaussian Mixture Model is applied to,and Related Support Method has been proposed to measure the extent of the strength of the association among indicators. Finally,through the analysis of the clinical parameters data stream which is from more than 100 hyperthyroidism patients,the correlation support of all pairwise indicators is calculated and the strength of the correlation between the indexes mutual has been generated.