Analysis of complex survival data subject to semi-competing risks
Analysis of complex survival data subject to semi-competing risks
复制标题
DOI：
10.32657/10220/47771
复制
发表时间：
2019-03
期刊：
影响因子：
--
通讯作者：
Mengjiao Peng
中科院分区：
文献类型：
--
作者：
Mengjiao Peng
半竞争风险下复杂生存数据的分析本论文致力于半竞争风险下复杂生存数据的分析。第一部分是通过灵活的半参数建模框架分析具有半竞争风险的聚类终点。我们开发了一个非参数的最大似然估计程序，通过Monte Carlo EM算法，并建立所需的渐近性质的估计。在第二部分中，我们研究了两组半竞争风险的多个事件发生时间终点。我们提出了一种新的统计方法，通过一对Copula来解释多个依赖结构，并开发了一个伪似然估计过程，从而共同对这些数据进行建模。在第三部分中，我们提出了一种无模型的联合相关排序筛选方法，用于选择现代癌症研究中经常收集的超高维基因特征的半竞争风险数据的重要特征。已经进行了广泛的仿真研究，结果表明有前途的有限样本性能的所有提出的方法。从各种医学研究的真实的数据集的分析表明，我们提出的方法的实用性。彭梦娇
ANALYSIS OF COMPLEX SURVIVAL DATA SUBJECT TO SEMI-COMPETING RISKS This thesis is devoted to the analysis of complex survival data subject to semi-competing risks. The first part is to analyse clustered endpoints with semi-competing risks via a flexible semiparametric modelling framework. We develop a nonparametric maximum likelihood estimation procedure via a Monte Carlo EM algorithm, and establish desirable asymptotic properties of the resulting estimators. In the second part, we study multiple time-to-event endpoints subject to two sets of semi-competing risks. We propose a novel statistical approach that jointly models such data via a pair of copulas to account for multiple dependence structures, and develop a pseudo-likelihood estimation procedure. In the third part, we propose a model-free joint correlation ranking screening method to select the important features for semi-competing risks data, which are often collected with ultrahigh-dimensional gene features in modern cancer studies. Extensive simulation studies have been carried out and results show promising finite sample performance of all proposed methods. Analysis of real data sets from various medical studies demonstrate the practical utility of our proposed methods. PENG MENGJIAO