Semiparametric and Empirical Process Methods in Oncology

肿瘤学中的半参数和经验过程方法

• 批准号: 6608179
• 负责人: MICHAEL R KOSOROK
• 金额: $ 16.9万
• 依托单位: UNIVERSITY OF WISCONSIN-MADISON
• 依托单位国家: 美国
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-07-01 至 2005-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6608179
• 关键词: clinical research; clinical trials; computer simulation; data collection methodology /evaluation; epidemiology; human data; mathematical model; method development; oncology; statistics /biometry

DESCRIPTION (provided by applicant): The long-term goal of the proposed project is to develop flexible semi parametric analysis tools for cancer research in both epidemiologic studies and clinical trials. The change in title from the previous award reflects the completion of most of the aims in the original grant (R29 CA75 142) as well as a greater emphasis on semi parametric and empirical process methods which played an increasingly important role throughout the previous grant's multivariate clinical trial research. Semi parametric models are useful in medical research because they involve both a parametric component which is usually easy to interpret and a nonparametric component which permits greater flexibility in the presence of biologic uncertainty. The three proposed interrelated aims are: (1) Develop and evaluate inferential methods for semi parametric proportional hazards frailty regression models to account for unobserved individual-level heterogeneity in time-to-event data; (2) Develop and evaluate tools for flexible semi parametric risk-factor testing in cancer studies; and (3) Create tools for computationally efficient inference for survival, quantile and hazard functions in semi parametric models. The primary statistical research methods used for these challenging and interrelated aims include empirical process and semi parametric inference techniques as well as Monte Carlo computational approaches. These aims will provide cancer researchers with a significantly improved collection of useful modeling tools.

描述(由申请人提供)：拟议项目的长期目标是为流行病学研究和临床试验中的癌症研究开发灵活的半参数分析工具。与前一个奖项相比，标题的改变反映了原始奖项(R29 CA75 142)中大部分目标的完成，以及对半参数和经验过程方法的更多重视，这些方法在前一个奖项的多元临床试验研究中发挥了越来越重要的作用。半参数模型在医学研究中很有用，因为它们既包括通常易于解释的参数分量，也包括允许在存在生物不确定性时具有更大灵活性的非参数分量。提出的三个相互关联的目标是：(1)开发和评估半参数比例风险脆弱回归模型的推断方法，以解释时间到事件数据中未观察到的个体水平的异质性；(2)开发和评估癌症研究中灵活的半参数风险因素测试工具；以及(3)创建半参数模型中生存、分位数和风险函数的计算高效推断的工具。用于这些具有挑战性和相互关联的目标的主要统计研究方法包括经验过程和半参数推理技术以及蒙特卡罗计算方法。这些目标将为癌症研究人员提供显著改进的有用建模工具集合。