|
Abstract:
|
This work aims to utilize the patient population of the Alzheimer Disease Research Center (ADRC), and to identify important risk factors and biomarkers for Alzheimer's Disease (AD) through novel statistical methods. To systematically evaluate all potential risk factors that are collected in the ADRC, we need a unifying metric that does not rely on any model assumptions. We utilize a recently developed methodology by our group, which explicitly evaluates the impact of a marker on multiple competing events simultaneously without any model assumption. This novel method is then used as a screening tool to rank the importance of all potential risk factors in ADRC. The theoretical properties of our screening method will be discussed. Its practical performance will be evaluated by simulations and the application of the ADRC to identify risk factors for AD progression.
|
