Abstract

The proper design of an fMRI experiment can involve many factors that are critical to its success. In this talk, I will focus on optimizing the statistical efficiency of designs. After reviewing the framework of the general linear model, I will introduce metrics for statistical efficiency and discuss how these metrics
strongly depend on the underlying assumptions that an investigator must inevitably make. I will also review methods for optimizing statistical efficiency, such as the use of m-sequences and genetic algorithms.