o
Friday,
February 3, 12:00 p.m. GWC 487
Ming-Hung Kao
School of Mathematical and Statistical Sciences
Experimental Designs for Event-Related fMRI Studies
Abstract
Well planned experimental designs are crucially important to
successfully achieving statistical goals under practical restrictions in
event-related functional magnetic resonance imaging (ER-fMRI) studies. With
carefully determined order and onset times of mental stimuli, researchers
can acquire informative fMRI time series and render precise and valid
statistical inference about functions of our brains. However, due to the
complex nature of ER-fMRI, obtaining an optimal ER-fMRI design is an
arduous task. While attaining high statistical efficiencies is important,
the selected designs should also satisfy some psychological constraints. In
addition, the developed approaches for obtaining optimal design should be
versatile to accommodate a wide variety of experimental settings considered
in practice. In this talk, an introduction to ER-fMRI design problems will
be provided. Some of our recent studies on this emerging research area and
future research questions will also be discussed.
