POSTPONED: Data-driven Brain Image Analysis: Algorithms & Applications
Speaker: Vince D. Calhoun , Yale University
Date: November 9 2006
Time: 2:00PM to 3:00PM
Location: 32-D507
Host: Polina Golland, CSAIL
Contact: Polina Golland, x3005, polina@csail.mit.edu
Relevant URL: The analysis of brain imaging data is challenging since, among other
reasons, our understand of the brain is far from complete and the contrast
to noise of the data is typically quite low. In the first half of this talk,
I will review introduce the need for data driven methods in the analysis of
brain imaging data and describe some of the work we have done in ICA of
functional magnetic resonance imaging (fMRI). I will then discuss in more
detail algorithmic issues including the consistency of the estimated sources
and methods for estimating the number of sources from spatially correlated
data. Additional applications which rely heavily upon the use of data driven
methods will be discussed including the analysis of fMRI data collected
during a simulated driving experiment while participants are either sober or
intoxicated and an analysis of the 'default mode' network, a set of regions
which typically show task-uncoupled signal decreases during the performance
of a task, in healthy controls and patients with schizophrenia. Time
permitting, I will discuss in more detail the use of ICA for combining
multiple modalities. In summary, data driven analysis methods play an
important and even increasing role in hypothesis-based brain imaging
studies.
See other events that are part of Biomedical Imaging and Analysis 2006/2007
See other events happening in November 2006
