Monday, December 7, 1998
4:30-5:30 pm
1003 EECS
Abstract -
Functional Magnetic Resonance Imaging (fMRI) is a relatively new
technique that allows imaging of the brain while it is
performing a task. Traditional MRI images take many minutes
to
collect but in the last 5 years or so, ultra fast
imaging methods have been developed that provide temporal
resolution down to about 1 second and spatial resolution
down to
about 1-2 millimeters. The possibility for the first time
of such high resolution noninvasive whole human-brain
imaging
has caused tremendous excitement in the Neuroscience
community.
However the problem of analysing fMRI data is a complex
one involving elements of time series, image processing and
inverse problems. In this talk we will give some background
to
the problem and describe some of our approaches to the
signal
processing issues. We illustrate with results from
a particular set of experimental data.
To link to Dr. Solo's Home Page just click here