Medical Research
University of Florida
Increasing medical understanding through signal processing
Reproduced with permission from NeXT
Computer, Inc.
A Reference
Guide to NeXT in Higher Education, Fall 1992
ยช
1992 NeXT Computer, Inc
For more than 14 years, Jose Principe, an associate
professor of electrical engineering at the University of
Florida, has explored biomedical engineering and signal
processing. He is particularly interested in research
involving the analysis of electroencephalograms
(EEGs)-tracings that show changes in electrical potential
caused by brain activity.
Working with 14 graduate students in a lab of networked
NeXT computers, Principe developed several interrelated
projects that explore the uses of digital signal processing
to medicine. Principe and the students rely primarily on
Mathematica and the Motorola DSP56001 digital
signal processor.
"The Mathematica environment is excellent for
prototyping, and it offers good plotting facilities," says
Principe. "Biomedical signals are concentrated at low
frequency, and you can do a lot of processing with the
DSP-even real-time processing. Along with the NeXT user
interface, this combination is what I've always wanted in a
computer.
In one project, Principe and graduate student Haan-go Choi
used a NeXT machine to build the Signal Editor, a software
package created for visualizing all types of electrical
signals-from EEGs to machine tool data to audio waves
(music).
"With the Signal Editor," says Principe, "the computer
becomes a scope, or, actually, more like a microscope,
because the computer offers very accurate time and
amplitude measurements of signals."
Working with the Signal Editor, a user can import a
previously collected signal file, zoom in on specific
portions, and measure amplitude/duration and slope. "The
Signal Editor has display capability, but there's more,"
says Principe. "It's structured so that we can create our
own signal processing functions and then visualize
graphically the results of the processing on the computer
screen. For instance, you can apply various digital filters
with different parameters and compare the results of the
filtering. So it offers an efficient way of visually
comparing signal processing algorithms."
According to Principe, the Signal Editor and
Mathematica can be used by medical professionals
to pinpoint the focus of epilepsy in the brain. Prior to
brain surgery, a doctor can place a two-dimensional grid of
electrodes on the exposed cerebral cortex of epileptics
before brain surgery. The Signal Editor can then compute a
spatial map of the brain's electrical field. Working with
Mathematica, the Signal Editor creates a graphic
representation that may indicate the region of epileptic
focus. This information provides neurosurgeons with the
ability to plan their incisions more precisely when
removing the affected brain tissue.
"My idea is to have this device in the operating room and
to work closely with the neurosurgeon, providing a visual
indication of the most probable focus of epileptic
activity," says Principe. "Of course, it cannot be done in
real time yet, but that is a goal of the project. We're
still developing the algorithm and validating the method,
but with the power of DSP, we will be able to do it in real
time."
For more information, please contact:
Jose Principe
Associate Professor of Electrical Engineering
University of Florida
405 CSE
University of Florida
Gainesville, Florida 32611
principe@brain.ee.ufl.edu