| |
NSF PA/M 96-47 - November 6, 1996
Medical Imaging Is Latest Tool in Doctors' High-Tech Black Bag
Physicians are dramatically improving patient care with imaging technologies
that are changing the way we look inside ourselves. The innovations are
made possible by an array of "smart" tools and systems developed by engineers
and scientists to help diagnose, evaluate treatment choices, plan and
simulate surgery, monitor progress of operations, improve surgical precision,
minimize the invasiveness of procedures (and so speed healing) and enhance
teaching and training.
Among these medical imaging advances:
- faster and more cost-effective ways to capture, compress, store and
transmit images
- computer-aided diagnoses and screening
- 3-D virtual reality technology
- enhanced imaging for X-Rays, CT, MRI, PET and ultrasound
- position sensors and computerized registration used to integrate precise
pre-operative plans and surgery guidance systems
- improved medical image processing to extract and analyze components
- telemedicine
The National Science Foundation (NSF) encourages
partnerships between academic researchers and businesses to fuel advances
in this emerging
field of 3-D image-guided medicine. NSF awards federal funds to researchers
pursuing these advances, and supported international conferences in
1993 and 1996.
For more information, contact:
George Chartier (703) 306-1070; gchartie@nsf.gov Beth
Gaston (703) 306-1070; egaston@nsf.gov
Research Examples
NSF-SPONSORED RESEARCH ON MEDICAL IMAGING:
- SURGICAL NAVIGATION SYSTEMS. NSF is funding research
between Carnegie Mellon University and Shadyside Hospital in Pittsburgh
to
improve understanding and integration of pre-operative planners and
surgical simulator coupled with a surgical guidance and navigation
system to improve the success of total joint replacement surgery.
Anthony M. DiGioia, M.D., director of the Center for Orthopaedic Research
at Shadyside Hospital and co-director of the Center for Medical Robotics
and Computer-Assisted Surgery at Carnegie Mellon University, and collaborators
are conducting this research. He also helped coordinate an international
workshop supported by NSF in June 1996 on "Robotics and Computer Assisted
Medical Interventions."
Contact: Anthony M. DiGioia, (312) 623-2673;
digioia@cs.cmu.edu
- VIRTUAL REALITY IMAGING OF THE COLON. Colorectal
carcinoma is the second leading cause of cancer deaths in the U.S.
Studies suggest
that early detection and removal of polyps and small carcinomas save
lives. Widely used procedures for viewing the colon use X-ray images
enhanced with a barium enema, or a fiber optic endoscope. Patients
don't like either method because they're uncomfortable and inconvenient,
and can be risky and expensive. An NSF-funded research team at Wake
Forest University is developing a new approach: Virtual Colonoscopy
combines helical CT scanning, 3-D reconstruction and high-performance
computing to rapidly examine the entire colon with minimum patient
discomfort. Researchers predict this new method will considerably
lower costs and risks.
Contact: radiologist David Vining, (910) 716-9619;
dvining@rlito.medeng.bgsm.edu
- COMPARING IMAGE QUALITY. In addition to studying
signal compression and related signal processing, a Stanford University
research team
is studying applications to digital mammography and developing a method
to quantitatively compare the analog film versions of the images to
a digital image. The research is important to the Food and Drug Administration
because as new techniques are introduced, there needs to be a valid
way of comparing the efficacy of the new and old. As digital images
become more widely used for medical applications, compression and
computer enhancement techniques also need to be proven as valid before
they are used by physicians.
Contact: Bob Gray, professor and vice
chair at Stanford University, (415) 723-4001; gray@isl.stanford.edu
Also
see:http://www-isl.stanford.edu/~gray/compression.html
http://www.rose.brandeis.edu/users/mammo/digital.html
- ULTRASOUND FOR BREAST CANCER DIAGNOSIS. John Wild,
who pioneered the use of ultrasound (the beginning of harmless medical
inspection
of living, functioning tissues) for medical applications, is researching
ultrasound for finding and diagnosing early growths in breast tissue.
X-ray mammograms not only subject women to radiation, but also do
not work well for young women who tend to have fattier breast tissues.
Ultrasound is also able to positively detect smaller lumps than X-rays,
down to about 1 mm. Wild says the nipple region of the breast, where
approximately 25 percent of cancerous tumors are found, is particularly
favorable to ultrasound detection. Milk ducts in the nipple run parallel
to each other and do not give a bounceback signal; "a tumor gives
an instant flashback like a star in the sky," which makes ultrasound
suitable for rapid, mass screening.
Contact: John Wild, director of
the nonprofit Medico-Technological Research Institute, (612) 920-8279.
- ROBOTIC ASSISTED THERAPY. An NSF-funded research team at
Johns Hopkins University is developing technologies and systems for
precise, minimally invasive delivery of localized treatment for tumors
and other soft tissue lesions. The goal is to register 3-D CT images
with real-time fluoroscopy to identify and locate tumors in the liver,
and to use this information to direct a robotic needle to deliver
fast and accurate therapeutic substances to the body through the skin.
Success with this 3-D research promises to replace more costly surgery
or provide an alternative to individuals who cannot undergo traditional
surgery.
Contact: James Anderson, director of radiology research, (410) 955-3536;
jander@rad.jhu.edu
- DEFINING AN IMAGE. There are two ways to extract
or segment out information from an image: The computer can look for
outlines
and lock in on an object or a surface and "define" a region of interest
from by the outline; or, the computer can find a given point and connect
similar points, in a sense "building" or growing a homogeneous region.
Each approach has advantages and drawbacks. James Duncan and his team
are using game theory to integrate both strategies, creating a compromise
segmentation that contains the strengths of each approach. Duncan
and his team are using this method on MRIs to observe how the left
ventricle of the heart moves and to compare the structure of autistic
children's brains to normal brains.
Contact: James Duncan, director, Image Processing and Analysis Group, Yale
School of Medicine, (203) 785-2427; duncan-james@yale.edu
Also see: http://noodle.med.yale.edu/
|
|
