IEEE/EMBS Summer School'96
Diagnostic Medical Imaging and the Big Picture
Diagnostic Medical Imaging and the Big Picture
Michael W. Vannier, MD
University of Iowa
Iowa City, IA USA
Business Week recently announced "The software revolution - The Web changes everything"
(December 4, 1995 cover story), and Byte's cover says, "Toss your TV - How the Internet
will replace broadcasting" (February 1996). We are in the midst of a revolution in computer and telecommunications technology that will profoundly affect our lives
- personally and professionally.
The need to quickly process and transmit complex images is now the driving force in
computer and telecommunication hardware and software developments. It has been said
that bandwidth, which provides the ability to transmit complex images quickly, is
the single most important characteristic determining the functionality of future computers.
And this future is notfar off.
The Internet will determine what the future will bring. A new generation of applications
designed to run on the Internet don't care what type of computer you have. All that
matters is how fast your computer can process and transmit information within the
network. Operating systems as we know them, such as Windows 95 or Mac OS or Unix, don't
matter in this environment. Bandwidth is the key!
Bandwidth to support all these computer functions is directly proportional to the
transmission speeds and is the driving force behind the rapid adoption of ATM or
A synchronous Transfer Mode standard for telecommunications. The ability to process
and transfer complex images atATM speeds will fundamentally affect organizations and how they
operate. In radiology, a moderate speed ATM link allows delivery of CT or MRI scan
images in a fraction of a second as compared with the standard ethernet widely used
today. At ATM speeds, CD sound, telephone, broadcast-quality video and full resolution
medical images can be delivered simultaneously anywhere on the network. The same
link can provide all of these services at once at a cost that makes its use practical.
While ATM to one's home may be several years away, interim solutions may be provided
by both local telephone and cable television providers. Modems which connect to
existing cable television networks will likely become widely available during the
next year. This ever increasing bandwidth will provide access to informational sources which
are currently limited to the low transmission capacity of standard telephone lines.
The future of computing will feature a migration to object-based programs retrieved
as network-based applets. Today we're locked into hardware so programs are written
for a specific "platform", such as the PC or Macintosh. Every two years or so, application and operating systems developers announce upgrades that utilize new features of
the then current hardware. We clamor to buy and install these upgrades such as Windows
95, forcing us to replace computers that still work perfectly well. The distribution
of software by floppy disks and CD ROMs adds to the cost and burden that computer
users must assume to access improvements or new programs.
In using program applets, whenever a specific application is needed it is downloaded
from the network and runs on the user's computer. For instanceif an x-ray object
is retrieved for after hours interpretation, an applet will be used to view the image
on the radiologist's home network terminal. Sun Microsystems has developed the Java programming
language which makes it all possible. Java operates on a "virtual machine", implemented
as a compact program that could reside in any computer including very modest and inexpensive ones. Once it's present, the computer can run any Java applet that
comes across the Network. In other words, to run a program, part is resident in your
machine and the rest resides somewhere else on the Internet. When the program is
finished, it will simply vanish, eliminating the need for hard disks, CD ROMs, floppy disks,
and other local storage devices. Given the anticipated distribution of applets rather
than multimegabyte applications, Sun Microsystems, Oracle, and other companies will introduce small, inexpensive computers selling for a few hundred dollars that access
The functionality that is now present on your desktop or in the electronic reading
room will be available literally everywhere. In the two years since the Web and the
Mosaic program for viewing its pages emerged from research labs, the Web has turned
into a huge virtual disk drive. For those who don't know, the Web refers to the World Wide
Web, a graphical subnetwork within the Internet. The technical details are unimportant
to most users who can navigate the Internet interactively using a mouse with point
and click commands that are easily understood and applied. It is likely that current
Web browsers, such as Netscape will serve as one of the display platforms for these
In the future, we will see hardware-independent standards (the ACR-NEMA standard is
a good example) that allow programs to run on any platform, regardless of the manufacturer.
Software on demand, like pay-per-view video, will be available. Object technology and applets allow you to purchase or even rent software as you need it. You can
try it out before you buy it. Like home shopping on television, you can select the
items you want after seeing them on a CRT, request and pay for them without leaving
your desk or home.
Network applets, intelligent agents, distributed objects, multidata bases and ever
increasing bandwidth will replace "bloatware" operating systems and applications,
as well as the hard drives required to store them. Greater bandwidth will benefit
Radiology by facilitating access to images, especially when compared to today's premium cost
of teleradiology services. Applets and network agents will be developed to assist
in the distribution and interpretation of images and communication of diagnostic
results to referring clinicians at the time and place where decisions on treatment and prognosis
must be made.
Back to the Summer School home page.