January 29, 2004

Programming and IT career opportunities in radiology

Author: Dee-Ann LeBlanc

With the aging population in many countries and the ever-increasing
speed of medical and technological advances, it's not a huge
intellectual leap to realize that if you're either entering the work world for the
first time or considering a change in focus or occupation, you should look at the health care industry. The need for programmers and other IT people in the medical field grows every year. Today we'll look specifically at radiology, and how you can prepare yourself for a research or IT-based career in this area.

Taking the higher education road

Many existing programmers in radiology came to it through Masters and
PhD Programs, since a broad number of the continuing advances in this field come from the academic sector. Many of
these people have focused their educations on topics other than computer science
while adding a programming component to the mix. The problem for some is
predicting what specialty to aim for, since it can take five years or
longer to earn these two advanced degrees. Fortunately, careful study of
trends can help choose a promising course of action, and continuing to
watch where your area is heading can help you correct course when
needed.

For example, Steve Langer, PhD, Associate Professor of Diagnostic
Physics and Imaging Informatics at Mayo Clinic Radiology, began his
university studies in physics, taking his Masters in Nuclear Physics
instrumentation hardware and
software. His initial goal was to work in High Energy Particle Physics
(picture that big particle accelerator in Terminator III), but since
much of this work is dependent on the attitude of the current
government, and the U.S. appetite for basic physics research has been diminishing in recent years, it looked at the time like over 2,000 PhD physicists would lose their jobs before he would be able to enter the work force, so he'd have a hard time finding a suitable physics-oriented position.

Rather than risk his ability to find a job, Langer reviewed his options,
looking for related specialties he might pursue on the PhD level.
Physics addresses objects that can be relegated into three
oversimplified categories: microscopic (up to the size of a molecule),
mesoscopic (larger than a nanometer, smaller than a micrometer), and
macroscopic (larger than a micrometer). While High Energy Particle
Physics is the child of the microscopic realm, Medical Physics (the
field responsible for, among other things, developing and ensuring the
safety and effectiveness of imaging and therapy devices in medicine)
mostly addresses the mesoscopic realm, so it's of great interest to
those who are fascinated with tiny things you can't necessarily see
with the naked eye.

Langer sees other benefits to a PhD in Medical Physics as well. For one
thing, he says, "There is the plus at the end of the day that you can develop
things that actually help people." On the more pragmatic side of things,
Langer's previous experience in the realms of instrumentation and
software development benefited him in two different Medical Physics
specialties: Therapy Medical Physics (radiation as a
curative, typically), and Diagnostic Medical Physics (devices and processes that help specialists see what is
happening inside the body). With these advantages, plus the feel-good
factor, he proceeded into the world of Medical Physics.

Many PhD candidates in the sciences will tell you that
earning your PhD is not always enough, especially if your goal is to
ultimately teach and/or go into a research position. In Langer's case,
he continued his studies by accepting a postdoc (post-doctoral)
position, which is an advanced research position available only to
PhDs, and in his field is often pursued by those who may someday to start their own businesses. Others see the postdoc as a
proving ground for those who intend to work in research and development,
or become professors. Ultimately, the purpose of a postdoc seems to
depend on the particular field your focus includes.

A career beginning or change with an aim to move to
medical programming at the research level is not something you do on a
whim. However, if your ultimate goal is the job satisfaction of
developing new, more effective ways to treat and diagnose
illnesses, it's certainly hard to beat a career in this area.

But if you would rather move into the medical IT world without endless schooling,
there are other approaches available.

Taking the Industry Route

Not everyone who makes a difference in radiology programming has a Masters or PhD. According to many in this industry -- including
Paul Nagy, PhD, an Assistant Professor of Radiology as well as the
Director of the Medical College of Wisconson's Radiology Informatics
Laboratory -- medical IT is behind the rest of the IT industry by a good
ten years. Issues such as a lack of centralization between the three
major medical cultures (administration, physicians and nurses, and the
engineering and IT staff) have made it virtually impossible to bring
medical infrastructures in most hospitals up to the level of automation
and integration most businesses have already reached.

This problem is definitely one area where the industry's loss is the
programmer's gain, even though the confusion leaves career path choices here a bit muddled. You
might think the best choice would be to join one of the many
companies supplying this field, but according to Nagy much of the
programming work here is being outsourced offshore, so the
opportunities are more limited than they might first appear. Instead,
the greatest need in North America and other areas from which programming jobs are being outsourced is for programmers who have the talent to
work directly with end users (administrators, physicians, and so on) to integrate disparate systems within hospitals and clinics.

With medical professionals suffering from "change toxicity" due to the
rapid amount of progress in both their specialties and their workplaces,
anything that helps health care workers do their jobs with a minimum of
training, duplication, and fuss is worth its weight in gold.

Duplication is, in fact, its own kind of cancer in health care. In
tumors, cells replicate at a rate that is out of step with the rest of
the body's functions, while in the average hospital both data and work are
routinely replicated, wasting hours upon hours of an already overloaded
worker's time. Take, for example, the "simple" act of checking into a
hospital and having all of your information follow you through your
stay, then to your doctors after you are released. You start your hospital stay by giving all of your information at the front desk, but often need to give the same information again and again. Why? Because
the computer systems from department to department aren't integrated.

The problem is so pervasive that Nagy refers to the various IT systems
in a hospital as "Galapagos Islands" -- each one heavily isolated, with
its own procedures and policies, and very little ability to work nicely
with the other systems within the the chain.

A number of projects and standards have been considered over the years
to try to help solve this and many other problems in medical IT. For
example, there is PACS -- the Picture Archiving and Communications
System -- which is used to digitize X-Rays and other forms of
radiologic imaging. PACS covers
everything from creating the images (with a wide variety of expensive
digital photography hardware); displaying them to a great enough degree of detail
that specialists can glean the diagnostic data they need from the images;
picture storage and management; metadata stored within the JPEGs to help
keep data that accompanies the image remain with the image; and more.

While PACS in one form or another has been available since the early
1990s, only around 15% of health care institutions are currently using it, and even institutions that already use PACS aren't often using this technology nearly to
its potential. Programmers who can help smooth the transition from
today's early adopter phase to the majority use phase will be in high
demand, especially those who can do so by working with a specific medical facility to fit a solution into its -- almost certainly unique -- existing IT infrastructure,

PACS is just the tip of the iceberg as far as opportunities in this
area. However, don't expect to reap massive financial rewards if you
change over to this field. Medical IT typically pays less than IT in
other areas, and even worse, IT managers in medical institutions often have less control over what is happening on their networks
than they might have in another field. If you're looking for a major
salary boost, consider becoming a radiologist instead! Today, there are
four jobs for every trained radiologist, and as the boomer generation
continues marching toward old age, this shortage will only increase.

Into the Future

Medicine is undergoing so many simultaneous changes that it's a wonder
anything gets done these days -- especially since it's an industry that is
heavily resistant to change. Whether you choose to
make the switch to this field by going back to school and entering the
world of research and development, or by doing your part to help
improve hospital efficiency, there's plenty of room for programmers with
great ideas and the knack for making life easier for people around
them, especially since the radiology issues discussed here are only a
fraction of what this one medical niche is currently dealing with.

The additional beauty of the medical field for folks who like their source code open is that
not everything happening in this realm is proprietary. There are strong
medical open source movements gaining steam, both in the developing world where expensive proprietary solutions
are simply not affordable, and in the U.S. and other countries where health care budgets are high but getting pinched by government austerity moves. Don't worry, though. Open source solutions for many medical IT problems
are in the works -- or already in use -- all over the world, in almost every part of the health care field.

We'll cover some of the major ones -- and the career opportunities they offer -- in future articles.

Dee-Ann LeBlanc is a trainer, writer, and
course developer who's been involved with Linux since 1994. With 12
books and over 100 articles to her credit, she is pleased to be able to
track her favorite technology for a living.

Click Here!