Information Technology
Digital technologies push images beyond
todays boundaries
With Internet2 in the waiting
room, a smarter Marcus Welby can be at patients bedside
electronically
By Kathy
Kincade
Talk to some of the worlds leading medical informaticists about the
digital future of radiology and you hear terms like multimedia
interfaces, physician dashboards, ubiquitous
distribution, and Internet2. Probe a little further and you
begin to believe that, 10 years from now, the terms RIS and
PACS will no longer exist, primary care physicians will be doing
their own interpretations, and radiologists will have much better tans.
Advances in computer processing and telecommunications capabilities are
changing the way diagnostic images are acquired, analyzed, displayed,
distributed, and stored. No longer are image data considered golden
bytes. Instead, to facilitate the free flow of diagnostic data across the
enterprise and beyond, image data are being treated more and more like other
data. Not less importantsimply as important. This distinction is critical
to the development of multimedia medical records that can draw information from
multiple sources and display it on a single, easy-to-use desktop application
able to access any number of devices, databases, and platforms at once.
This trend toward the ubiquitous is being fueled in part by the medical
communitys growing dependence on personal computers and the Internet. In
addition, Web-based tools have enabled new compression and data-streaming
products that work within existing bandwidth and image resolution limitations to
provide high-quality images over public networks.
The evolution is having a parallel impact on workflow processes. The lower
costs associated with standard Internet transmission cannot help but shift the
focus of image management away from the pictures and onto the communication and
workflow. In addition, the ability to stream images quickly and efficiently on
demand expands the use of images in the diagnostic, collaborative realm. New
kinds of devices that delve deeper into cellular and subcellular processes are
also changing the way diseases are diagnosed and treated.
In addition to the explosion of data, there are also more ways to communicate
and share clinical information and more users clamoring for access to it.
Fortunately, advances in high-bandwidth communications make wide-area
distribution practical and affordable for hospitals and imaging centers of all
sizes. [Fig. 1]
Just two years ago, asynchronous transfer mode (ATM) networks were considered
the best choice for high-end, mission-critical data-transfer applications. In
the past year, however, gigabit Ethernet has begun to overshadow ATM in
local-area and even wide-area network configurations, and it appears that this
trend will continue, especially with the advent of 10 GigE. Some healthcare
organizations are also looking at optical networks.
The Internet has proved itself an adequate and less expensive means of
transport for medical images. As a full-scale WAN backbone for enterprise
distribution of diagnostic images and patient data, however, the Net still has
drawbacks. While compression techniques help cut transmission times, they cannot
overcome the Internets notorious bottlenecks, especially during peak usage
times. There are also concerns about reliability, security, and quality.
The traditional Internet is still bandwidth-limited and doesnt
have good quality of service (QOS), said Dr. Paul Chang, director of
radiology informatics at the University of Pittsburgh. That is why a lot
of us are holding off in truly embracing the Internet until we get to Internet2.
Not just because it is a bigger, fatter pipe, but because we want reliability
and guaranteed QOS.
Internet2 is a collaborative effort aimed at creating the next generation of
Internet tools for research and education. More than 170 universities across the
U.S. are members of Internet2, and more than 50 corporations are cooperating in
its development. The nationwide network supporting Internet2 is known as
Abilene, a high-performance backbone that operates at 2.4 Gbps, nearly 45,000
times faster than a 56 Kbps modem. The network uses 13,000 miles of fiber-optic
cable to transmit information.
Despite the hefty admission price ($50,000 per year to join and use
Internet2), several academic medical facilities are involved in Internet2
research projects through the National Library of Medicines Next
Generation Internet (NGI) program. Researchers at the University of California,
San Francisco, and UCLA are evaluating Internet2 for high-performance broadband
imaging applications, including telemammography consultations. Similarly, the
University of Pennsylvania is working with the University of Chicago, University
of Toronto, and University of North Carolina to develop the National Digital
Mammography Archive, using Internet2 technologies.
Beyond The EMR
While such behind-the-scenes advances are fundamental to making clinical data
available whenever and wherever they are needed, critical advances are also
taking place on the front end. Once all these networks and databases are in
place, how will physicians access, view, interpret, and manipulate the desired
data?
For years the pot of gold at the end of the integration rainbow has been the
electronic medical record. Vendor attempts to develop and implement a successful
EMR, however, have met with limited success, and homegrown versions of EMRs
dominate the field. But Internet-based technologies could pave the way for a
practical multimedia medical record that draws patient data and images not only
from legacy information systems but also from PACS, RIS, and even the
modalities.
Images will be an integral part of the patients EMR, said
Dr. David Channin, an assistant professor of radiology and chief of imaging
informatics at Northwestern University Medical School. The only reason
that hasnt happened yet is that, until recently, images have been
difficult to deal with. But the barriers are beginning to come down. Ultimately,
whoever uses the patients chart should be able, with appropriate
authentication and security, to access images just like any other part of the
EMR.
How the EMR is structured is also critical to its adoption. A monolithic EMR,
with departments forwarding data directly from the modalities or the point of
care into one big enterprise-wide repository is a complex and expensive
approach, according to Rik Primo, director of new business development for
Siemens Medical Solutions. Instead, departments will have local logical
databases that forward meta-data to their own repositories and a Web-enabled
front end that will allow users to access and update the EMR from anywhere.
A number of Internet-based tools are already making it possible to create a
more ubiquitous version of the EMR. Rather than impose an additional layer of
software on top of the existing IS infrastructure, the idea is to create a
smart, Web-based user interfacea kind of dashboardthat gathers
whatever data are requested from whatever resources the user is interested in
accessing and displays them in a customized format tailored to the specific
needs, specialty, and even habits of each user.
The issue is no longer any image, anywhere, anytime,
said Michael Sullivan, director of product line management and business
development for Kodak Health Imaging in Paris. It is getting the right
information to the right person with bulletproof reliability, and making sure
you have intelligent filters in there so you dont overwhelm the
user.
Such physician dashboards are already commercially available. The
Microsoft Physician Digital Dashboard allows information on a specific patient
to be pulled from various systems and displayed simultaneously on the dashboard.
Siemens Medical Solutions is marketing a browser-based, personalized portal
product called Health Enterprise Dashboard that gives physicians secure access
to clinical and financial applications and medical and business application
service provider offerings, such as scheduling, procurement, e-mail, and
wireless messaging. The HED also provides access to patient lifetime clinical
records, test results, and radiology exams.
Eventually such products will be customizable for each specialty, with smart
interfaces that tailor the system to individual users. Each time a physician
logs on, the system will remember where they went, what information they
requested, how they requested it, and for whom they requested it. It will also
anticipate such actions, remembering where images should be sent and who should
receive alerts. Such capabilities are already in place at the University of
Pittsburgh, where 10 separate Web-based interfaces tailor the universitys
dynamic transfer syntax image-streaming technology to each department.
All multimedia components will inevitably be available ubiquitously in
digital format, full fidelity, throughout the enterprise and beyond, via
Web-based technologies and the Internet, Chang said. Everyone who
needs to make medical management and policy decisions should, must, and will
have access. And the interface should be flexible and nimble enough to bend to
the needs of the user.
Patient-Centric Approach
Physicians outside of radiology are not the only users clamoring for better
access to digital medical records. There is a growing trend toward
patient-centric medical data management, which involves creating longitudinal
records that follow, or are carried by, patients throughout their lifetime.
Virtual private networks are emerging as a secure method for patient/provider
data access and sharing. VPNs work in conjunction with the Internet to bring
limited wide-area capabilities to the medical facility and give patients outside
the hospital limited access to their physicians and, in some cases, their
medical records.
There is also a push to give patients more direct control over that data. In
the next few years, industry experts envision a VPN scenario that uses the
peer-to-peer model, where the patient is the center of a virtual healthcare
delivery space. Each patient is allocated a private, secure directory on a
central server that can be accessed using a standard Web browser. Providers can
visit the site and use the information stored there for diagnosis and treatment.
Clinical notes are stored and updated as needed, creating a patient-centric
medical record.
This approach is already fairly common outside the U.S. The international Web
site www.doctorglobal.com allows patients to upload diagnostic images into their
personal health records and manage those records themselves. The U.S. is slowly
beginning to embrace this concept, and some products and services exist.
Medscape offers an online version of its Logician EMR product and a personal
health record service called AboutMyHealth, which together create an electronic
bridge between patients and providers that includes e-mail messaging, online
scheduling, and even patient access to medical charts.
In the next five to 10 years we will see more movement toward an
integrated EMR, and the bounds of that record will extend all the way to the
patient, said Aaron Waitz, chief systems architect for Kodak Health
Imaging. Right now, EMRs tend to be the purview of the hospitals and
medical professionals. But they will extend out to patient access and perhaps
even patient entry.
Unleashing The Lion
Despite the virtual nature of data, usage has long been tied to physical
forms of communicationpaper, phone lines, cables, floppy disks, and the
like. But the emerging wireless infrastructure is making data more dynamic and
communicable than ever before. So far, however, healthcare has only flirted with
the notion of integrating wireless technologies into the clinical workflow to
improve the delivery of critical information to the point of care.
But this scenario will change dramatically in the next few years as wireless
LANs become more widely deployed in hospitals, bandwidth capabilities improve,
and data security is no longer an issue. Driving this is the need for
efficiency; doctors can better use their time if they can enter patient records
on a laptop or handheld device and send the information to a centralized
database via a wireless link. And in the near future, as images become just
another part of the data chain, they will be more easily accessed via that same
link.
Satellite-based cameras have been sending megapixel images for decades
over wireless networks, Primo said. TV, another wireless imaging
application, has been ubiquitous for five decades. Now, with the cost of
wireless going down rapidly, specific medical applications for wireless imaging
will become a real possibility.
Current wireless devices, such as laptops and personal digital assistants,
must overcome some fundamental technical challenges, like battery life and
screen resolution, before they will become truly useful diagnostic tools. Such
limitations have prompted developers to look to new kinds of wireless devices
that are between a PDA and a laptop in size but still offer the front-end
capabilities of a full-size computer, with some unique display capabilities.
If you are truly mobile, where is your monitor? In todays
paradigm, it is on your laptop, Waitz said. There will be full
diagnostic resolution laptops, but in 2020, your computer will be the size of
your PDA and you are either going to have a flexible monitor that you unroll and
lay down on a table for viewing, or youll wear special eyeglasses that
will project the image directly onto your retina.
Meanwhile, efforts are under way to improve the ability to access and view
images from any wireless platform. [Fig. 2] Researchers at UCSF have been
working with LizardTech to refine medical image applications of the
companys wavelet-based software for handhelds. Other vendors are focused
on developing infrastructure to enable seamless data sharing between mobile
devices and legacy information systems, including PACS.
For many medical informaticists, the ability of wireless technologies to
untether healthcare providers from their computers is critical to moving
medicine to the next level. It isnt just about improving the ability to
move diagnostic images out of radiology and into the hands of whoever needs
them, wherever and whenever they need them. It is about taking healthcare to a
whole new plane, where new kinds of diagnostic information are made available to
both patients and providers in a more timely and comprehensive manner. [Fig.
3]
Everyone thinks the goal of PACS is to get images everywhere, but
thats only half the puzzle, Chang said. The goal of
informatics is really to recreate Marcus Welby in an environment where Marcus
Welby as a human being could not exist. The real goal is not just to disseminate
information everywhere but to eliminate isolation and create an environment of
collaboration by combining infrastructure with collaborative tools.