FINDING THE RIGHT SYSTEM DEMANDS CAREFUL THOUGHT AS TECHNOLOGY EVOLVES FOR
COMPUTED AND DIRECT RADIOGRAPHY
Role of DR hinges on efficiency and technology
By: Jane Lowers
Pure digital radiography can make imaging suites more efficient but is not yet
available for portable or bedside applications.
As radiography becomes increasingly digital, it is stretching the capacity of
PACS-based imaging departments. Larry Filipow, Ph.D., a radiation physicist and
an assistant professor of radiology at the University of Alberta, also works as
a full-time consultant on PACS implementation, radiology equipment installation,
department layout and design, and operations review. He spoke with Diagnostic
Imaging about the current and future roles of digital radiography.
DI: What advantages does direct digital radiography offer over conventional
Filipow: Direct radiography is still the new kid on the block, and there are
maybe 200 sites installed worldwide. The main advantages are better dynamic
range and contrast detectability-the ability to see more gray levels. It has
better throughput simply because it eliminates handling cassettes. If it's
running properly, one DR room could process as many patients as three
conventional film rooms. In that case, you would use one tech to perform the
imaging and one to position the patients, whereas a conventional film room
typically has one tech alone. It depends on your staffing levels, but it could
be more efficient in human resource allocation if you've got two techs doing
what was previously a patient load for three.
DI: What factors are departments considering in deciding whether to
Filipow: You can't install new technology unless it's better, faster, and
cheaper. You need to achieve at least two of those. Computed radiography systems
that were installed in the early '90s were faster but not cheap. They didn't
become cost-effective until PACS became more common and computing power and the
price of memory dropped. If you're converting to DR, you still need to achieve
two of the three. It's faster, yes, and the image quality is superior. It's not
cheaper, but it might be fast enough to overcome the cost.
The hospitals I've seen going to DR are large teaching hospitals with large
film loads and film handling problems. They also already have PACS. PACS is
almost a prerequisite. It makes sense to do a lot of ultrasound and CT with a
miniPACS, and many departments expand from there. Those modalities can be
converted relatively inexpensively because their resolution, at 512 x 512 pixels
and 256 gray scale, can be read on just about any monitor. The challenge is
chest images. To read a chest image, you're dealing with 4000 x 3000 pixels and
4000 gray levels. Not many monitors can handle that. DR is a move to make once
you've made film management on PACS a routine part of your department. A DR room
can cost as much as a PACS.
DI: What are the limitations of DR?
Filipow: In many cases, the conversion takes place when there is new capital
funding or a new building. One of the main clinical disadvantages of DR is that
the equipment typically is fixed to a table or wall, so you can't do lateral,
out-of-bucky exposures. Several vendors are starting to offer U-arms and
multiposition detectors, so that problem may disappear. For now, a DR room will
need to have CR as well, unless it is a dedicated chest room.
DR isn't portable, and there are some limits to what the detectors can do.
You've also got 16 million transistors, so even when new, they won't all work,
and you're not getting 100% of the true image. Some pixels will "drop out," like
on a laptop screen, or an entire row or column won't respond. The systems are
configured to compensate and are designed to work around this problem.
DI: What adjustments to workflow does DR require compared with CR or plain
Filipow: A lot of people go digital with PACS on all modalities before
switching over general radiography. If you're going to look at chest images,
many experts will say you need new monitors, though the jury is still out on
this. A 1600 x 1200 monitor will take care of most of your needs, but a chest
image is acquired at 4000 x 3000, so you need one that size to see the whole
thing. At the moment, those can cost more than $20,000 each. Many radiologists
want four monitors to look at chest images so they can see the PA and lateral
views of both current and previous exams. That just adds to the expense. If you
have one monitor, you can look at images one at a time and flip back and forth.
With two, you can switch from current to previous, and that seems to be the
The majority of PACS monitors are 1.5K or 2K, so either you aren't getting
the full resolution in a shrunken image, or you can see only a portion at a
time. Most radiologists want the full resolution, but it's just not
cost-effective to buy the big monitors. So their behavior is going to have to
change the most-away from the habit of full simultaneous viewing. Still, they
get the images faster in an all-digital department, and the time savings might
be 30% or more just by getting rid of film handling.
DI: How does the cost-efficiency work out?
Filipow: A department will save time on consumables, tech staff, and film
handling, but it also will have to upgrade the infrastructure of
equipment and monitors, which can be necessary every time the warranty
expires. Things change very rapidly. In an old radiography room, you had simple
tables and light boxes and the electronics were very simple but built to last.
Now the equipment is sophisticated but also complicated and designed for rapid
evolution, and a department that has just spent millions will have to budget for
upgrades. One colleague has compared PACS to jumping on a galloping horse: You
just have to hang on and hope you eventually get control.
DI: How do CR and DR compare?
Filipow: All the CR vendors are planning new technology that will rival or
supersede DR. Whereas DR plates' pixels are limited by the size of the detector
element to the 200- or 140-micron level, a CR plate picks up images essentially
on a molecular level, like film. The resolution of the image is governed by the
resolution of the readout laser and photo pickup. Potentially, then, CR can
offer higher resolution, without the loss of pixels found in DR. Still, CR
cassettes have to be handled and fed into a reader, although even that might
evolve in favor of a reader built into a bucky in a few years.
DR does appear to provide better dynamic range-more contrast range-than CR,
and radiologists by and large are very pleased with DR images. New DR technology
may also provide improved intrinsic spatial resolution, approaching that of
conventional film performance, on the order of five line pairs per mm.
DI: What about dose?
Filipow: When CR became popular back in the '80s, techs discovered that you
can overexpose the imaging plate-and hence the patient-and also underexpose it
to a certain extent, and still get diagnostic-quality images, because you can
make adjustments during digital processing. CR became popular for ICU and
portable units because it required only one exposure to get something readable.
However, the images weren't stored electronically, but printed out as film.
CR has about twice the dose of screen-film, and you have to take that into
consideration compared with the range and manipulating ability you get. The
added radiation isn't a big issue if, for a small potential risk, you get a
considerable diagnostic benefit. DR seems to be faster, and the exposure levels
may be more similar to plain film. If dose is a concern, DR may be a preference,
but that has to be put into perspective. Diagnostic imaging doses are low,
generally. CR rates may be the highest of the three, but they are not so high as
to be a cause for concern.