 |
Workflow is never the same after a mammography service converts from conventional analog imaging to full-field digital. Initially, at least, some breast centers find the changeover adds—rather than reduces—steps.
“In the beginning, it took us as much as 20 minutes longer to read each screening mammography study simply because digital gave us the ability to see more than we ever could with film,” says Stamatia V. Destounis, MD, staff radiologist at the Elizabeth Wende Breast Clinic in Rochester, NY, which acquired its first digital mammography system 3 years ago. “Instead of four views, hanging on an alternator, digital gave us the ability to window level, magnify, view at full resolution, and make an assortment of other helpful adjustments.”
Mainly, radiologist time is eaten up pondering the richness of detail that emerges in those digital images, she says. Reading is slowed too because of the need to repeatedly look away from the monitor in order to assess prior film images positioned on an alternator. “For the first year of digital, your priors will be the studies produced on the analog systems the digital machine was brought in to replace,” says Destounis. “And, in addition to craning your head back and forth between the alternator and the monitor, you’ve also got the extra step of physically hanging the film priors.”
Toss computer-assisted detection (CAD) into the mix, and read times can take longer still. “CAD has been well tested with analog mammography, but not so with digital,” Destounis reminds. “So, you use CAD with your digital imaging and wonder, ‘How much can I trust these marks?’ That leads you to proceed with even greater caution than before. You look twice at every image. But looking twice means you go slower.”
The Elizabeth Wende Breast Clinic owns more than one digital mammography system, and each comes from a different manufacturer. As such, images produced on the first do not necessarily match the look of images generated by the second. The same holds true among the handful of analog units still in operation. Says Destounis, “The presentation of the same breast can appear different in the images acquired by different systems, which is something the radiologist takes into account as she interprets current studies and compares those to priors. It takes extra steps and time to sort out the differences.”
BORN OF NECESSITY
The Importance of Monitor QA |
|
Proper quality-assurance (QA) testing and adjustment of the monitors used in digital mammography are important for the reason that even a minimally out-of-kilter display can serve as a drag on radiologist productivity. The biggest source of inefficiency is the presence of artifact. Artifact can manifest as pixels that are either too bright or completely missing, as random noise spikes caused by problems with the monitor’s display card, or as loss of spatial resolution resulting from changes to the workstation setup. However, anything odd that shows up in an image will cause the radiologist to pause and puzzle over the find. All fine and well if the oddity is an actual physiologic pathology. But it amounts to a detrimental waste of time if the blip on the radar, so to speak, turns out to be nothing more than a couple of dropped pixels, warns John Weiser, PhD, DABR. “Even if the radiologist knows there are missing pixels and compensates by moving the image off to one side or the other, that’s still an extra step that has to be taken,” says Weiser, medical physicist at Xtria Healthcare, Digital Imaging Solutions, in Frederick, Md. “The goal is to have excellent QA so that the radiologist has the ability to do the best job possible with the fewest steps required. Monitor QA is important for all types of studies, but especially so for digital mammography because mammograms make for such complicated reading to begin with—often, in early detection screening exams, the patient has no symptoms, which means the radiologist has to make a determination from sometimes very, very subtle signs of disease process.” The monitors used in digital mammography are specifically designed to meet the quality standards required by the federal Food and Drug Administration for that particular type of imaging, Weiser adds. “A mammography monitor must have a screen resolution of at least 5 megapixels and a white level of at least 500 candelas per square meter,” he says. Either CRT or flat panel LCD-type monitors can be used in mammography, although LCD is the more popular of the two. “Flat panel LCD monitors take up less space and generate less heat—which offer certain environmental advantages in the reading room,” Weiser notes. The American Association of Physicists in Medicine (AAPM) recently came out with a report detailing the different types of tests that should be performed on digital mammography monitors, and Weiser advocates for their adoption by mammography services. “The AAPM in conjunction with its recommendations has produced a set of artifact-revealing test images, which several monitor vendors have already incorporated into their software,” he says. “You can bring up these images, evaluate them, and record the results; the software produces a report for you with regard to grayscale accuracy, resolution, and anything else that could affect monitor display quality.” To assist in the QA process, software is available (usually sold with the monitor or workstation) to automatically maintain monitor brightness. Says Weiser, “It checks backlighting and adjusts brightness accordingly. As long as the monitor has been calibrated to the DICOM display standard, the software will see to it that the calibration remains locked in.” Remote monitoring software is also commercially available that allows a technologist or a physicist sitting at a console elsewhere in the facility to keep tabs on the QA status of each monitor. “That person can remotely verify that the white level is where it is supposed to be, that the monitor is functioning properly,” says Weiser. The danger of using QA software is that it can lull one into a false sense of security. “It encourages a set-it-and-forget-it frame of mind,” Weiser contends. “Only by keeping a steady visual watch over a monitor can you be assured of detecting and being able to address QA problems, such as artifact. Software cannot tell whether a pixel has dropped out partially or entirely. The only way this can be determined is by having a technologist or a physicist go in and bring up certain test images on a regular basis in accordance with manufacturer recommendations.” |
| —R. Smith |
After a few months of working in the newly digital environment, mammography radiologists gain proficiency and confidence sufficient to boost their reading speeds back up to where they were in the analog days. “For instance, you soon learn which images don’t need window level or magnification,” says Destounis.
However, digital offers the ability to improve radiologist productivity beyond anything possible with film. To realize that potential, the Elizabeth Wende Breast Clinic has sought to develop formal protocols and standards for digital mammography based on workflow initiatives originating with the Society for Computer Applications in Radiology, the Radiological Society of North America, and the Integrating the Healthcare Enterprise initiative. Radiologists also participate in weekly meetings during which digital mammography workflow is a standing topic of discussion. “It’s largely a matter of educating one another as we together try to find the best solutions for the problems we encounter,” says Destounis. “And, if a problem comes up in the normal course of daily business—such as we’re not sure what this particular CAD mark here means or why CAD isn’t coming up at all on the image—we just walk down the hall and solicit ideas from one another on how to solve it.”
Many of the process improvements at the Elizabeth Wende Breast Clinic were born of necessity. Take, as one illustration, case annotations—a cumbersome task that originally required the radiologist (or a clerk) to hand-deliver to the technologist a printout of the image for which extra views were sought. “It’s not particularly inefficient or economically wasteful if annotation is required only once in a while, but it is on both counts if annotation requests are coming in at a rate of 20 a day as they were for us,” says Destounis, explaining that the problem was solved by installing a digital workstation in the technologist’s room so that images in question could be instantaneously relayed from point to point with just a couple of clicks of the radiologist’s mouse.
TRIAGE OF PATIENTS
The majority of service revolves around mammography screening examinations—about 250 to 280 per day. These are routinely double-read, as are images produced for diagnostic assessment of discharges, lumps, or calcifications. Within 3 years, Destounis expects the entirety of her enterprise’s breast imaging work to be performed digitally.
“Digital is inevitable,” she says, adding that the clinic plans to attain that status through the attrition of its older analog cameras. “We’re replacing our film systems with digital as the film systems age and die off. It’s a slow process, but that’s the only way a small, freestanding imaging center can manage the costs.”
Importantly, the Elizabeth Wende Breast Clinic has determined that the expedited workflow of digital will be such that it need not replace each of its analog systems one-for-one with a digital product. “A single digital unit can handle the caseload of two analog systems,” Destounis says. “That fact helps us offset some of the cost of our digital purchases.”
Being a mix of digital and analog systems presents some interesting workflow challenges in and of itself. One of them arises from the matter of deciding which patients to image digitally and which to image on film—an issue of no small consequence in light of recently reported findings by the Digital Mammographic Imaging Screening Trial (DMIST), which indicates that digital mammography screening reliably detects 14% to 27% more cases of cancer in women under the age of 50, in women with denser breasts, and in women who are pre- and peri-menopausal. Destounis says that her enterprise has embraced the DMIST findings, and now routinely sends women meeting those criteria to the digital procedure room instead of randomly assigning them to whichever room happens to be open, digital or analog.
Also predesignated to receive digital mammograms are those patients whose prior-year examination was likewise digital. “Once we do a digital mammogram on a patient, we see it as very important to continue doing them that way from then on,” Destounis conveys. “It just makes workflow so much easier if you’re comparing digital to digital when you look at current and prior exams. We don’t want to take a step backward by doing digital one year and then film the next.”
Fewer than 10% of the clinic’s patients specifically request digital mammography. Sometimes, when they do, it is on the advice of a doctor or at the behest of an informed friend or relative. But, usually, they ask for digital after having read about it in newspapers and consumer-oriented magazines or seen it demonstrated on television. “If we were in a major metropolitan market like New York or Los Angeles, the number of women specifying digital mammography would be much greater,” Destounis says. “The message about digital mammography hasn’t yet saturated our market. Our local media don’t talk as frequently as they do in the bigger markets about new digital mammography techniques.”
CART BEFORE THE HORSE
The temporary hamstringing of workflow that often develops with the introduction of digital mammography can be minimized by taking a proactive approach, such as the one employed at Straub Clinic and Hospital in Honolulu, an affiliate of Hawaii Pacific Health, the largest health care provider in Hawaii . Straub has yet to install its first digital mammography system., but when the day arrives, it will already have at least 1 year’s worth of digitized film priors in hand.
“We currently produce about 30 mammography studies a day at our main campus and another 30 combined at our half-dozen satellites; these studies are entirely analog and are sent through a CAD system that includes built-in digitizer technology,” says Robert Lipman, MD, staff radiologist. “We are in the process of building a new clinic, and it’s designed to accommodate full-field digital mammography. That’s where we will install our first such system. We acquired the CAD-digitizer system in advance of digital mammography so that we could have an archive of digitized priors to immediately begin working with once the new clinic opens. This will allow us to right from the start make digital-to-digital comparisons of current and prior images, rather than have to be looking at a monitor for the new studies and a film alternator for the old ones. We felt this would be a more efficient way to work. We’re not going to be in the position of having to manually verify that the prior films we’re looking at are the correct ones and that they are hung on the alternator in the proper order. We’re also not going to have the problem of lights on for viewing film priors, which produces suboptimal conditions for viewing digital images on the workstation monitor.”
Another advantage of digitizing images ahead of installing a digital mammography system is the elimination of film handling. “Rather than send films from place to place, we wanted to have digital access to all images from just one place,” says Lipman, noting that the digitized views are stored on an enterprise picture archiving and communications system (PACS).
The economics of this approach were quite attractive, and that is why Straub Clinic and Hospital decided to go forward with it. “We could have waited until we actually had our digital mammography system up and running before we invested in the CAD-digitizer technology,” says Lipman. “That would have enabled us to hold onto some capital for a year or two longer. But the CAD component of the product allowed us to generate extra income and thereby leverage the investment. In other words, we could begin digital archiving of our filmed mammograms, and make a profit at the same time.”
No matter how one proceeds, the fact remains that digital mammography changes everything. In some regards, productivity will immediately rise, while in others it will just as swiftly decline. The trick is to be able to manage those changes so that the positive is accentuated while the negative is minimized. As the experiences of some mammography enterprises suggest, it is by no means a trick easily performed, but a trick that must nonetheless be performed.
Rich Smith is a contributing writer for Decisions in Axis Imaging News.
