About a year before implementing its picture archiving and communications system (PACS)

in 1999, Santa Clara Valley Medical Center (SCVMC), San Jose, Calif, took the intermediate step of introducing computed radiography (CR). This enabled the county-owned, Stanford University-affiliated facility to generate images in soft-copy format, thereby permitting radiologists and referring physicians to review radiographic scans both on workstation monitors and on film.

"We did this to give our staff a sense of what it would be like to work in a filmless environment," Deb Lopez, CRT, director of diagnostic imaging, reports. "It was a smart move because, a year later, when we brought up our enterprise-wide PACS, we had a much smoother transition than almost surely would have been the case had we elected simply to go straight to PACS from a film environment. By the time we converted to PACS, we were quite accustomed to and comfortable with soft copy."

To abet its CR strategy, SCVMC, which is the main hospital of the Santa Clara Valley Health & Hospital System, acquired not one, but three, CR processor systems from Agfa. The official name of Agfa's CR product is Agfa Diagnostic Center (ADC). First to receive an ADC at Santa Clara Valley Medical Center was the hospital's emergency department. This was followed by installation of two more ADCs in the radiology department when that department moved into its current home, a high-tech building that opened on the SCVMC main campus in 1999.

CATCHING UP

In times past, hospitals like SCVMC might have had second thoughts about using CR as an entry-level PACS for the reason that images typically lacked spatial resolution on a par with that of conventional film-screen radiography. "CR had superior contrast resolution, so it could see a wider range of tissue densities and contribute to faster diagnoses, but many radiologists and referring physicians still felt it was more of a specialty application," according to M. Ted Ciona, senior marketing manager for Agfa. Ciona, who is responsible for marketing digital projection radiography systems, adds, "Now, however, CR is catching up to and overtaking conventional film-screen radiography on the spatial resolution side. In addition, the technology has advanced to a point at which it can now generate true-size images, which makes it conveniently useful to orthopedic surgeons in the operating room. Until just a few years ago, orthopedists relying on soft-copy images to guide them in placing artificial hips, for instance, first were obliged to overlay those scans with templates in order to determine the correct size of implant needed and to map out the incision depths." Now Agfa's ADC employs a traditional cassette to capture digital input, but the cassette contains no film for storing images. Instead, it uses a phosphor plate. "The image exposures are stored on the phosphor plate; then, a laser built into the system scans the plate to retrieve those images," Ciona explains. "Next, the laser-retrieved images are fed electronically into a computer, where they can then be routed to a monitor or to a PACS, or printed on film."

Figure. Diagnostic imaging procedures performed at Santa Clara Valley Medical Center, San Jose, Calif.

ADC is sold as a ready-to-use system onto which modules can be added to increase its image storage and processing capabilities, Ciona notes. "A basic ADC includes the CR cassette, along with a Windows NT®-based PC server loaded with ADC software," he says. "We've designed our ADC systems so that a user can start with the basic configuration, then easily move up to increasingly sophisticated features and capabilities simply by plugging in more modules. There is no need to discard and replace previous investments in hardware and software just to gain enhanced functionality."

The basic system generates images only on film. Step-up modules include the software and connectors needed to generate soft copy, as well as to perform specialty tasks (such as imaging for pediatric patients). Ciona says that it is usually unnecessary to alter the physical plant in which the ADC will be installed, since the equipment requires no more space than a conventional film processor. "The basic ADC can be installed right alongside the control panel of the x-ray room," he suggests. "If output is to be read as soft copy, however, there must be Ethernet™ of at least 10 megabits per second already in place to carry the digital images."

As for consumables in a soft-copy environment, Ciona states that the only one is the phosphor plate. "Historically, plates have had to be replaced more frequently than manufacturers have predicted, although plate technology today is much improved," he says. "We estimate that, under normal use conditions, the plates will need to be replaced at most once during the lifetime of the system. They should survive between 25,000 to 30,000 cycles, depending on the model."

No special skills are required of the radiology technologists who will operate the system, Ciona reports. "We do recommend that customers assign to the system a technologist who would be considered the superuser," he says. "This superuser would be trained more extensively than any other. The superuser would become the person to go to for anyone who has questions about how to perform various tasks with the system, how to tailor images, or how to troubleshoot."

BIG PRODUCTIVITY GAINS

Once ADC arrived at SCVMC, the systems were deployed in the most strategic locations possible. "The two we have in our own department are placed in the middle of a pod of eight general radiology rooms, which makes access very convenient for the technologists," Lopez says. "When we set up the department in this new building of ours, we organized it in a work-group configuration with the ADCs at the center. We did this because we were going from 17,000 sq ft to 45,000 sq ft of floor space with no increase in staff size. That meant that we had to develop a deployment strategy that emphasized efficiency. The work-group approach allowed the technologists to use a common piece of equipment and, at the same time, to assist one another with patient care and lifting."

The introduction of the ADCs had a significant impact on SCVMC's workflow which, in turn, improved productivity. That, of course, was to be expected, Ciona states. "Each time an enterprise advances from one level of technology to the next, that enterprise can expect about a 15% increase in productivity," he says. "Going from a film-screen system to a CR system will give you that 15% gain in productivity. Exactly how does ADC influence workflow? It affects workflow at several levels, not the least of which is the ability to eliminate paper for requisitions. With the aid of bar codes, you can easily download patient demographic information from the ADC's computer."

Ciona continues, "Most important, the ADC technology enables technologists to reduce, or even eliminate, time spent away from the patient. You expose the plate and identify the plate. The digitizer is right next to the control panel; you put the plate in, and you've never left the room. By the time you've completed your second exposure, your first image is already available on the monitor, so you can see whether you positioned it correctly. Within a minute or two after the final exposure, you can release the patient."

In addition, Ciona says, "The need to repeat exposures drops to virtually nothing because of the system. That happens not only because of its ability to compensate for exposure variation, but also because the fact that it eliminates steps in the production of the image means the technologist can-without adverse consequences-devote more time to ensuring that the examination is done correctly in the first place, making sure that the patient is positioned correctly and that the modality is properly aligned. With that gain, an enterprise has the option of doing more examinations with the same amount of available resources, or of doing the same number of examinations as before, but in less time, or with a reduction in available resources."

Integrating ADC within a PACS environment can present some challenges. "If the enterprise is just starting to get into electronic imaging," Ciona says, "there will, of course, be the need for everyone to learn the technology. Training for the technologists will have to be provided so that they understand how and why CR differs from film-screen radiography and how those differences will affect the way that they perform their jobs. Training for the radiologists will have to be provided, as well, so that they will be comfortable working with images that are substantially different in composition and content than what they're accustomed to seeing in traditional film-screen images. It will take a solid month or two of training to bring everyone up to speed on CR," he advises. "If the enterprise is already well along in electronic imaging, training will have to be provided to the referring physicians, because images will be moving out on the network."

Lopez warns that users should not expect film to vanish overnight. It will not, even with a full PACS up and running. "We still print film for the operating room," she says. "Some of the surgeons still like to have a hard copy on the viewbox as well as the soft copy on the review station. Film, today, accounts for roughly 10% of our total imaging procedures, primarily due to current use of film-screen mammography. That amount is shrinking, however. The film budget dropped from about $380,000 in 1998 to just $70,000 in 2001. For that we can thank PACS-and ADC."

ADC Technical Innovation

Agfa Diagnostic Center (ADC) systems for computed radiography (CR) first reached the market in the early 1990s, starting with Europe and, about 3 years later, the United States. In 1998, US market share for ADC stood at a meager 15%. By 2000, market share had climbed to nearly 45%, making ADC the leading product in its class. One reason for this dramatic rise is that, from the beginning, ADC has employed an open-architecture design.

Scan head device illustrating position of storage phosphor plate along with a coin for size reference.

"Agfa has always taken the approach that our systems should be open to the outside world," M. Ted Ciona, senior marketing manager for Agfa, reports. "A customer who wants to be able to use the Agfa system to query a radiology information system (RIS) or hospital information systems (HIS) have found our open architecture extremely accommodating of attempts to create fluent RIS and HIS interfaces."

As Agfa sees it, however, this still is only the beginning. The company is innovating in CR processing at a dramatic pace. A good example is its forthcoming needle-based storage phosphor technology. "Image quality of this storage phosphor is equal to if not superior to the best CR detectors currently on the market," Ciona says. "This storage phosphor could be built into a system that would deliver image quality on a par with that of direct [digital capture] radiography (DR), plus the flexibility of using a cassette-based system. We could also, potentially, build this into a device that would replace the detector in a DR device with a more economical storage phosphor device."

Illustration of potential for small cassette based scan head direct capture device for bucky replacement. System height is less than 3 inches.

In another example, Agfa is developing a  scan-head technology that it hopes to put into commercial use within the next 18 months or so. "Our scan-head technology features a single-line charge-coupled device (CCD) detector and laser-diode assembly that scans the needle-based storage phosphor and displays a high-end, 17x14-in image scan in less than 5 seconds," Ciona says. "It has variable pixel sizes of down to 50 mm (half the pixel size of traditional CR). This, obviously, results in much higher resolution. Image display time and recharge times are both about 10 to 15 seconds. This would allow the development of systems capable of handling 210 plates per hour, about three times more than the fastest CR systems today can manage. The image quality would be near that of DR." Ciona continues, "The first machine we're planning to introduce with this technology will be a small, single-plate reader that is essentially a tabletop device. This will be followed soon thereafter by an entire family of products."

Agfa also is developing a direct detector panel. Here, the scan head would be built into a circuit board, permitting construction of a very thin unit. "We estimate that, with this approach, we can put the entire CR system into a space that is approximately 18x24x2 in," Ciona says. "Inside, it would have a fixed, needle-based phosphor plate. The scanning engine would be approximately 3-in wide and be positioned at one end of that plate. As soon as the exposure was made, the scanning engine would slide across the plate and shine its laser diode on the phosphor plate. The energy from the phosphor plate would then be picked up by the single-line CCD detector array during the scan across the entire 17-in plate. At the end of the scan, which would last about 5 seconds, the image would already be displayed. Then, as the scanning engine returned to its rest position, an eraser lamp built into the assembly would purge the plate along the way. By the time the assembly returned to the rest position, the plate would be ready for another exposure cycle. This entire process would take, we estimate, about 10 to 15 seconds, meaning the user would be able to produce images at the rate of one every 20 seconds."

Scan head device for reflective scanning and reading of storage phosphor plates utilizes a single line CCD array with 50 micron pixels.

Agfa expects to begin introducing systems employing this technology by the end of 2002, Ciona adds. "The cost is expected to be substantially less than that of DR detectors," he says. "The image quality will be similar to that of DR, and the fact that we can use very small pixel sizes will make these perfect for high-definition imaging applications such as digital mammography."

Once ADC arrived at SCVMC, the systems were deployed in the most s

Agfa Diagnostic Center (ADC) systems for computed radiography (CR) first reached the market in the early 1990s, starting with Europe and, about 3 years later, the United States. In 1998, US market share for ADC stood at a meager 15%. By 2000, market share had climbed to nearly 45%, making ADC the leading product in its class. One reason for this dramatic rise is that, from the beginning, ADC has employed an open-architecture design.

"Agfa has always taken the approach that our systems should be open to the outside world," M. Ted Ciona, senior marketing manager for Agfa, reports. "A customer who wants to be able to use the Agfa system to query a radiology information system (RIS) or hospital information systems (HIS) have found our open architecture extremely accommodating of attempts to create fluent RIS and HIS interfaces."

As Agfa sees it, however, this still is only the beginning. The company is innovating in CR processing at a dramatic pace. A good example is its forthcoming needle-based storage phosphor technology. "Image quality of this storage phosphor is equal to if not superior to the best CR detectors currently on the market," Ciona says. "This storage phosphor could be built into a system that would deliver image quality on a par with that of direct [digital capture] radiography (DR), plus the flexibility of using a cassette-based system. We could also, potentially, build this into a device that would replace the detector in a DR device with a more economical storage phosphor device."

In another example, Agfa is developing a  scan-head technology that it hopes to put into commercial use within the next 18 months or so. "Our scan-head technology features a single-line charge-coupled device (CCD) detector and laser-diode assembly that scans the needle-based storage phosphor and displays a high-end, 17x14-in image scan in less than 5 seconds," Ciona says. "It has variable pixel sizes of down to 50 mm (half the pixel size of traditional CR). This, obviously, results in much higher resolution. Image display time and recharge times are both about 10 to 15 seconds. This would allow the development of systems capable of handling 210 plates per hour, about three times more than the fastest CR systems today can manage. The image quality would be near that of DR." Ciona continues, "The first machine we're planning to introduce with this technology will be a small, single-plate reader that is essentially a tabletop device. This will be followed soon thereafter by an entire family of products."

Agfa also is developing a direct detector panel. Here, the scan head would be built into a circuit board, permitting construction of a very thin unit. "We estimate that, with this approach, we can put the entire CR system into a space that is approximately 18x24x2 in," Ciona says. "Inside, it would have a fixed, needle-based phosphor plate. The scanning engine would be approximately 3-in wide and be positioned at one end of that plate. As soon as the exposure was made, the scanning engine would slide across the plate and shine its laser diode on the phosphor plate. The energy from the phosphor plate would then be picked up by the single-line CCD detector array during the scan across the entire 17-in plate. At the end of the scan, which would last about 5 seconds, the image would already be displayed. Then, as the scanning engine returned to its rest position, an eraser lamp built into the assembly would purge the plate along the way. By the time the assembly returned to the rest position, the plate would be ready for another exposure cycle. This entire process would take, we estimate, about 10 to 15 seconds, meaning the user would be able to produce images at the rate of one every 20 seconds."

Agfa expects to begin introducing systems employing this technology by the end of 2002, Ciona adds. "The cost is expected to be substantially less than that of DR detectors," he says. "The image quality will be similar to that of DR, and the fact that we can use very small pixel sizes will make these perfect for high-definition imaging applications such as digital mammography."

Rich Smith is a contributing writer for Decisions in Imaging Economics.