While there is no clear consensus that the use of full-field digital mammography (FFDM) improves mammography outcomes, many practices and institutions across the United States are not waiting for consensus and have purchased FFDM, or plan to purchase the technology in the near future. Some cite marketing pressure, others the allure of the “filmless” department, and still others simply want to stay at what they consider the leading edge of breast imaging. Whatever the reasoning, FFDM purchases are accelerating.

The implementation of FFDM in a breast imaging practice or center is not a simple one-to-one replacement of equipment. Rather, it is the adoption of a technology that will alter work flow at every level, from scheduling, through the delivery of the mammogram, to the interpretation and rendering of the report, and on to billing and collecting for the service. The author believes that successful implementation of this technology will largely be the result of understanding its challenges, and carefully and comprehensively planning and preparing for DdigitalDay.

Think Image

Before proceeding, it is important to begin thinking in terms of the mammographic image. Mammographic images have heretofore been captured on film. FFDM and other digital technologies (eg, computed radiography-mammography) create that image in a digital format, rather than on film, producing many of the advantages of digital, as well as many of its challenges.

Imagine a system with no film. No processor, no messy developer, no waiting for films, no lost films, no film library, no film library staffand that is just on the technical side. From the interpretive perspective, FFDM promises the ability to optimize and manipulate the image, create digital magnification views, and otherwise use the tremendous postprocessing capabilities inherent in a world where all of the data is stored as a series of 0s and 1s. For the present, however, even the “digital” breast imaging practice will remain rooted in film, a world that will challenge the ability of adopters to live digitally, while continuing to accommodate film-based comparisons.

The Promise of Digital

Leaving, for a moment, the current challenges of FFDM, one of its main benefits is enhanced work flow. Using FFDM, the whole process of developing film is eliminated, quality assurance can be performed without the technologist leaving the room, and a mammogram can, realistically, be delivered in 56 minutes. This is image throughput, however, and that is not the whole story. Image throughput is the time from positioning through the availability of a set of images that can be interpreted by the radiologist, but does not consider the ancillary patient handling that is a part of every mammogram.

Image throughput for FFDM is approximately twice as fast as for film screen mammography. Therein lies both the benefit and the challenge of throughput. On the one hand, at 6 minutes per patient and 8 hours of operation per day, the FFDM unit can produce 20,000 mammograms in 250 workdays per year. Compared with the rapid-throughput 15-minute schedule for film screen mammography at approximately 7,500 per year, the increase in capacity is very substantial.

Utilizing this capacity is the first, and the major, long-term, implementation challenge. The fact that most mammography facilities underutilize a $100,000 piece of equipment is a far different matter than underutilization of equipment costing five times as much. The Advisory Board, Washington, DC, in a study published in 2002,1 found that the average mammography unit was  utilized only 65% of the available time. Simply put, this kind of inefficiency cannot be accommodated in a digital facility.

The Work-Flow Challenge

Utilizing the capacity of FFDM, however, requires more than just optimizing the current delivery paradigm and adding FFDM equipment. US patients will not tolerate being treated like cattle and rushed through a mammogram in the 6 minutes that may be possible with FFDM. Film-screen equipment has created a time interval that allows the patient and technologist to bond. Reassessing the patient history, small talk, and showing everyday concern for the patient take time and cannot be rushed. While there may be other ways to provide both rapid equipment throughput and a quality patient experience, room pairing has been found to be effective.

With this approach, the center uses two technologists, two adjacent rooms with a pass door between them, and the FFDM unit in one of the rooms. The patient is prepared by one of the technologists in one room, then taken into the FFDM room for her examination, and released when the examination is complete. One patient and one technologist are in each room at any given time and, to provide continuity, the technologist stays with her patient throughout the entire period. From the patient perspective, her mammogram appointment takes 12 minutes, even though a mammogram is being produced every 6 minutes by the FFDM equipment. Because the technologist never has to leave the patient, the quality of the patient experience is enhanced, despite the fact that total appointment time may actually be reduced.

Room pairing will work only if the delivery system has been fine-tuned to optimize the schedule. Scheduling must be refined to accurately segregate screening mammograms on one schedule from diagnostic examinations. Technologists must be committed to a fast-paced schedule and more “face” time with the patient than is included in the typical film-screen visit (there is no leaving the room to develop films). It is strongly recommended that film-screen programs optimize their delivery systems before implementing FFDM, and that FFDM be adopted with room pairing appointments at 20 minutes initially, reducing to 12 minutes within 3 months.

On the diagnostic side, FFDM can reduce the time normally associated with developing film studies; however, interaction with the interpreting physician is still required to determine which additional views may be required, and the number of those views. Because the physician has several digital tools at his or her disposal, the number of those views may be reduced, although at the cost of additional physician review time.

Digital Technical Requirements

The basic FFDM system consists of one or more acquisition units (including the acquisition workstation), one or more interpretation workstations, storage, and a high-quality laser printer.2 Facilities that already incorporate a PACS will need the appropriate connectivity software, and those wishing to incorporate computer-aided detection (CAD) in their review will need to obtain compatible software and hardware from the FFDM or CAD vendor.

Unfortunately for most users, the typical FFDM system will not be “basic.” Digital mammography files are very large, requiring both substantial storage capability, and rapid access media to allow the interpretation of large numbers of cases in succession without undue delay in presenting the images at the workstation. When planning the implementation of FFDM, it is of critical importance to consider the volume of studies that can be anticipated, both immediately and in the future, and the demands that this volume will place on storage and network capabilities. Many installations consider separate mini PACS for breast imaging.

The Challenge of Prior Studies

In the event that this had been part of the plan, do not get rid of that file room yet.  Beyond the statutory requirement to archive mammographic studies, consider the extent to which prior studies are used: (1) the technologist reviews the immediate prior year’s films for positioning and preexisting conditions (eg, density, surgical scars, etc) that may affect the way she sets up the mammography unit; and (2) films from 1, 2, or 3 years prior to the current study for comparison with the current study.

Given existing technology, the practical impact of the foregoing is that the interpretation of FFDM screening mammograms will exist in a world that includes films, viewers, and film files for the foreseeable future. Even if this was simply a matter of moving films and film files around, it would be problematic, but it is with respect to the actual temporal comparison of the current images with the prior films that the challenge moves beyond money to interpretive quality.

Film to film temporal comparison, particularly for films that have been taken and developed at the same site, presents relatively few problems. Images are the same size, exposures are typically close, and positioning is usually fairly comparable.  Differences, therefore, are relatively easy to identify, and changes in the size of tissue features can be recognized or calculated without a great deal of difficulty.

Mammography workstations, on the other hand, present a digitally acquired image that is quite different in appearance from the image captured on mammography film. Even when positioning is not an issue, the image is quite likely a different size, and appears as if it was exposed at quite different settings than the film images. Exacerbating this problem is the fact that the light illuminating the film images on a typical motorized viewer is at approximately three times the light level of the images on the workstation. The low light level of the current FFDM workstations also makes them very susceptible to monitor reflections off of the motorized viewer, and sensitive to environmental light in the reading room.

The result of these differences is to increase the time required for interpretation. The comparison that formerly was an easy 1:1 now requires mental adjustments.  While it would be nice to be able to say that this becomes automatic with time, there is no evidence to support this statement. When the object of the search is a subtle change in density that may indicate an invasive cancer, the stakes are simply too high.

Interpretive time for FFDM has been found to be approximately 2.5 times the time required for film screen studies.3 Part of this time is involved in the radiologist adjusting contrast and leveling to obtain the “best view” of the current studies, but a significant portion of the increase is the result of the challenge of performing a comparison of prior studies with the current digital study presented on a soft-copy monitor.

But this, too, will pass. Systems are currently being developed that will digitize prior film studies, and present them in the soft-copy environment in comparable form to the digitally acquired image.  Workstation technology will also improve to take advantage of postprocessing technology that will automatically review the “raw” digital image and present it to the interpreting physician with the appropriate adjustments to optimize the probability of finding cancers.

Benefits of The Digital World

The acquisition and implementation of FFDM technology represent taking a first step into the world of digital mammography.  The key to FFDM is the acquisition of an image that is digital. At its current market level, digital mammography is limited to the substitution of digital images for the four views in a conventional film screen mammogram. The “digital” image, however, raises the prospect of new methods of manipulating digital data as well as acquiring images that address the shortcomings of conventional mammography and present the opportunity to dramatically improve the ability of the interpreting physician to find cancers that have hitherto been masked or otherwise incapable of perception. CAD, of course, is made for FFDM,4 but it is only the beginning. Digital tomosynthesis (tomo)5,6 is clearly on the horizon and is showing great promise in its ability to minimize the effects of tissue density on finding and recognizing breast lesions. Contrast enhancement of FFDM and tomo studies can be expected to provide further diagnostic certainty, just as it has for other radiographic studies.

Practices and institutions that acquire and implement FFDM should do so with their eyes wide open to the realities and challenges presented by this new and exciting technology. FFDM is far from perfect, but it clearly represents the dawn of a new era in mammography that will see improvements in the interpretive value of mammograms and further reductions in the mortality rate from breast cancer.

Gerald R. Kolb, JD, is president of Breast Health Management Inc in Bend, Ore, [email protected].

References:

  1. The Advisory Board, Next-Generation Breast Centers. Washington, DC: Advisory Board Company; 2002.
  2. ACR Standard for the Performance of Whole Breast Digital Mammography, Section V, B, effective 1/1/02, American College of Radiology, Reston, Va. Available at: www.acr.org/dyna/?doc=frames/main-departments.html. Accessed June 10, 2003.
  3. Newstead G. Presentation at Mammography in the Next Millennium: Screening and Beyond; May 24, 2003; Santa Fe, NM.
  4. Baum F, Fischer U, Obenauer S, et al. Computer-aided detection in direct digital full-field mammography: initial results. Eur Radiol. 2002;12:3015-7.
  5. Wu T, Stewart A, Stanton M, et al. Tomographic mammography using a limited number of low-dose cone-beam projection images. Med Phys. 2003;30:365-80.
  6. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in breast imaging. Radiology. 1997;205:399-406.
  7. Diekmann F, Diekmann S, Taupitz M, et al. Use of iodine-based contrast media in digital full-field mammography??”initial experience. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr. 2003;175:342-5.