Digital Mammography: The Promise of Improved Breast Cancer Detection

Kevin D. Evans, PhD, RT(R)(M)(BD), RDMS, RVS, FSDMS

    ^*Director/Assistant Professor, Radiologic Sciences and Therapy Division, The Ohio State University, Columbus, Ohio.
    Address correspondence to: Kevin D. Evans, PhD, RT(R)(M)(BD), RDMS,RVS, FSDMS, Director/Assistant Professor, Radiologic Sciences and Therapy Division, The Ohio State University, 453 West 10th Avenue, 340 A. Atwell Hall, Columbus, OH 43210. E-mail: Kevin.Evans@osumc.edu.

ABSTRACT

The incidence of invasive breast cancer is now estimated to be approximately 13%. To realize the goal of early detection, breast cancer must be recognized not only at the tissue level but also at the cellular level. Film-screen mammography has improved detection of primary breast cancer and significantly decreased the rate of breast cancer mortality. Depending on the independent statistical model used to determine incidence and mortality, the mortality rate decrease ranges from 28% to 65%. Unfortunately, film-screen mammography only depicts changes at the breast tissue level.

Results of the Digital Mammographic Imaging Screening Trial (N = 49 500 women in the United States and Canada) reported that women younger than 50 years had 22 malignancies that were identified only by full-field digital mammography, although 6 cancers were found using film-screen mammography. This study represents the largest study of digital mammography to date.

Digital mammography holds the promise to provide better resolution on breast tissue, especially for women with dense breast tissue, who are typically younger than 50 years.

The goal of mammography is to provide early detection of breast cancer through low-dose imaging of the breast. The incidence of invasive breast cancer is now estimated to be approximately 1 in 8 (13%) women.¹ In 2006, approximately 212 920 new cases of invasive breast cancer were diagnosed among women in the United States. Increasing age is associated with an increasing risk for breast cancer, highlighting the need to provide an accurate imaging technique for screening women.

The primary goal of mammography is to accurately visualize the breast tissue, record the presence of any breast masses, and identify characteristics of calcifications or architectural distortion of the normal breast. It is also important to provide an accurate diagnosis based on the appearance of the breast tissue. To realize the goal of early detection, breast cancer must be recognized not only at the tissue level but also at the cellular level.

Film-screen mammography (FSM) has improved detection of primary breast cancer and significantly decreased the rate of breast cancer mortality. Depending on the independent statistical model used to determine incidence and mortality, the mortality rate decrease ranges from 28% to 65%.² Unfortunately, FSM only depicts changes at the breast tissue level. Specific mammographic features with the highest positive predictive value include masses with spiculated margins and/or irregular shape, in addition to calcifications with linear morphology and/or segmental distribution.³ Because many cancers begin at the cellular level, microscopic changes can pass undetected during the screening mammography examination.

Mammographic Features of Breast Cancer

Mammography is a diagnostic tool that plays a substantial role in documenting early changes in the disease process. It also demonstrates pathologic and metabolic processes that would otherwise be hidden. Finally, it provides evidence that prompts the pathologist to obtain a more exact diagnosis.⁴

Primary signs of breast cancer include mass, calcification, or a combination of both radiologic signs. Secondary signs include skin thickening, nipple inversion, adenopathy, skin densities, and architectural distortion.⁴ Physical presentations can be vital in developing an accurate diagnosis. Associated changes in the skin, nipples, contour, movement, and size of the breast are subtle findings that may impact the final diagnosis. Additionally, lumps and nodularity are physical features that may help in correlating the mammographic findings (Figure 1).⁴

Angiogenesis is a complex process occurring at the cellular level. As endothelial cells lining the ducts of breast tissue change and adopt a malignant character, the rapidly dividing cells demand an increased vascular supply. Angiogenesis is the development of neovascular networks of blood arising within the tissue to support malignant growth. Detection of this phase of breast cancer development means enhancing our imaging capability.

Full-field digital mammography (FFDM) has been developed to provide better visualization of tissue and the changes that may be occurring at a more cellular level. The development and release of FFDM has provided several well-devised studies to determine the accuracy and predictability of this new method for breast imaging.

Evidenced-Based Practice: Full-Field Digital Mammography

When performing mammography, it is essential to provide the highest image quality at the lowest possible radiation dose. Studies are currently under way to build scientific evidence to inform the practice of FFDM.

Patient Dose with Full-Field Digital Mammography
Ideally, the dose for a patient is expected to be lower with FFDM because it takes less time to capture an image digitally compared with FSM. The quality of each image is governed by the signal-to-noise ratio (SNR). The value of SNR is the square root of the patient dose. A study of an FFDM scanning phosphor system demonstrated an increased dose per view; however, it did not show a commensurate increase in the SNR.⁵ This study was one of the first to evaluate the use of higher tube potentials that produce X rays suitable for mammography. The breast phantom used in this study contained an outer layer of 0.5 cm thickness of fat that may have contributed to absorbing the lower energy photons, raising the mean glandular dose. Increasing filtration removes the lower radiation photons; therefore, making the beam contain higher energy.

The dose of radiation given during an FFDM can be controversial, but increasing resolution assists in distinguishing between different types of breast tissue, which is paramount. Providing a low-dose mammogram is useless if it does not provide detail that will promote the detection of small cancers that have not spread to the axillary lymph nodes.⁶ Rigorous study of patient dose with FFDM is important but must be tempered with the need to provide increased resolution.

Researchers in Germany conducted a study that compared FFDM and FSM dose differences by using different target and filter materials (tungsten and molybdenum) to produce digital and film images.⁷ The detection of microcalcifications was selected as the resolution benchmark for the comparison study. This small study found that a combination of molybdenum target and rhodium filter with 31 kVp and rhodium target and rhodium filter with 32 kVp were the most beneficial for imaging. These dose reductions staged by altering target and filter types in FFDM compared with FSM demonstrated no significant differences in detecting microcalcifications.⁷ This provides evidence that FFDM may be accomplished with a lower dose if the equipment has a higher potential for generating radiation.

The results of a small Italian study indicated that 4 machines tested provided consistent performance for output, in addition to sensitivity of the photo timing device.⁸ Additionally, the dose calculations conducted on test phantoms indicated that mean average glandular dose was higher compared with FSM levels.⁸ Additional studies need to be conducted and phantom models may help yield accurate radiation readings. Larger studies with varied types of FFDM equipment are still needed to inform practice patterns. An additional concern is that FFDM provides the ability to adjust image contrast that often obviates the need to repeat an image. Reduction of repeat images that stem from FSM could be the best way to demonstrate overall patient dose reductions.

Comparing FFDM and FSM Image Quality

Several studies have been conducted to compare the difference in image quality between FFDM and FSM.^9,10 A study conducted in 2002 provided a comparison of FFDM and FSM across 55 patients.⁹ The results of this comparison demonstrated that FFDM provided more consistent image contrast, substantially fewer overexposures, and no artifacts compared with FSM. Of the patients who were screened, malignancies were detected with equal accuracy; however, the radiologists were better able to classify the tumors with the FFDM images.^9,11 An analogous study also compared FFDM and FSM for the detection of microcalcifications in 55 different patients.¹¹ Each patient received an FFDM and an FSM of the area where these calcifications were located. The result was that FFDM provided 41% more calcifications than FSM.¹¹ The next study to rigorously provide a comparison of FFDM and FSM was conducted in Oslo, Norway, and recruited 3683 women aged 50 to 69.¹⁰ Each woman received 2 standard views with FFDM and FSM. The results demonstrated no statistical difference in breast cancer detection rate between the 2 modalities. Even with the 2 techniques having equal ability to detect abnormalities, the positive predictive value, based on needle biopsy, was 46% for FSM and 39% for FFDM.¹⁰ This indicates a need for continued refinement in screening mammography efficiency to accurately visualize small cancers at their earliest stage.

The utility of FFDM was confirmed by the Digital Mammographic Imaging Screening Trial, which reached its target goal of enrolling 49 500 women in the United States and Canada. Each participant had FFDM and FSM performed at the time of enrollment and was asked to return 1 year later for a follow-up mammogram.¹² The follow-up examination allowed researchers to assess each participant's breast health and gather key data for comparing the 2 technologies. This landmark study found that digital mammography was more sensitive in women younger than 50 years, women with dense breasts, and women in the perimenopausal and premenopausal age groups. Approximately 43 000 women were screened with digital and film mammography at 33 sites across the country. The diagnostic accuracy of FFDM and FSM was similar. Both technologies detected 122 malignancies; 52 were found exclusively on film and 63 were found only on FFDM. Digital mammography detected 3 additional invasive cancers and 8 cases of carcinoma in situ. The women younger than 50 years had 22 malignancies that were identified only by FFDM, although 6 cancers were found using FSM. This study represents the largest study of FFDM to date.

Film-Screen Mammography

Although FSM is the standard of reference for detecting breast cancer, approximately 10% to 20% of breast cancers detected during self-examination or physical examination are not visible during FSM.¹³ Additionally, only 5% to 40% of detected lesions found during FSM and recommended for biopsies are determined to be malignant.¹⁴ The high rate of false positives results in unnecessary biopsies and additional psychological strain on patients. Unfortunately, although FSM is effective, it is not highly sensitive or specific.

FSM exposes the breasts to low-dose radiation followed by transmission and scattering through the breast tissue. Following attenuation, the photons passing through the grid interact with the image receptor and are absorbed as a latent image on the recording device. The entire process is captured, displayed, and archived with the film, a single medium.

Full-Field Digital Mammography

Image Acquisition
The processes of image acquisition, display, and storage in FFDM are performed by distinct individual systems that may be optimized. The digital detector can be designed to efficiently absorb low-dose radiation, produce an electronic signal, digitize the signal, and store the results in computer memory. The output image is saved as a 2-dimensional matrix, where each element represents the radiation transmission corresponding to a particular path through the breast. This image can be digitally processed such that, when it is displayed in soft-copy form on a high-resolution computer monitor or printed on laser film, it will demonstrate the key features required for mammographic interpretation (Figure 2).

FFDM uses detectors that modify transmitted photons into electrical signals. These signals are then transferred to a digital receptor that converts the transmitted energy to numbers, processes the numbers, and produces an image that can be displayed on a monitor or printed on a high-resolution laser printer.

There are 3 types of detectors that are currently available for FFDM: photostimulable storage phosphors (PSP), active matrix detectors, and scanning systems. Active matrix detectors and scanning systems can be direct or indirect methods of image capture.¹⁵ The indirect method of image capture employs charge-coupled devices (CCDs), self-scanning devices that provide an electronic readout of an image produced by a scintillator. The image produced by the scintillator is then digitized to produce an image. The direct method of image capture involves the use of a photoconductor, typically amorphous selenium, which converts the photon to an electric charge that is processed as a digital signal.

Digital mammography has an intrinsically wider dynamic range, displaying all breast structures from dense parenchyma to skin. Because there is a lower level of intrinsic noise, a lower dose of X ray may be used than is currently used in conventional mammography with no loss in accuracy. The imaging process can also be used to create display "windows" for masses and calcifications (Figure 3).

Indirect Conversion

Photostimulable Storage Phosphors
Computed radiography systems similar to the Computed Radiography for Mammography system (Fujifilm Medical Systems USA, Inc, Stamford, CT) are widely used outside of mammography and employ, as the X-ray absorber, a phosphor screen that has a property called PSP. Energy from X-ray absorption causes electrons in the phosphor crystal to be temporarily freed from the crystal matrix and then captured and stored in traps within the crystal lattice. The number of filled traps is proportional to the absorbed X-ray signal. The image is then read out by placing the screen in a separate reader, where it is scanned with a red laser beam. This causes the electrons to be liberated from the traps and to return to their original resting state. In doing so, they may pass between energy levels in the crystal structure created by doping the crystal with certain materials. The difference in these energy levels corresponds to the energy of blue light, which is given off by the phosphor when such transitions occur. Thus, the amount of blue light emitted and measured by an optical collecting system and a photomultiplier tube is proportional to the energy of X rays absorbed by the phosphor. A filter in the optical chain prevents the stimulating red light from interfering with the measurement. The time at which the laser beam strikes a given location on the screen gives the x-y coordinate of each image location. In this system, the spatial sampling is determined by the size of the laser spot (detector-element size) and the distance between sample measurements (pitch). The spatial resolution for this type of system is 50 µm.¹⁵

Full-Field Active Matrix Detectors—Indirect
The flat-panel detectors involve the detection of light by a series of smaller (100 µm) photo diodes arrayed over a flat panel measuring 19 x 23 cm or 24 x 31 cm. The spatial resolution is 100 µm and can be used to cover the entire breast. In the flat-panel phosphor system, X rays are detected at a thallium (TI)-activated CsI, or CsI(TI) phosphor layer.¹⁶ The phosphor crystals are layered to form needlelike elements that reduce the spread of light. The phosphor crystals are deposited on an amorphous silicon flat plate that contains a rectangular array of photodiodes, which are used to record the emitted light from the phosphor. Each of the detector elements on the array contain the photodiode and a thin-film transistor switch. The control lines located in each row of the array are sequentially energized to activate all switches in the row. The latent image is released with a laser beam, scanning to de-energize the plate, thereby allowing the light to be detected by a photomultiplier tube that produces a digital image.

Scanning Phosphor-CCD System
In a scanning phosphor-CCD mammography system, the detector has a long, narrow, rectangular shape measuring approximately 1 x 24 cm.¹⁶ The X-ray beam is collimated to match the format. This technique depends on synchronization of the mechanical scanning system and the readout electronics.¹⁵ The phosphor in the scanning system is deposited on a coupling plate composed of optical fibers. These fibers conduct light from the phosphor to a CCD array that converts the light into an electronic signal that is digitized. In addition, the fiber optics block radiation that is not absorbed by the phosphor serving to protect the CCD from radiation damage. The actual CCD is a chip that contains rows and columns of light-sensitive elements. The charge on each element in response to the light exposure is transferred down the columns and read out in a single line. The X-ray beam and detector scan across the breast to capture the image. The total scanning time for the system is 5 to 6 seconds with a spatial resolution of 54 µm.¹⁵ Multi-slit scanning systems exist that provide direct conversion of the transmitted radiation to digital signal.

Direct Conversion

Full-Field Active Matrix Detectors—Direct
The X-ray absorber of this detector is composed of amorphous selenium. Selenium was used in xeroradiography and is still used in photocopiers today. When selenium absorbs X rays, an electric charge is liberated in the material. If electrodes are placed on the upper and lower surfaces of the selenium, and an electric field is applied between the electrodes, the charge signal can be collected onto a readout surface. This surface can be created on a plate of amorphous silicon similar to the phosphor flat-plate system. In this case, however, the photodiodes are replaced by a set of simple electrode pads to collect the charge. The spatial resolution for this system is 70 µm.¹⁵ Details are generally proprietary (Figure 4).

Digital Mammographic Image Display

The "soft-copy" image provides a direct-view display of the digital image on cathode ray tube monitors or liquid crystal display, allowing for unlimited manipulation of the contrast and brightness of the image, including image reversal and electronic magnification.

There are several important characteristics of FFDM display:

• Appropriate calibration
• Appropriate environment
• Transition from screen to digital

Although it is possible to display digital images in a hard-copy format, the advantages of digital technology cannot be fully realized without soft-copy display of mammograms, making electronic display a key and inseparable component of digital mammography.¹⁷ The quality of the display has a direct effect on radiologic interpretation. A faulty, inadequately calibrated, or improperly setup display device can compromise the overall quality of the diagnostic procedure.^18,19

Archive

The archive device for digital mammography must support the Digital Imaging and Communications in Medicine receipt of FFDM images. Images must be stored to allow radiologists to reproduce the original images that were used for the interpretation and archived to picture archiving and communication systems (PACS). Two important issues related to the archival of digital mammography are the need for storage capacity and multiple access. Storage capacity relates to the need for PACS to allow for the storage of several megabytes of data that are contained in 1 patient study. Because mammography uses the review of previous studies, the computer demand for retrieving previous digital studies, in addition to current studies, is quite taxing.

The demand for multiple users to access digital mammography images is also critical. Radiologists, pathologists, mammographers, and clerical staff need to have computer terminals available for viewing studies, presenting an additional load on the system. Studies should be available and easy to retrieve for everyone in the department. The quality of the study must be comparable on a multitude of workstations.

Digital mammogram compression can provide more efficient transmission and storage. The digital image is an exact representation of an inexact noisy signal, with finite limits to the amount of compression that can be applied. The advantages of compression include faster transmission time and smaller storage requirements.

Advantages and Disadvantages

Advantages to having digital mammography available for patients include:

• Elimination of processor artifacts
• Contrast enhancement
• Ability to perform invasive procedures faster
• Potentially better resolution of breast tissue for women younger than 50 years
• Reduce examination time for patients
• Increased production of examinations
• Images are immediately available

Disadvantages of this innovation in breast imaging include:

• Cost of the equipment
• Integration of the equipment into the department
• Comparison of FSM images with FFDM images for repeat patients
• Large megabyte images that must be handled by the computer and viewing stations

One of the most important diagnostic advantages to instituting FFDM is the use of computer-aided detection (CAD). CAD provides a software program that examines the digital image for increased density, such as microcalcifications. Using CAD allows for an "over read" or second opinion. The CAD program is consistent and unaffected by fatigue. Using a CAD program to increase the detection of pixel-sized differences in dense breast tissue is moving toward higher diagnostic efficiency.

Studies have been conducted on the use of CAD as a detection device and its ability to raise detection rates. Studies have also been conducted on the disagreement of interpreting physicians when viewing FFDM and FSM.²⁰ CAD provides an objective method to avoid a delay in treatment by having to secure a second opinion with difficult cases. Studies have also demonstrated an increase in the detection of microcalcifications.²¹

Computer-Aided Detection: Mammography

CAD systems use a digitized mammographic image obtained from FSM or FFDM. The computer software scans for abnormal areas of density, mass, or calcification and highlights these areas on the images. CAD allows for the location of digitally dramatic differences and highlights them. Highlighted areas overlaying the image alert the radiologist to the need for additional analysis. The radiologist decides whether more evaluation is needed in the area that is highlighted. With or without the CAD highlighting, the radiologist makes the diagnosis if a clinically significant abnormality exists, and decides whether further diagnostic evaluation is indicated.

Although evidence shows that the use of CAD with FSM is equal to or better than single reading of the FSM images, there is limited information on the performance of CAD with FFDM. The differences between film and digital mammography preclude extrapolating from the impact of CAD with FSM to CAD with FFDM. The large increase in the magnitude of the data collected by FFDM, the ability to fine-tune the digitally acquired images, and the elimination of the digitization step make FFDM sufficiently different from FSM, such that separate studies on the impact of CAD on FFDM are needed. Until results from better studies focusing on the use of CAD with FFDM become available, the benefits of CAD with FFDM cannot be determined.^22,23

Conclusions

Digital mammography holds the promise to provide better resolution on breast tissue, especially for women with dense breasts, who are typically younger than 50. A variety of digital image acquisition methods exist (indirect and direct) that may assist in providing increased resolution. Patient dose for FFDM continues to be an issue compared with FSM; however, the trade-off may be well worth the benefit of increased resolution. The high cost of FFDM will hopefully decrease with continued use and innovations by equipment vendors. CAD is a digital advantage that could help in raising the detection of tissue abnormalities at the level of pixel size. FFDM holds the key to providing detection of breast cancer at earlier stages, especially for younger women with dense breasts.

References

1. American Cancer Society. What are the key statistics for breast cancer? Available at: http://www.cancer.org/docroot/CRI/content/CRI_2_4_1X_What_are_the_key_statistics_for_breast_cancer_5.asp?sitearea. Accessed January 15, 2007.

2. Berry DA, Cronin KA, Plevritis SK, et al. Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med. 2005;353:1784-1792.

3. Liberman L, Abramson AF, Squires FB, et al. The breast imaging reporting and data system: positive predictive value of mammographic features and final assessment categories. AJR Am J Roentgenol. 1998;171:35-40.

4. Andolina VF, Willison KM, Lille SL. Mammographic Imaging: A Practical Guide. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001.

5. Court LE, Speller R. A multiparameter optimization of digital mammography. Phys Med Biol. 1995;40:1841-1861.

6. Haus AG, Yaffe MJ. Screen-film and digital mammography. Image quality and radiation dose considerations. Radiol Clin North Am. 2000;38:871-898.

7. Obenauer S, Hermann KP, Grabbe E. Dose reduction in full-field digital mammography: an anthropomorphic breast phantom study. Br J Radiol. 2003;76:478-482.

8. Gennaro G, Baldelli P, Taibi A, et al. Patient dose in full-field digital mammography: an Italian survey. Eur Radiol. 2004;14:645-652.

9. Obenauer S, Luftner-Nagel S, von HD, et al. Screen film vs full-field digital mammography: image quality, detectability and characterization of lesions. Eur Radiol. 2002;12:1697-1702.

10. Skaane P, Young K, Skjennald A. Population-based mammography screening: comparison of screen-film and full-field digital mammography with soft-copy reading. Radiology. 2003;229:877-884.

11. Fischer U, Baum F, Obenauer S, et al. Comparative study in patients with microcalcifications: full-field digital mammography vs screen-film mammography. Eur Radiol. 2002;12:2679-2683.

12. Pisano ED, Gatsonis CA, Yaffe MJ, et al. American College of Radiology Imaging Network digital mammographic imaging screening trial: objectives and methodology. Radiology. 2005;236:404-412.

13. Mahesh M. AAPM/RSNA physics tutorial for residents. Digital mammography: an overview. Radiographics. 2004;24:1747-1760.

14. Burrell HC, Pinder SE, Wilson AR, et al. The positive predictive value of mammographic signs: a review of 425 non-palpable breast lesions. Clin Radiol. 1996;51:277-281.

15. Noel A, Thibault F. Digital detectors for mammography: the technical challenges. Eur Radiol. 2004;14:1990-1998.

16. Pisano ED, Yaffe MJ. Digital mammography. Radiology. 2005;234:353-362.

17. Badano A, Flynn MJ, et al. High Fidelity Medical Imaging Displays. Bellingham, WA: SPIE Press; 2004.

18. Badano A, Flynn MJ, Martin S, Kanicki J. Angular dependence of the luminance and contrast in medical monochrome liquid crystal displays. Med Phys. 2003;30:2602-2613.

19. Baker JA, Rosen EL, Crockett MM, Lo JY. Accuracy of segmentation of a commercial computer-aided detection system for mammography. Radiology. 2005;235:385-390.

20. Venta LA, Hendrick RE, Adler YT, et al. Rates and causes of disagreement in interpretation of full-field digital mammography and film-screen mammography in a diagnostic setting. AJR Am J Roentgenol. 2001;176:1241-1248.

21. Kegelmeyer WP Jr, Pruneda JM, Bourland PD, Hillis A, et al. Computer-aided mammographic screening for spiculated lesions. Radiology. 1994;191:331-337.

22. Obenauer S, Sohns C, Werner C, Grabbe E. Computer-aided detection in full-field digital mammography: detection in dependence of the BI-RADS categories. Breast J. 2006;12:16-19.

23. Obenauer S, Sohns C, Werner C, Grabbe E. Impact of breast density on computer-aided detection in full-field digital mammography. J Digit Imaging. 2006;19:258-263.